Compare commits

..

No commits in common. "master" and "v0.12.11" have entirely different histories.

285 changed files with 10745 additions and 30363 deletions

View File

@ -5,18 +5,18 @@ info:
repository_url: https://github.com/dosco/super-graph
options:
commits:
filters:
Type:
- feat
- fix
- perf
- refactor
# filters:
# Type:
# - feat
# - fix
# - perf
# - refactor
commit_groups:
title_maps:
feat: Features
fix: Bug Fixes
perf: Performance Improvements
refactor: Code Refactoring
# title_maps:
# feat: Features
# fix: Bug Fixes
# perf: Performance Improvements
# refactor: Code Refactoring
header:
pattern: "^((\\w+)\\s.*)$"
pattern_maps:

View File

@ -1,8 +0,0 @@
version = 1
[[analyzers]]
name = "go"
enabled = true
[analyzers.meta]
import_path = "github.com/dosco/super-graph"

11
.gitignore vendored
View File

@ -23,20 +23,15 @@
/tmp/runner-build
/demo/tmp
.idea
*.iml
.vscode
main
.DS_Store
.swp
.release
main
super-graph
supergraph
*-fuzz.zip
crashers
suppressions
release
.gofuzz
*-fuzz.zip
*.test
.firebase
release

View File

@ -1,371 +1,401 @@
<a name="unreleased"></a>
## [Unreleased]
<a name="v0.13.22"></a>
## [v0.13.22] - 2020-05-01
<a name="v0.13.21"></a>
## [v0.13.21] - 2020-04-24
<a name="v0.13.20"></a>
## [v0.13.20] - 2020-04-24
<a name="v0.13.19"></a>
## [v0.13.19] - 2020-04-23
<a name="v0.13.18"></a>
## [v0.13.18] - 2020-04-23
<a name="v0.13.17"></a>
## [v0.13.17] - 2020-04-22
<a name="v0.13.16"></a>
## [v0.13.16] - 2020-04-21
### Features
- feat : improve the generated introspection schema and avoid the chirino/graphql api leaking through the core api. ([#53](https://github.com/dosco/super-graph/issues/53))
<a name="v0.13.15"></a>
## [v0.13.15] - 2020-04-20
<a name="v0.13.14"></a>
## [v0.13.14] - 2020-04-19
<a name="v0.13.13"></a>
## [v0.13.13] - 2020-04-19
<a name="v0.13.12"></a>
## [v0.13.12] - 2020-04-19
<a name="v0.13.11"></a>
## [v0.13.11] - 2020-04-18
<a name="v0.13.10"></a>
## [v0.13.10] - 2020-04-17
<a name="v0.13.9"></a>
## [v0.13.9] - 2020-04-16
<a name="v0.13.8"></a>
## [v0.13.8] - 2020-04-16
<a name="v0.13.7"></a>
## [v0.13.7] - 2020-04-16
<a name="v0.13.6"></a>
## [v0.13.6] - 2020-04-13
<a name="v0.13.5"></a>
## [v0.13.5] - 2020-04-13
<a name="v0.13.4"></a>
## [v0.13.4] - 2020-04-12
<a name="v0.13.3"></a>
## [v0.13.3] - 2020-04-12
<a name="v0.13.2"></a>
## [v0.13.2] - 2020-04-11
<a name="v0.13.1"></a>
## [v0.13.1] - 2020-04-11
<a name="v0.13.0"></a>
## [v0.13.0] - 2020-04-10
<a name="v0.12.49"></a>
## [v0.12.49] - 2020-04-01
<a name="v0.12.48"></a>
## [v0.12.48] - 2020-03-31
<a name="v0.12.47"></a>
## [v0.12.47] - 2020-03-30
<a name="v0.12.46"></a>
## [v0.12.46] - 2020-03-21
<a name="v0.12.45"></a>
## [v0.12.45] - 2020-03-18
<a name="v0.12.44"></a>
## [v0.12.44] - 2020-03-16
<a name="v0.12.43"></a>
## [v0.12.43] - 2020-03-16
<a name="v0.12.42"></a>
## [v0.12.42] - 2020-03-14
<a name="v0.12.41"></a>
## [v0.12.41] - 2020-03-06
<a name="v0.12.40"></a>
## [v0.12.40] - 2020-03-06
<a name="v0.12.39"></a>
## [v0.12.39] - 2020-03-06
<a name="v0.12.38"></a>
## [v0.12.38] - 2020-03-05
<a name="v0.12.37"></a>
## [v0.12.37] - 2020-03-04
<a name="v0.12.36"></a>
## [v0.12.36] - 2020-03-04
<a name="v0.12.35"></a>
## [v0.12.35] - 2020-03-03
<a name="v0.12.34"></a>
## [v0.12.34] - 2020-03-03
<a name="v0.12.33"></a>
## [v0.12.33] - 2020-02-29
<a name="v0.12.32"></a>
## [v0.12.32] - 2020-02-24
### Bug Fixes
- fix "Try the demo app" in docs ([#38](https://github.com/dosco/super-graph/issues/38))
<a name="v0.12.31"></a>
## [v0.12.31] - 2020-02-23
<a name="v0.12.30"></a>
## [v0.12.30] - 2020-02-23
<a name="v0.12.29"></a>
## [v0.12.29] - 2020-02-21
<a name="v0.12.28"></a>
## [v0.12.28] - 2020-02-20
<a name="v0.12.27"></a>
## [v0.12.27] - 2020-02-19
<a name="v0.12.26"></a>
## [v0.12.26] - 2020-02-11
<a name="v0.12.25"></a>
## [v0.12.25] - 2020-02-10
<a name="v0.12.24"></a>
## [v0.12.24] - 2020-02-03
<a name="v0.12.23"></a>
## [v0.12.23] - 2020-02-02
<a name="v0.12.22"></a>
## [v0.12.22] - 2020-02-01
<a name="v0.12.21"></a>
## [v0.12.21] - 2020-01-31
<a name="v0.12.20"></a>
## [v0.12.20] - 2020-01-28
<a name="v0.12.19"></a>
## [v0.12.19] - 2020-01-26
<a name="v0.12.18"></a>
## [v0.12.18] - 2020-01-20
<a name="v0.12.17"></a>
## [v0.12.17] - 2020-01-20
<a name="v0.12.16"></a>
## [v0.12.16] - 2020-01-19
<a name="v0.12.15"></a>
## [v0.12.15] - 2020-01-17
<a name="v0.12.14"></a>
## [v0.12.14] - 2020-01-17
<a name="v0.12.13"></a>
## [v0.12.13] - 2020-01-16
<a name="v0.12.12"></a>
## [v0.12.12] - 2020-01-15
<a name="v0.12.11"></a>
## [v0.12.11] - 2020-01-14
<a name="v0.12.10"></a>
## [v0.12.10] - 2020-01-14
<a name="v0.12.9"></a>
## [v0.12.9] - 2020-01-14
<a name="v0.12.8"></a>
## [v0.12.8] - 2020-01-13
<a name="v0.12.7"></a>
## [v0.12.7] - 2020-01-11
### Pull Requests
- Merge pull request [#22](https://github.com/dosco/super-graph/issues/22) from bhaskarmurthy/fix-grammer-syntax
### Add
- Add config driven custom table relationships
- Add support for `websearch_to_tsquery` in PG 11
### Create
- Create CODE_OF_CONDUCT.md
### Fix
- Fix bug with remote join example
- Fix grammer / syntax
### Update
- Update issue templates
- Update CONTRIBUTING.md
- Update issue templates
- Update feature_request.md
<a name="v0.12.6"></a>
## [v0.12.6] - 2019-12-02
### Add
- Add support for `websearch_to_tsquery` in PG 11
<a name="v0.12.5"></a>
## [v0.12.5] - 2019-11-30
### Add
- Add a guide to the internals of the codebase
- Add a CONTRIBUTING.md guide for contributors
- Add a CHANGLOG.md
- Add issue templates
### Fix
- Fix for missing filters on nested selectors
### Refactor
- Refactor rename 'Select.Table` to `Select.Name`
<a name="v0.12.4"></a>
## [v0.12.4] - 2019-11-28
### Move
- Move license from MIT to Apache 2.0. Add Makefile
<a name="v0.12.3"></a>
## [v0.12.3] - 2019-11-26
### Added
- Added support for query names to the allow.list
<a name="v0.12.2"></a>
## [v0.12.2] - 2019-11-25
### Fix
- Fix bug with compiling anon queries
<a name="v0.12.1"></a>
## [v0.12.1] - 2019-11-22
### Move
- Move sql query logging from info to debug
<a name="v0.12.0"></a>
## [v0.12.0] - 2019-11-22
### Use
- Use logger error instead of panic in goja handlers
<a name="v0.11.9"></a>
## [v0.11.9] - 2019-11-22
### Add
- Add a db:reset command only for dev mode
<a name="v0.11.8"></a>
## [v0.11.8] - 2019-11-21
### Optimize
- Optimize db queries limit use of transactions
<a name="v0.11.7"></a>
## [v0.11.7] - 2019-11-19
### Added
- Added support for multi-root queries
<a name="v0.11.6"></a>
## [v0.11.6] - 2019-11-15
### Fix
- Fix issues with JWT auth
- Fix bug with migration filename generation
- Fix bug with migration file name
<a name="v0.11.5"></a>
## [v0.11.5] - 2019-11-10
### Fix
- Fix bug with migration template name
<a name="v0.11.4"></a>
## [v0.11.4] - 2019-11-10
### Fix
- Fix bug with creating new migrations
<a name="v0.11.3"></a>
## [v0.11.3] - 2019-11-09
### Fix
- Fix macro syntax bug in app templates
<a name="v0.11.2"></a>
## [v0.11.2] - 2019-11-07
### Fix
- Fix bugs and add new production mode
<a name="v0.11.1"></a>
## [v0.11.1] - 2019-11-05
### Add
- Add nested where clause to filter based on related tables
### Block
- Block unauthorized requests when 'anon' role is not defined
### Update
- Update docs and website with new features
<a name="v0.11"></a>
## [v0.11] - 2019-11-01
### Add
- Add config driven presets for insert, update and upsert
- Add config driven presets for insert, update and upserta
- Add RBAC option to disable functions eg. count
- Add fuzz testing to 'serv' for the GQL hash parser
- Add fuzz testing to 'jsn' and 'qcode'
- Add ability to block queries and mutations by role
- Add built in 'anon' and 'user' roles
- Add role based access control
### Allow
- Allow config files to inherit from other config files
### Change
- Change config key inherit to inherits
### Get
- Get RBAC working for queries and mutations
### Optimize
- Optimize prepared statement flow for RBAC
### Preserve
- Preserve allow.list ordering on save
### Update
- Update filters section in guide
### Pull Requests
- Merge pull request [#11](https://github.com/dosco/super-graph/issues/11) from dosco/rbac
<a name="v0.10.1"></a>
## [v0.10.1] - 2019-10-06
### Add
- Add ability to set filters per operation / action
- Add upsert mutation
### Pull Requests
- Merge pull request [#10](https://github.com/dosco/super-graph/issues/10) from FourSigma/sm-examples-folder
<a name="v0.10"></a>
## [v0.10] - 2019-10-04
### Fix
- Fix return values for bulk mutations and delete
- Fix issues with mutation SQL
- Fix broken demo app
- Fix typo in 'across'
### Remove
- Remove extra link from README
### Update
- Update docs, getting started guide and mutations
### Pull Requests
- Merge pull request [#6](https://github.com/dosco/super-graph/issues/6) from muesli/typo-fixes
<a name="v0.9"></a>
## [v0.9] - 2019-10-01
### Fix
- Fix demo rails app broken build
<a name="v0.8"></a>
## [v0.8] - 2019-09-30
### Fix
- Fix invalid import bug
### Update
- Update documentation site
<a name="v0.7"></a>
## [v0.7] - 2019-09-29
### Failure
- Failure to prepare statements should be a warning
### Fix
- Fix duplicte column bug
<a name="v0.6"></a>
## [v0.6] - 2019-09-29
### Add
- Add database setup commands
- Add binary compression back to Dockerfile
- Add initialization command to setup new apps
- Add migrate command
- Add database seeding capability
- Add session variable for user id
- Add delete mutation
- Add update mutation
- Add insert mutation with bulk insert
- Add GoTO Aug, 19 presentation
- Add support for prepared statements
- Add end-to-end benchmaking
- Add object pooling for parser expressions
- Add request / response debugging for remote joins
- Add a presentation about GraphQL
- Add validation for remote JSON
- Add tracing for API stitching
- Add REST API stitching
- Add SQL query cacheing
- Add support for GraphQL variables
- Add fuzz testing to qcode
- Add test for Rails Redis cookie store integration
- Add an install guide
### Change
- Change fuzz test name to qcode
- Change logo from PNG to SVG
### Enabke
- Enabke reload on config change
### Fix
- Fix missing config name bug
- Fix new app templates
- Fix help message for migrate
- Fix session variable bug
- Fix test failures in `psql` and `serv`
- Fix demo docker services startup order
- Fix wrong value for false token bug. Reported by [@ThisIsMissEm](https://github.com/ThisIsMissEm)
- Fix allow.list file discovery bug
- Fix bug with allow list path
- Fix wrong value for use_allow_list in dev config
- Fix startup bug in demo script
- Fix url bug in allow list
- Fix bug [#676](https://github.com/dosco/super-graph/issues/676) found by fuzzer
- Fix race-condition in remote joins
- Fix cookie passing in web ui
- Fix bug with passing cookies in web ui
- Fix null pointer with invalid argument values
- Fix infinite loop bug in lexer
- Fix null pointer issue found by fuzz test
- Fix issue with fuzzbuzz config
- Fix demo to run as memory only
- Fix auth documentation
- Fix issue with web ui sizing
- Fix issue preventing docker-compose deploy
- Fix try demo documentation
### Futher
- Futher reduce allocations across hot paths
- Futher reduce allocations on the compiler hot path
- Futher optimize json parsing and editing performance
### Highlight
- Highlight top features better on the site
### Improve
- Improve readability of json parser code
- Improve the motivation section in the readme
- Improve the demo experience
### Make
- Make remote joins use parallel http requests
### Merge
- Merge branch 'master' into optimize-psql
### New
- New low allocation fast json parsing and editing library
### Optimize
- Optimize lexer and fix bugs
- Optimize the sql generator hot path
### Reduce
- Reduce alllocations done by the stack
- Reduce steps to run the demo
- Reduce allocations and improve perf over 50%
### Remove
- Remove unused packages
- Remove the 'hello' test app folder
- Remove other allocations in psql
### Use
- Use hash's as ids for table relationships
### Watch
- Watch and reload on config changes
<a name="v0.5"></a>
## [v0.5] - 2019-04-10
### Add
- Add supprt for new Rails 5.2 aes-256-gcm cookies
- Add query support for ts_rank and ts_headline
- Add full text search support using TSV indexes
- Add missing assets folder
- Add fetch by ID feature
- Add documentation
### Cleanup
- Cleanup and redesign config files
### Fix
- Fix bug with auth config parsing
### Redesign
- Redesign config file architecture
### Reduce
- Reduce realloc of maps and slices
### Update
- Update docs with full-text search information
<a name="v0.4"></a>
## [v0.4] - 2019-04-01
<a name="v0.3"></a>
## [v0.3] - 2019-04-01
### Add
- Add SQL execution timing and tracing
- Add support for HAVING with aggregate queries
- Add aggregrate functions to GQL queries
- Add Auth0 JWT support
- Add React UI building to the docker build flow
- Add compiler profiling
- Add bechmarks for GQL to SQL compile
- Add tests for gql to sql compile
### Cleanup
- Cleanup Dockerfile
### Fix
- Fix recurring packer issue docker hub builds
- Fix issue with asset packer breaking Docker builds
- Fix missing git package in Dockerfile
- Fix docker ignore values
- Fix image build failure on docker hub
- Fix build issue in Dockerfile
- Fix bugs and document the 'where' clause
- Fix perf issue with inflections
### Optimize
- Optimize docker image
### Pack
- Pack web UI with app into a single binary
### Upgrade
- Upgrade web UI packages
<a name="0.3"></a>
## 0.3 - 2019-03-24
### First
- First commit
[Unreleased]: https://github.com/dosco/super-graph/compare/v0.13.22...HEAD
[v0.13.22]: https://github.com/dosco/super-graph/compare/v0.13.21...v0.13.22
[v0.13.21]: https://github.com/dosco/super-graph/compare/v0.13.20...v0.13.21
[v0.13.20]: https://github.com/dosco/super-graph/compare/v0.13.19...v0.13.20
[v0.13.19]: https://github.com/dosco/super-graph/compare/v0.13.18...v0.13.19
[v0.13.18]: https://github.com/dosco/super-graph/compare/v0.13.17...v0.13.18
[v0.13.17]: https://github.com/dosco/super-graph/compare/v0.13.16...v0.13.17
[v0.13.16]: https://github.com/dosco/super-graph/compare/v0.13.15...v0.13.16
[v0.13.15]: https://github.com/dosco/super-graph/compare/v0.13.14...v0.13.15
[v0.13.14]: https://github.com/dosco/super-graph/compare/v0.13.13...v0.13.14
[v0.13.13]: https://github.com/dosco/super-graph/compare/v0.13.12...v0.13.13
[v0.13.12]: https://github.com/dosco/super-graph/compare/v0.13.11...v0.13.12
[v0.13.11]: https://github.com/dosco/super-graph/compare/v0.13.10...v0.13.11
[v0.13.10]: https://github.com/dosco/super-graph/compare/v0.13.9...v0.13.10
[v0.13.9]: https://github.com/dosco/super-graph/compare/v0.13.8...v0.13.9
[v0.13.8]: https://github.com/dosco/super-graph/compare/v0.13.7...v0.13.8
[v0.13.7]: https://github.com/dosco/super-graph/compare/v0.13.6...v0.13.7
[v0.13.6]: https://github.com/dosco/super-graph/compare/v0.13.5...v0.13.6
[v0.13.5]: https://github.com/dosco/super-graph/compare/v0.13.4...v0.13.5
[v0.13.4]: https://github.com/dosco/super-graph/compare/v0.13.3...v0.13.4
[v0.13.3]: https://github.com/dosco/super-graph/compare/v0.13.2...v0.13.3
[v0.13.2]: https://github.com/dosco/super-graph/compare/v0.13.1...v0.13.2
[v0.13.1]: https://github.com/dosco/super-graph/compare/v0.13.0...v0.13.1
[v0.13.0]: https://github.com/dosco/super-graph/compare/v0.12.49...v0.13.0
[v0.12.49]: https://github.com/dosco/super-graph/compare/v0.12.48...v0.12.49
[v0.12.48]: https://github.com/dosco/super-graph/compare/v0.12.47...v0.12.48
[v0.12.47]: https://github.com/dosco/super-graph/compare/v0.12.46...v0.12.47
[v0.12.46]: https://github.com/dosco/super-graph/compare/v0.12.45...v0.12.46
[v0.12.45]: https://github.com/dosco/super-graph/compare/v0.12.44...v0.12.45
[v0.12.44]: https://github.com/dosco/super-graph/compare/v0.12.43...v0.12.44
[v0.12.43]: https://github.com/dosco/super-graph/compare/v0.12.42...v0.12.43
[v0.12.42]: https://github.com/dosco/super-graph/compare/v0.12.41...v0.12.42
[v0.12.41]: https://github.com/dosco/super-graph/compare/v0.12.40...v0.12.41
[v0.12.40]: https://github.com/dosco/super-graph/compare/v0.12.39...v0.12.40
[v0.12.39]: https://github.com/dosco/super-graph/compare/v0.12.38...v0.12.39
[v0.12.38]: https://github.com/dosco/super-graph/compare/v0.12.37...v0.12.38
[v0.12.37]: https://github.com/dosco/super-graph/compare/v0.12.36...v0.12.37
[v0.12.36]: https://github.com/dosco/super-graph/compare/v0.12.35...v0.12.36
[v0.12.35]: https://github.com/dosco/super-graph/compare/v0.12.34...v0.12.35
[v0.12.34]: https://github.com/dosco/super-graph/compare/v0.12.33...v0.12.34
[v0.12.33]: https://github.com/dosco/super-graph/compare/v0.12.32...v0.12.33
[v0.12.32]: https://github.com/dosco/super-graph/compare/v0.12.31...v0.12.32
[v0.12.31]: https://github.com/dosco/super-graph/compare/v0.12.30...v0.12.31
[v0.12.30]: https://github.com/dosco/super-graph/compare/v0.12.29...v0.12.30
[v0.12.29]: https://github.com/dosco/super-graph/compare/v0.12.28...v0.12.29
[v0.12.28]: https://github.com/dosco/super-graph/compare/v0.12.27...v0.12.28
[v0.12.27]: https://github.com/dosco/super-graph/compare/v0.12.26...v0.12.27
[v0.12.26]: https://github.com/dosco/super-graph/compare/v0.12.25...v0.12.26
[v0.12.25]: https://github.com/dosco/super-graph/compare/v0.12.24...v0.12.25
[v0.12.24]: https://github.com/dosco/super-graph/compare/v0.12.23...v0.12.24
[v0.12.23]: https://github.com/dosco/super-graph/compare/v0.12.22...v0.12.23
[v0.12.22]: https://github.com/dosco/super-graph/compare/v0.12.21...v0.12.22
[v0.12.21]: https://github.com/dosco/super-graph/compare/v0.12.20...v0.12.21
[v0.12.20]: https://github.com/dosco/super-graph/compare/v0.12.19...v0.12.20
[v0.12.19]: https://github.com/dosco/super-graph/compare/v0.12.18...v0.12.19
[v0.12.18]: https://github.com/dosco/super-graph/compare/v0.12.17...v0.12.18
[v0.12.17]: https://github.com/dosco/super-graph/compare/v0.12.16...v0.12.17
[v0.12.16]: https://github.com/dosco/super-graph/compare/v0.12.15...v0.12.16
[v0.12.15]: https://github.com/dosco/super-graph/compare/v0.12.14...v0.12.15
[v0.12.14]: https://github.com/dosco/super-graph/compare/v0.12.13...v0.12.14
[v0.12.13]: https://github.com/dosco/super-graph/compare/v0.12.12...v0.12.13
[v0.12.12]: https://github.com/dosco/super-graph/compare/v0.12.11...v0.12.12
[v0.12.11]: https://github.com/dosco/super-graph/compare/v0.12.10...v0.12.11
[v0.12.10]: https://github.com/dosco/super-graph/compare/v0.12.9...v0.12.10
[v0.12.9]: https://github.com/dosco/super-graph/compare/v0.12.8...v0.12.9
[v0.12.8]: https://github.com/dosco/super-graph/compare/v0.12.7...v0.12.8
[v0.12.7]: https://github.com/dosco/super-graph/compare/v0.12.6...v0.12.7
### Fix
- Fix license to MIT
[Unreleased]: https://github.com/dosco/super-graph/compare/v0.12.6...HEAD
[v0.12.6]: https://github.com/dosco/super-graph/compare/v0.12.5...v0.12.6
[v0.12.5]: https://github.com/dosco/super-graph/compare/v0.12.4...v0.12.5
[v0.12.4]: https://github.com/dosco/super-graph/compare/v0.12.3...v0.12.4

View File

@ -1,38 +1,30 @@
# stage: 1
FROM node:10 as react-build
WORKDIR /web
COPY /internal/serv/web/ ./
COPY web/ ./
RUN yarn
RUN yarn build
# stage: 2
FROM golang:1.14-alpine as go-build
FROM golang:1.13.4-alpine as go-build
RUN apk update && \
apk add --no-cache make && \
apk add --no-cache git && \
apk add --no-cache jq
apk add --no-cache upx=3.95-r2
RUN GO111MODULE=off go get -u github.com/rafaelsq/wtc
ARG SOPS_VERSION=3.5.0
ADD https://github.com/mozilla/sops/releases/download/v${SOPS_VERSION}/sops-v${SOPS_VERSION}.linux /usr/local/bin/sops
RUN chmod 755 /usr/local/bin/sops
WORKDIR /app
COPY . /app
RUN mkdir -p /app/internal/serv/web/build
COPY --from=react-build /web/build/ ./internal/serv/web/build
RUN mkdir -p /app/web/build
COPY --from=react-build /web/build/ ./web/build/
RUN go mod vendor
RUN make build
# RUN echo "Compressing binary, will take a bit of time..." && \
# upx --ultra-brute -qq super-graph && \
# upx -t super-graph
RUN echo "Compressing binary, will take a bit of time..." && \
upx --ultra-brute -qq super-graph && \
upx -t super-graph
# stage: 3
FROM alpine:latest
@ -44,15 +36,10 @@ RUN mkdir -p /config
COPY --from=go-build /etc/ssl/certs/ca-certificates.crt /etc/ssl/certs/
COPY --from=go-build /app/config/* /config/
COPY --from=go-build /app/super-graph .
COPY --from=go-build /app/internal/scripts/start.sh .
COPY --from=go-build /usr/local/bin/sops .
RUN chmod +x /super-graph
RUN chmod +x /start.sh
USER nobody
ENV GO_ENV production
EXPOSE 8080
ENTRYPOINT ["./start.sh"]
CMD ["./super-graph", "serv"]
CMD ./super-graph serv

View File

@ -12,30 +12,30 @@ endif
export GO111MODULE := on
# Build-time Go variables
version = github.com/dosco/super-graph/internal/serv.version
gitBranch = github.com/dosco/super-graph/internal/serv.gitBranch
lastCommitSHA = github.com/dosco/super-graph/internal/serv.lastCommitSHA
lastCommitTime = github.com/dosco/super-graph/internal/serv.lastCommitTime
version = github.com/dosco/super-graph/serv.version
gitBranch = github.com/dosco/super-graph/serv.gitBranch
lastCommitSHA = github.com/dosco/super-graph/serv.lastCommitSHA
lastCommitTime = github.com/dosco/super-graph/serv.lastCommitTime
BUILD_FLAGS ?= -ldflags '-s -w -X ${lastCommitSHA}=${BUILD} -X "${lastCommitTime}=${BUILD_DATE}" -X "${version}=${BUILD_VERSION}" -X ${gitBranch}=${BUILD_BRANCH}'
.PHONY: all build gen clean test run lint changlog release version help $(PLATFORMS)
test:
@go test -v -short -race ./...
@go test -v ./...
BIN_DIR := $(GOPATH)/bin
GORICE := $(BIN_DIR)/rice
GOLANGCILINT := $(BIN_DIR)/golangci-lint
GITCHGLOG := $(BIN_DIR)/git-chglog
WEB_BUILD_DIR := ./internal/serv/web/build/manifest.json
WEB_BUILD_DIR := ./web/build/manifest.json
$(GORICE):
@GO111MODULE=off go get -u github.com/GeertJohan/go.rice/rice
$(WEB_BUILD_DIR):
@echo "First install Yarn and create a build of the web UI then re-run make install"
@echo "Run this command: yarn --cwd internal/serv/web/ build"
@echo "First install Yarn and create a build of the web UI found under ./web"
@echo "Command: cd web && yarn build"
@exit 1
$(GITCHGLOG):
@ -45,7 +45,7 @@ changelog: $(GITCHGLOG)
@git-chglog $(ARGS)
$(GOLANGCILINT):
@GO111MODULE=off curl -sfL https://raw.githubusercontent.com/golangci/golangci-lint/master/install.sh| sh -s -- -b $(GOPATH)/bin v1.25.1
@GO111MODULE=off curl -sfL https://raw.githubusercontent.com/golangci/golangci-lint/master/install.sh| sh -s -- -b $(GOPATH)/bin v1.21.0
lint: $(GOLANGCILINT)
@golangci-lint run ./... --skip-dirs-use-default
@ -57,7 +57,7 @@ os = $(word 1, $@)
$(PLATFORMS): lint test
@mkdir -p release
@GOOS=$(os) GOARCH=amd64 go build $(BUILD_FLAGS) -o release/$(BINARY)-$(BUILD_VERSION)-$(os)-amd64 main.go
@GOOS=$(os) GOARCH=amd64 go build $(BUILD_FLAGS) -o release/$(BINARY)-$(BUILD_VERSION)-$(os)-amd64
release: windows linux darwin
@ -69,7 +69,7 @@ gen: $(GORICE) $(WEB_BUILD_DIR)
@go generate ./...
$(BINARY): clean
@go build $(BUILD_FLAGS) -o $(BINARY) main.go
@go build $(BUILD_FLAGS) -o $(BINARY)
clean:
@rm -f $(BINARY)
@ -77,10 +77,11 @@ clean:
run: clean
@go run $(BUILD_FLAGS) main.go $(ARGS)
install: clean build
install:
@echo $(GOPATH)
@echo "Commit Hash: `git rev-parse HEAD`"
@echo "Old Hash: `shasum $(GOPATH)/bin/$(BINARY) 2>/dev/null | cut -c -32`"
@mv $(BINARY) $(GOPATH)/bin/$(BINARY)
@go install $(BUILD_FLAGS)
@echo "New Hash:" `shasum $(GOPATH)/bin/$(BINARY) 2>/dev/null | cut -c -32`
uninstall: clean

101
README.md
View File

@ -1,107 +1,72 @@
<img src="docs/website/static/img/super-graph-logo.svg" width="80" />
<!-- <a href="https://supergraph.dev"><img src="https://supergraph.dev/hologram.svg" width="100" height="100" align="right" /></a> -->
# Super Graph - Fetch data without code!
<img src="docs/.vuepress/public/super-graph.png" width="250" />
[![GoDoc](https://img.shields.io/badge/godoc-reference-5272B4.svg)](https://pkg.go.dev/github.com/dosco/super-graph/core?tab=doc)
![Apache 2.0](https://img.shields.io/github/license/dosco/super-graph.svg?style=flat-square)
![Docker build](https://img.shields.io/docker/cloud/build/dosco/super-graph.svg?style=flat-square)
[![Discord Chat](https://img.shields.io/discord/628796009539043348.svg)](https://discord.gg/6pSWCTZ)
### Build web products faster. Secure high performance GraphQL
Super Graph gives you a high performance GraphQL API without you having to write any code. GraphQL is automagically compiled into an efficient SQL query. Use it either as a library or a standalone service.
![Apache Public License 2.0](https://img.shields.io/github/license/dosco/super-graph.svg)
![Docker build](https://img.shields.io/docker/cloud/build/dosco/super-graph.svg)
![Cloud native](https://img.shields.io/badge/cloud--native-enabled-blue.svg)
[![Discord Chat](https://img.shields.io/discord/628796009539043348.svg)](https://discord.gg/6pSWCTZ)
## Using it as a service
```console
go get github.com/dosco/super-graph
super-graph new <app_name>
```
## What is Super Graph
## Using it in your own code
Is designed to 100x your developer productivity. Super Graph will instantly and without you writing code provide you a high performance and secure GraphQL API for Postgres DB. GraphQL queries are translated into a single fast SQL query. No more writing API code as you develop
your web frontend just make the query you need and Super Graph will do the rest.
```console
go get github.com/dosco/super-graph/core
```
Super Graph has a rich feature set like integrating with your existing Ruby on Rails apps, joining your DB with data from remote APIs, role and attribute based access control, support for JWT tokens, built-in DB mutations and seeding, and a lot more.
```golang
package main
![GraphQL](docs/.vuepress/public/graphql.png?raw=true "")
import (
"database/sql"
"fmt"
"time"
"github.com/dosco/super-graph/core"
_ "github.com/jackc/pgx/v4/stdlib"
)
func main() {
db, err := sql.Open("pgx", "postgres://postgrs:@localhost:5432/example_db")
if err != nil {
log.Fatal(err)
}
sg, err := core.NewSuperGraph(nil, db)
if err != nil {
log.Fatal(err)
}
query := `
query {
posts {
id
title
}
}`
ctx = context.WithValue(ctx, core.UserIDKey, 1)
res, err := sg.GraphQL(ctx, query, nil)
if err != nil {
log.Fatal(err)
}
fmt.Println(string(res.Data))
}
```
## About Super Graph
After working on several products through my career I found that we spend way too much time on building API backends. Most APIs also need constant updating, and this costs time and money.
## The story of Super Graph?
After working on several products through my career I find that we spend way too much time on building API backends. Most APIs also require constant updating, this costs real time and money.
It's always the same thing, figure out what the UI needs then build an endpoint for it. Most API code involves struggling with an ORM to query a database and mangle the data into a shape that the UI expects to see.
I didn't want to write this code anymore, I wanted the computer to do it. Enter GraphQL, to me it sounded great, but it still required me to write all the same database query code.
Having worked with compilers before I saw this as a compiler problem. Why not build a compiler that converts GraphQL to highly efficient SQL.
This compiler is what sits at the heart of Super Graph, with layers of useful functionality around it like authentication, remote joins, rails integration, database migrations, and everything else needed for you to build production-ready apps with it.
This compiler is what sits at the heart of Super Graph with layers of useful functionality around it like authentication, remote joins, rails integration, database migrations and everything else needed for you to build production ready apps with it.
## Features
- Complex nested queries and mutations
- Auto learns database tables and relationships
- Role and Attribute-based access control
- Opaque cursor-based efficient pagination
- Full-text search and aggregations
- Role and Attribute based access control
- Full text search and aggregations
- JWT tokens supported (Auth0, etc)
- Join database queries with remote REST APIs
- Also works with existing Ruby-On-Rails apps
- Rails authentication supported (Redis, Memcache, Cookie)
- A simple config file
- High performance Go codebase
- High performance GO codebase
- Tiny docker image and low memory requirements
- Fuzz tested for security
- Database migrations tool
- Database seeding tool
- Works with Postgres and Yugabyte DB
- OpenCensus Support: Zipkin, Prometheus, X-Ray, Stackdriver
## Get started
```
git clone https://github.com/dosco/super-graph
cd ./super-graph
make install
super-graph new <app_name>
```
## Documentation
[supergraph.dev](https://supergraph.dev)
## Reach out
## Contact me
We're happy to help you leverage Super Graph reach out if you have questions
I'm happy to help you deploy Super Graph so feel free to reach out over
Twitter or Discord.
[twitter/dosco](https://twitter.com/dosco)
@ -112,3 +77,5 @@ We're happy to help you leverage Super Graph reach out if you have questions
[Apache Public License 2.0](https://opensource.org/licenses/Apache-2.0)
Copyright (c) 2019-present Vikram Rangnekar

View File

@ -2,13 +2,13 @@ app_name: "Super Graph Development"
host_port: 0.0.0.0:8080
web_ui: true
# debug, error, warn, info, none
# debug, info, warn, error, fatal, panic
log_level: "debug"
# enable or disable http compression (uses gzip)
http_compress: true
# When production mode is 'true' only queries
# When production mode is 'true' only queries
# from the allow list are permitted.
# When it's 'false' all queries are saved to the
# the allow list in ./config/allow.list
@ -30,27 +30,7 @@ reload_on_config_change: true
# seed_file: seed.js
# Path pointing to where the migrations can be found
migrations_path: ./migrations
# Secret key for general encryption operations like
# encrypting the cursor data
secret_key: supercalifajalistics
# CORS: A list of origins a cross-domain request can be executed from.
# If the special * value is present in the list, all origins will be allowed.
# An origin may contain a wildcard (*) to replace 0 or more
# characters (i.e.: http://*.domain.com).
cors_allowed_origins: ["*"]
# Debug Cross Origin Resource Sharing requests
cors_debug: true
# Default API path prefix is /api you can change it if you like
# api_path: "/data"
# Cache-Control header can help cache queries if your CDN supports cache-control
# on POST requests (does not work with not mutations)
# cache_control: "public, max-age=300, s-maxage=600"
migrations_path: ./config/migrations
# Postgres related environment Variables
# SG_DATABASE_HOST
@ -68,25 +48,13 @@ cors_debug: true
# person: people
# sheep: sheep
# open opencensus tracing and metrics
# telemetry:
# debug: true
# metrics:
# exporter: "prometheus"
# tracing:
# exporter: "zipkin"
# endpoint: "http://zipkin:9411/api/v2/spans"
# sample: 0.2
# include_query: false
# include_params: false
auth:
# Can be 'rails' or 'jwt'
type: rails
cookie: _app_session
# Comment this out if you want to disable setting
# the user_id via a header for testing.
# the user_id via a header for testing.
# Disable in production
creds_in_header: true
@ -103,6 +71,7 @@ auth:
# password: ""
# max_idle: 80
# max_active: 12000
# In most cases you don't need these
# salt: "encrypted cookie"
# sign_salt: "signed encrypted cookie"
@ -120,7 +89,7 @@ database:
port: 5432
dbname: app_development
user: postgres
password: postgres
password: ''
#schema: "public"
#pool_size: 10
@ -134,18 +103,18 @@ database:
# database ping timeout is used for db health checking
ping_timeout: 1m
# Define additional variables here to be used with filters
variables:
admin_account_id: "5"
# Define additional variables here to be used with filters
variables:
admin_account_id: "5"
# Field and table names that you wish to block
blocklist:
- ar_internal_metadata
- schema_migrations
- secret
- password
- encrypted
- token
# Field and table names that you wish to block
blocklist:
- ar_internal_metadata
- schema_migrations
- secret
- password
- encrypted
- token
tables:
- name: customers
@ -155,7 +124,7 @@ tables:
url: http://rails_app:3000/stripe/$id
path: data
# debug: true
pass_headers:
pass_headers:
- cookie
set_headers:
- name: Host
@ -176,6 +145,7 @@ tables:
- name: email
related_to: products.name
roles_query: "SELECT * FROM users WHERE id = $user_id"
roles:
@ -184,12 +154,12 @@ roles:
- name: products
query:
limit: 10
columns: ["id", "name", "description"]
columns: ["id", "name", "description" ]
aggregation: false
insert:
block: false
update:
block: false
@ -197,13 +167,10 @@ roles:
block: false
- name: deals
query:
limit: 3
aggregation: false
- name: purchases
query:
limit: 3
columns: ["name", "description" ]
aggregation: false
- name: user
@ -216,10 +183,12 @@ roles:
query:
limit: 50
filters: ["{ user_id: { eq: $user_id } }"]
columns: ["id", "name", "description", "search_rank", "search_headline_description" ]
disable_functions: false
insert:
filters: ["{ user_id: { eq: $user_id } }"]
columns: ["id", "name", "description" ]
presets:
- user_id: "$user_id"
- created_at: "now"

View File

@ -6,13 +6,13 @@ app_name: "Super Graph Production"
host_port: 0.0.0.0:8080
web_ui: false
# debug, error, warn, info, none
# debug, info, warn, error, fatal, panic, disable
log_level: "info"
# enable or disable http compression (uses gzip)
http_compress: true
# When production mode is 'true' only queries
# When production mode is 'true' only queries
# from the allow list are permitted.
# When it's 'false' all queries are saved to the
# the allow list in ./config/allow.list
@ -30,11 +30,7 @@ enable_tracing: true
# seed_file: seed.js
# Path pointing to where the migrations can be found
# migrations_path: ./migrations
# Secret key for general encryption operations like
# encrypting the cursor data
# secret_key: supercalifajalistics
# migrations_path: migrations
# Postgres related environment Variables
# SG_DATABASE_HOST
@ -54,23 +50,14 @@ database:
port: 5432
dbname: app_production
user: postgres
password: postgres
password: ''
#pool_size: 10
#max_retries: 0
#log_level: "debug"
#log_level: "debug"
# Set session variable "user.id" to the user id
# Enable this if you need the user id in triggers, etc
set_user_id: false
# database ping timeout is used for db health checking
ping_timeout: 5m
# open opencensus tracing and metrics
# telemetry:
# debug: false
# metrics:
# exporter: "prometheus"
# tracing:
# exporter: "zipkin"
# endpoint: "http://zipkin:9411/api/v2/spans"
# sample: 0.6
ping_timeout: 5m

View File

@ -11,7 +11,7 @@ for (i = 0; i < user_count; i++) {
var pwd = fake.password()
var data = {
full_name: fake.name(),
avatar: fake.avatar_url(200),
avatar: fake.image_url(),
phone: fake.phone(),
email: fake.email(),
password: pwd,

View File

@ -1,228 +0,0 @@
// Package core provides the primary API to include and use Super Graph with your own code.
// For detailed documentation visit https://supergraph.dev
//
// Example usage:
/*
package main
import (
"database/sql"
"fmt"
"time"
"github.com/dosco/super-graph/core"
_ "github.com/jackc/pgx/v4/stdlib"
)
func main() {
db, err := sql.Open("pgx", "postgres://postgrs:@localhost:5432/example_db")
if err != nil {
log.Fatal(err)
}
sg, err := core.NewSuperGraph(nil, db)
if err != nil {
log.Fatal(err)
}
query := `
query {
posts {
id
title
}
}`
ctx = context.WithValue(ctx, core.UserIDKey, 1)
res, err := sg.GraphQL(ctx, query, nil)
if err != nil {
log.Fatal(err)
}
fmt.Println(string(res.Data))
}
*/
package core
import (
"context"
"crypto/sha256"
"database/sql"
"encoding/json"
"hash/maphash"
_log "log"
"os"
"github.com/chirino/graphql"
"github.com/dosco/super-graph/core/internal/allow"
"github.com/dosco/super-graph/core/internal/crypto"
"github.com/dosco/super-graph/core/internal/psql"
"github.com/dosco/super-graph/core/internal/qcode"
)
type contextkey int
// Constants to set values on the context passed to the NewSuperGraph function
const (
// Name of the authentication provider. Eg. google, github, etc
UserIDProviderKey contextkey = iota
// User ID value for authenticated users
UserIDKey
// User role if pre-defined
UserRoleKey
)
// SuperGraph struct is an instance of the Super Graph engine it holds all the required information like
// datase schemas, relationships, etc that the GraphQL to SQL compiler would need to do it's job.
type SuperGraph struct {
conf *Config
db *sql.DB
log *_log.Logger
dbinfo *psql.DBInfo
schema *psql.DBSchema
allowList *allow.List
encKey [32]byte
hashSeed maphash.Seed
queries map[uint64]*query
roles map[string]*Role
getRole *sql.Stmt
rmap map[uint64]resolvFn
abacEnabled bool
qc *qcode.Compiler
pc *psql.Compiler
ge *graphql.Engine
}
// NewSuperGraph creates the SuperGraph struct, this involves querying the database to learn its
// schemas and relationships
func NewSuperGraph(conf *Config, db *sql.DB) (*SuperGraph, error) {
return newSuperGraph(conf, db, nil)
}
// newSuperGraph helps with writing tests and benchmarks
func newSuperGraph(conf *Config, db *sql.DB, dbinfo *psql.DBInfo) (*SuperGraph, error) {
if conf == nil {
conf = &Config{}
}
sg := &SuperGraph{
conf: conf,
db: db,
dbinfo: dbinfo,
log: _log.New(os.Stdout, "", 0),
hashSeed: maphash.MakeSeed(),
}
if err := sg.initConfig(); err != nil {
return nil, err
}
if err := sg.initCompilers(); err != nil {
return nil, err
}
if err := sg.initAllowList(); err != nil {
return nil, err
}
if err := sg.initPrepared(); err != nil {
return nil, err
}
if err := sg.initResolvers(); err != nil {
return nil, err
}
if err := sg.initGraphQLEgine(); err != nil {
return nil, err
}
if conf.SecretKey != "" {
sk := sha256.Sum256([]byte(conf.SecretKey))
conf.SecretKey = ""
sg.encKey = sk
} else {
sg.encKey = crypto.NewEncryptionKey()
}
return sg, nil
}
// Result struct contains the output of the GraphQL function this includes resulting json from the
// database query and any error information
type Result struct {
op qcode.QType
name string
sql string
role string
Error string `json:"message,omitempty"`
Data json.RawMessage `json:"data,omitempty"`
Extensions *extensions `json:"extensions,omitempty"`
}
// GraphQL function is called on the SuperGraph struct to convert the provided GraphQL query into an
// SQL query and execute it on the database. In production mode prepared statements are directly used
// and no query compiling takes places.
//
// In developer mode all names queries are saved into a file `allow.list` and in production mode only
// queries from this file can be run.
func (sg *SuperGraph) GraphQL(c context.Context, query string, vars json.RawMessage) (*Result, error) {
var res Result
res.op = qcode.GetQType(query)
res.name = allow.QueryName(query)
// use the chirino/graphql library for introspection queries
// disabled when allow list is enforced
if !sg.conf.UseAllowList && res.name == "IntrospectionQuery" {
r := sg.ge.ServeGraphQL(&graphql.Request{Query: query})
res.Data = r.Data
if r.Error() != nil {
res.Error = r.Error().Error()
}
return &res, r.Error()
}
ct := scontext{Context: c, sg: sg, query: query, vars: vars, res: res}
if len(vars) <= 2 {
ct.vars = nil
}
if keyExists(c, UserIDKey) {
ct.role = "user"
} else {
ct.role = "anon"
}
data, err := ct.execQuery()
if err != nil {
return &ct.res, err
}
ct.res.Data = json.RawMessage(data)
return &ct.res, nil
}
// GraphQLSchema function return the GraphQL schema for the underlying database connected
// to this instance of Super Graph
func (sg *SuperGraph) GraphQLSchema() (string, error) {
return sg.ge.Schema.String(), nil
}
// Operation function return the operation type from the query. It uses a very fast algorithm to
// extract the operation without having to parse the query.
func Operation(query string) OpType {
return OpType(qcode.GetQType(query))
}
// Name function return the operation name from the query. It uses a very fast algorithm to
// extract the operation name without having to parse the query.
func Name(query string) string {
return allow.QueryName(query)
}

View File

@ -1,62 +0,0 @@
package core
import (
"context"
"fmt"
"testing"
"github.com/DATA-DOG/go-sqlmock"
"github.com/dosco/super-graph/core/internal/psql"
)
func BenchmarkGraphQL(b *testing.B) {
ct := context.WithValue(context.Background(), UserIDKey, "1")
db, _, err := sqlmock.New()
if err != nil {
b.Fatal(err)
}
defer db.Close()
// mock.ExpectQuery(`^SELECT jsonb_build_object`).WithArgs()
c := &Config{}
sg, err := newSuperGraph(c, db, psql.GetTestDBInfo())
if err != nil {
b.Fatal(err)
}
query := `
query {
products {
id
name
user {
full_name
phone
email
}
customers {
id
email
}
}
users {
id
name
}
}`
b.ResetTimer()
b.ReportAllocs()
b.RunParallel(func(pb *testing.PB) {
for pb.Next() {
_, err = sg.GraphQL(ct, query, nil)
}
})
fmt.Println(err)
//fmt.Println(mock.ExpectationsWereMet())
}

View File

@ -1,96 +0,0 @@
package core
import (
"encoding/json"
"fmt"
"github.com/dosco/super-graph/core/internal/psql"
"github.com/dosco/super-graph/jsn"
)
// argList function is used to create a list of arguments to pass
// to a prepared statement.
func (c *scontext) argList(md psql.Metadata) ([]interface{}, error) {
params := md.Params()
vars := make([]interface{}, len(params))
var fields map[string]json.RawMessage
var err error
if len(c.vars) != 0 {
fields, _, err = jsn.Tree(c.vars)
if err != nil {
return nil, err
}
}
for i, p := range params {
switch p.Name {
case "user_id":
if v := c.Value(UserIDKey); v != nil {
vars[i] = v.(string)
} else {
return nil, argErr(p)
}
case "user_id_provider":
if v := c.Value(UserIDProviderKey); v != nil {
vars[i] = v.(string)
} else {
return nil, argErr(p)
}
case "user_role":
if v := c.Value(UserRoleKey); v != nil {
vars[i] = v.(string)
} else {
return nil, argErr(p)
}
case "cursor":
if v, ok := fields["cursor"]; ok && v[0] == '"' {
v1, err := c.sg.decrypt(string(v[1 : len(v)-1]))
if err != nil {
return nil, err
}
vars[i] = v1
} else {
return nil, argErr(p)
}
default:
if v, ok := fields[p.Name]; ok {
switch {
case p.IsArray && v[0] != '[':
return nil, fmt.Errorf("variable '%s' should be an array of type '%s'", p.Name, p.Type)
case p.Type == "json" && v[0] != '[' && v[0] != '{':
return nil, fmt.Errorf("variable '%s' should be an array or object", p.Name)
}
switch v[0] {
case '[', '{':
vars[i] = v
default:
var val interface{}
if err := json.Unmarshal(v, &val); err != nil {
return nil, err
}
vars[i] = val
}
} else {
return nil, argErr(p)
}
}
}
return vars, nil
}
func argErr(p psql.Param) error {
return fmt.Errorf("required variable '%s' of type '%s' must be set", p.Name, p.Type)
}

View File

@ -1,256 +0,0 @@
package core
import (
"fmt"
"path"
"path/filepath"
"strings"
"github.com/spf13/viper"
)
// Core struct contains core specific config value
type Config struct {
// SecretKey is used to encrypt opaque values such as
// the cursor. Auto-generated if not set
SecretKey string `mapstructure:"secret_key"`
// UseAllowList (aka production mode) when set to true ensures
// only queries lists in the allow.list file can be used. All
// queries are pre-prepared so no compiling happens and things are
// very fast.
UseAllowList bool `mapstructure:"use_allow_list"`
// AllowListFile if the path to allow list file if not set the
// path is assumed to tbe the same as the config path (allow.list)
AllowListFile string `mapstructure:"allow_list_file"`
// SetUserID forces the database session variable `user.id` to
// be set to the user id. This variables can be used by triggers
// or other database functions
SetUserID bool `mapstructure:"set_user_id"`
// DefaultBlock ensures that in anonymous mode (role 'anon') all tables
// are blocked from queries and mutations. To open access to tables in
// anonymous mode they have to be added to the 'anon' role config.
DefaultBlock bool `mapstructure:"default_block"`
// Vars is a map of hardcoded variables that can be leveraged in your
// queries (eg variable admin_id will be $admin_id in the query)
Vars map[string]string `mapstructure:"variables"`
// Blocklist is a list of tables and columns that should be filtered
// out from any and all queries
Blocklist []string
// Tables contains all table specific configuration such as aliased tables
// creating relationships between tables, etc
Tables []Table
// RolesQuery if set enabled attributed based access control. This query
// is use to fetch the user attributes that then dynamically define the users
// role.
RolesQuery string `mapstructure:"roles_query"`
// Roles contains all the configuration for all the roles you want to support
// `user` and `anon` are two default roles. User role is for when a user ID is
// available and Anon when it's not.
//
// If you're using the RolesQuery config to enable atribute based acess control then
// you can add more custom roles.
Roles []Role
// Inflections is to add additionally singular to plural mappings
// to the engine (eg. sheep: sheep)
Inflections map[string]string `mapstructure:"inflections"`
// Database schema name. Defaults to 'public'
DBSchema string `mapstructure:"db_schema"`
}
// Table struct defines a database table
type Table struct {
Name string
Table string
Type string
Blocklist []string
Remotes []Remote
Columns []Column
}
// Column struct defines a database column
type Column struct {
Name string
Type string
ForeignKey string `mapstructure:"related_to"`
}
// Remote struct defines a remote API endpoint
type Remote struct {
Name string
ID string
Path string
URL string
Debug bool
PassHeaders []string `mapstructure:"pass_headers"`
SetHeaders []struct {
Name string
Value string
} `mapstructure:"set_headers"`
}
// Role struct contains role specific access control values for for all database tables
type Role struct {
Name string
Match string
Tables []RoleTable
tm map[string]*RoleTable
}
// RoleTable struct contains role specific access control values for a database table
type RoleTable struct {
Name string
ReadOnly bool `mapstructure:"read_only"`
Query *Query
Insert *Insert
Update *Update
Delete *Delete
}
// Query struct contains access control values for query operations
type Query struct {
Limit int
Filters []string
Columns []string
DisableFunctions bool `mapstructure:"disable_functions"`
Block bool
}
// Insert struct contains access control values for insert operations
type Insert struct {
Filters []string
Columns []string
Presets map[string]string
Block bool
}
// Insert struct contains access control values for update operations
type Update struct {
Filters []string
Columns []string
Presets map[string]string
Block bool
}
// Delete struct contains access control values for delete operations
type Delete struct {
Filters []string
Columns []string
Block bool
}
// AddRoleTable function is a helper function to make it easy to add per-table
// row-level config
func (c *Config) AddRoleTable(role, table string, conf interface{}) error {
var r *Role
for i := range c.Roles {
if strings.EqualFold(c.Roles[i].Name, role) {
r = &c.Roles[i]
break
}
}
if r == nil {
nr := Role{Name: role}
c.Roles = append(c.Roles, nr)
r = &nr
}
var t *RoleTable
for i := range r.Tables {
if strings.EqualFold(r.Tables[i].Name, table) {
t = &r.Tables[i]
break
}
}
if t == nil {
nt := RoleTable{Name: table}
r.Tables = append(r.Tables, nt)
t = &nt
}
switch v := conf.(type) {
case Query:
t.Query = &v
case Insert:
t.Insert = &v
case Update:
t.Update = &v
case Delete:
t.Delete = &v
default:
return fmt.Errorf("unsupported object type: %t", v)
}
return nil
}
// ReadInConfig function reads in the config file for the environment specified in the GO_ENV
// environment variable. This is the best way to create a new Super Graph config.
func ReadInConfig(configFile string) (*Config, error) {
cp := path.Dir(configFile)
vi := newViper(cp, path.Base(configFile))
if err := vi.ReadInConfig(); err != nil {
return nil, err
}
if pcf := vi.GetString("inherits"); pcf != "" {
cf := vi.ConfigFileUsed()
vi = newViper(cp, pcf)
if err := vi.ReadInConfig(); err != nil {
return nil, err
}
if v := vi.GetString("inherits"); v != "" {
return nil, fmt.Errorf("inherited config (%s) cannot itself inherit (%s)", pcf, v)
}
vi.SetConfigFile(cf)
if err := vi.MergeInConfig(); err != nil {
return nil, err
}
}
c := &Config{}
if err := vi.Unmarshal(&c); err != nil {
return nil, fmt.Errorf("failed to decode config, %v", err)
}
if c.AllowListFile == "" {
c.AllowListFile = path.Join(cp, "allow.list")
}
return c, nil
}
func newViper(configPath, configFile string) *viper.Viper {
vi := viper.New()
vi.SetEnvPrefix("SG")
vi.SetEnvKeyReplacer(strings.NewReplacer(".", "_"))
vi.AutomaticEnv()
if filepath.Ext(configFile) != "" {
vi.SetConfigFile(path.Join(configPath, configFile))
} else {
vi.SetConfigName(configFile)
vi.AddConfigPath(configPath)
vi.AddConfigPath("./config")
}
return vi
}

View File

@ -1,14 +0,0 @@
package core
import (
"context"
"errors"
)
var (
errNotFound = errors.New("not found in prepared statements")
)
func keyExists(ct context.Context, key contextkey) bool {
return ct.Value(key) != nil
}

View File

@ -1,446 +0,0 @@
package core
import (
"context"
"database/sql"
"encoding/json"
"fmt"
"hash/maphash"
"time"
"github.com/dosco/super-graph/core/internal/psql"
"github.com/dosco/super-graph/core/internal/qcode"
)
type OpType int
const (
OpQuery OpType = iota
OpMutation
)
type extensions struct {
Tracing *trace `json:"tracing,omitempty"`
}
type trace struct {
Version int `json:"version"`
StartTime time.Time `json:"startTime"`
EndTime time.Time `json:"endTime"`
Duration time.Duration `json:"duration"`
Execution execution `json:"execution"`
}
type execution struct {
Resolvers []resolver `json:"resolvers"`
}
type resolver struct {
Path []string `json:"path"`
ParentType string `json:"parentType"`
FieldName string `json:"fieldName"`
ReturnType string `json:"returnType"`
StartOffset int `json:"startOffset"`
Duration time.Duration `json:"duration"`
}
type scontext struct {
context.Context
sg *SuperGraph
query string
vars json.RawMessage
role string
res Result
}
func (sg *SuperGraph) initCompilers() error {
var err error
var schema string
if sg.conf.DBSchema == "" {
schema = "public"
} else {
schema = sg.conf.DBSchema
}
// If sg.di is not null then it's probably set
// for tests
if sg.dbinfo == nil {
sg.dbinfo, err = psql.GetDBInfo(sg.db, schema)
if err != nil {
return err
}
}
if len(sg.dbinfo.Tables) == 0 {
return fmt.Errorf("no tables found in database (schema: %s)", schema)
}
if err = addTables(sg.conf, sg.dbinfo); err != nil {
return err
}
if err = addForeignKeys(sg.conf, sg.dbinfo); err != nil {
return err
}
sg.schema, err = psql.NewDBSchema(sg.dbinfo, getDBTableAliases(sg.conf))
if err != nil {
return err
}
sg.qc, err = qcode.NewCompiler(qcode.Config{
DefaultBlock: sg.conf.DefaultBlock,
Blocklist: sg.conf.Blocklist,
})
if err != nil {
return err
}
if err := addRoles(sg.conf, sg.qc); err != nil {
return err
}
sg.pc = psql.NewCompiler(psql.Config{
Schema: sg.schema,
Vars: sg.conf.Vars,
})
return nil
}
func (c *scontext) execQuery() ([]byte, error) {
var data []byte
var st *stmt
var err error
if c.sg.conf.UseAllowList {
data, st, err = c.resolvePreparedSQL()
} else {
data, st, err = c.resolveSQL()
}
if err != nil {
return nil, err
}
if len(data) == 0 || st.md.Skipped() == 0 {
return data, nil
}
// return c.sg.execRemoteJoin(st, data, c.req.hdr)
return c.sg.execRemoteJoin(st, data, nil)
}
func (c *scontext) resolvePreparedSQL() ([]byte, *stmt, error) {
var tx *sql.Tx
var err error
mutation := (c.res.op == qcode.QTMutation)
useRoleQuery := c.sg.abacEnabled && mutation
useTx := useRoleQuery || c.sg.conf.SetUserID
if useTx {
if tx, err = c.sg.db.BeginTx(c, nil); err != nil {
return nil, nil, err
}
defer tx.Rollback() //nolint: errcheck
}
if c.sg.conf.SetUserID {
if err := setLocalUserID(c, tx); err != nil {
return nil, nil, err
}
}
var role string
if useRoleQuery {
if role, err = c.executeRoleQuery(tx); err != nil {
return nil, nil, err
}
} else if v := c.Value(UserRoleKey); v != nil {
role = v.(string)
} else {
role = c.role
}
c.res.role = role
h := maphash.Hash{}
h.SetSeed(c.sg.hashSeed)
id := queryID(&h, c.res.name, role)
q, ok := c.sg.queries[id]
if !ok {
return nil, nil, errNotFound
}
if q.sd == nil {
q.Do(func() { c.sg.prepare(q, role) })
if q.err != nil {
return nil, nil, err
}
}
c.res.sql = q.st.sql
var root []byte
var row *sql.Row
varsList, err := c.argList(q.st.md)
if err != nil {
return nil, nil, err
}
if useTx {
row = tx.Stmt(q.sd).QueryRow(varsList...)
} else {
row = q.sd.QueryRow(varsList...)
}
if q.roleArg {
err = row.Scan(&role, &root)
} else {
err = row.Scan(&root)
}
if err != nil {
return nil, nil, err
}
c.role = role
if useTx {
if err := tx.Commit(); err != nil {
return nil, nil, q.err
}
}
if root, err = c.sg.encryptCursor(q.st.qc, root); err != nil {
return nil, nil, err
}
return root, &q.st, nil
}
func (c *scontext) resolveSQL() ([]byte, *stmt, error) {
var tx *sql.Tx
var err error
mutation := (c.res.op == qcode.QTMutation)
useRoleQuery := c.sg.abacEnabled && mutation
useTx := useRoleQuery || c.sg.conf.SetUserID
if useTx {
if tx, err = c.sg.db.BeginTx(c, nil); err != nil {
return nil, nil, err
}
defer tx.Rollback() //nolint: errcheck
}
if c.sg.conf.SetUserID {
if err := setLocalUserID(c, tx); err != nil {
return nil, nil, err
}
}
if useRoleQuery {
if c.role, err = c.executeRoleQuery(tx); err != nil {
return nil, nil, err
}
} else if v := c.Value(UserRoleKey); v != nil {
c.role = v.(string)
}
stmts, err := c.sg.buildStmt(c.res.op, []byte(c.query), c.vars, c.role)
if err != nil {
return nil, nil, err
}
st := &stmts[0]
c.res.sql = st.sql
varList, err := c.argList(st.md)
if err != nil {
return nil, nil, err
}
// finalSQL := buf.String()
////
// _, err = t.ExecuteFunc(buf, c.argMap(st.md))
// if err != nil {
// return nil, nil, err
// }
// finalSQL := buf.String()
/////
// var stime time.Time
// if c.sg.conf.EnableTracing {
// stime = time.Now()
// }
var root []byte
var role string
var row *sql.Row
// defaultRole := c.role
if useTx {
row = tx.QueryRowContext(c, st.sql, varList...)
} else {
row = c.sg.db.QueryRowContext(c, st.sql, varList...)
}
if len(stmts) > 1 {
err = row.Scan(&role, &root)
} else {
err = row.Scan(&root)
}
if role == "" {
c.res.role = c.role
} else {
c.res.role = role
}
if err != nil {
return nil, nil, err
}
if useTx {
if err := tx.Commit(); err != nil {
return nil, nil, err
}
}
if root, err = c.sg.encryptCursor(st.qc, root); err != nil {
return nil, nil, err
}
if c.sg.allowList.IsPersist() {
if err := c.sg.allowList.Set(c.vars, c.query, ""); err != nil {
return nil, nil, err
}
}
if len(stmts) > 1 {
if st = findStmt(role, stmts); st == nil {
return nil, nil, fmt.Errorf("invalid role '%s' returned", role)
}
}
// if c.sg.conf.EnableTracing {
// for _, id := range st.qc.Roots {
// c.addTrace(st.qc.Selects, id, stime)
// }
// }
return root, st, nil
}
func (c *scontext) executeRoleQuery(tx *sql.Tx) (string, error) {
userID := c.Value(UserIDKey)
if userID == nil {
return "anon", nil
}
var role string
row := c.sg.getRole.QueryRow(userID, c.role)
if err := row.Scan(&role); err != nil {
return "", err
}
return role, nil
}
func (r *Result) Operation() OpType {
switch r.op {
case qcode.QTQuery:
return OpQuery
case qcode.QTMutation, qcode.QTInsert, qcode.QTUpdate, qcode.QTUpsert, qcode.QTDelete:
return OpMutation
default:
return -1
}
}
func (r *Result) OperationName() string {
return r.op.String()
}
func (r *Result) QueryName() string {
return r.name
}
func (r *Result) Role() string {
return r.role
}
func (r *Result) SQL() string {
return r.sql
}
// func (c *scontext) addTrace(sel []qcode.Select, id int32, st time.Time) {
// et := time.Now()
// du := et.Sub(st)
// if c.res.Extensions == nil {
// c.res.Extensions = &extensions{&trace{
// Version: 1,
// StartTime: st,
// Execution: execution{},
// }}
// }
// c.res.Extensions.Tracing.EndTime = et
// c.res.Extensions.Tracing.Duration = du
// n := 1
// for i := id; i != -1; i = sel[i].ParentID {
// n++
// }
// path := make([]string, n)
// n--
// for i := id; ; i = sel[i].ParentID {
// path[n] = sel[i].Name
// if sel[i].ParentID == -1 {
// break
// }
// n--
// }
// tr := resolver{
// Path: path,
// ParentType: "Query",
// FieldName: sel[id].Name,
// ReturnType: "object",
// StartOffset: 1,
// Duration: du,
// }
// c.res.Extensions.Tracing.Execution.Resolvers =
// append(c.res.Extensions.Tracing.Execution.Resolvers, tr)
// }
func findStmt(role string, stmts []stmt) *stmt {
for i := range stmts {
if stmts[i].role.Name != role {
continue
}
return &stmts[i]
}
return nil
}

View File

@ -1,72 +0,0 @@
package core
import (
"bytes"
"encoding/base64"
"github.com/dosco/super-graph/core/internal/crypto"
"github.com/dosco/super-graph/core/internal/qcode"
"github.com/dosco/super-graph/jsn"
)
func (sg *SuperGraph) encryptCursor(qc *qcode.QCode, data []byte) ([]byte, error) {
var keys [][]byte
for _, s := range qc.Selects {
if s.Paging.Type != qcode.PtOffset {
var buf bytes.Buffer
buf.WriteString(s.FieldName)
buf.WriteString("_cursor")
keys = append(keys, buf.Bytes())
}
}
if len(keys) == 0 {
return data, nil
}
from := jsn.Get(data, keys)
to := make([]jsn.Field, len(from))
for i, f := range from {
to[i].Key = f.Key
if f.Value[0] != '"' || f.Value[len(f.Value)-1] != '"' {
continue
}
var buf bytes.Buffer
if len(f.Value) > 2 {
v, err := crypto.Encrypt(f.Value[1:len(f.Value)-1], &sg.encKey)
if err != nil {
return nil, err
}
buf.WriteByte('"')
buf.WriteString(base64.StdEncoding.EncodeToString(v))
buf.WriteByte('"')
} else {
buf.WriteString(`null`)
}
to[i].Value = buf.Bytes()
}
var buf bytes.Buffer
if err := jsn.Replace(&buf, data, from, to); err != nil {
return nil, err
}
return buf.Bytes(), nil
}
func (sg *SuperGraph) decrypt(data string) ([]byte, error) {
v, err := base64.StdEncoding.DecodeString(data)
if err != nil {
return nil, err
}
return crypto.Decrypt(v, &sg.encKey)
}

View File

@ -1,15 +0,0 @@
package core
import (
"context"
"database/sql"
)
func setLocalUserID(c context.Context, tx *sql.Tx) error {
var err error
if v := c.Value(UserIDKey); v != nil {
_, err = tx.Exec(`SET LOCAL "user.id" = ?`, v)
}
return err
}

View File

@ -1,341 +0,0 @@
package core
import (
"fmt"
"strings"
"github.com/dosco/super-graph/core/internal/psql"
"github.com/dosco/super-graph/core/internal/qcode"
"github.com/gobuffalo/flect"
)
func (sg *SuperGraph) initConfig() error {
c := sg.conf
for k, v := range c.Inflections {
flect.AddPlural(k, v)
}
// Tables: Validate and sanitize
tm := make(map[string]struct{})
for i := 0; i < len(c.Tables); i++ {
t := &c.Tables[i]
// t.Name = flect.Pluralize(strings.ToLower(t.Name))
if _, ok := tm[t.Name]; ok {
sg.conf.Tables = append(c.Tables[:i], c.Tables[i+1:]...)
sg.log.Printf("WRN duplicate table found: %s", t.Name)
}
tm[t.Name] = struct{}{}
t.Table = flect.Pluralize(strings.ToLower(t.Table))
}
sg.roles = make(map[string]*Role)
for i := 0; i < len(c.Roles); i++ {
role := &c.Roles[i]
role.Name = sanitize(role.Name)
if _, ok := sg.roles[role.Name]; ok {
c.Roles = append(c.Roles[:i], c.Roles[i+1:]...)
sg.log.Printf("WRN duplicate role found: %s", role.Name)
}
role.Match = sanitize(role.Match)
role.tm = make(map[string]*RoleTable)
for n, table := range role.Tables {
role.tm[table.Name] = &role.Tables[n]
}
sg.roles[role.Name] = role
}
// If user role not defined then create it
if _, ok := sg.roles["user"]; !ok {
ur := Role{
Name: "user",
tm: make(map[string]*RoleTable),
}
c.Roles = append(c.Roles, ur)
sg.roles["user"] = &ur
}
// If anon role is not defined then create it
if _, ok := sg.roles["anon"]; !ok {
ur := Role{
Name: "anon",
tm: make(map[string]*RoleTable),
}
c.Roles = append(c.Roles, ur)
sg.roles["anon"] = &ur
}
if c.RolesQuery == "" {
sg.log.Printf("INF roles_query not defined: attribute based access control disabled")
} else {
n := 0
for k, v := range sg.roles {
if k == "user" || k == "anon" {
n++
} else if v.Match != "" {
n++
}
}
sg.abacEnabled = (n > 2)
if !sg.abacEnabled {
sg.log.Printf("WRN attribute based access control disabled: no custom roles found (with 'match' defined)")
}
}
return nil
}
func getDBTableAliases(c *Config) map[string][]string {
m := make(map[string][]string, len(c.Tables))
for i := range c.Tables {
t := c.Tables[i]
if t.Table != "" && t.Type == "" {
m[t.Table] = append(m[t.Table], t.Name)
}
}
return m
}
func addTables(c *Config, di *psql.DBInfo) error {
var err error
for _, t := range c.Tables {
switch t.Type {
case "json", "jsonb":
err = addJsonTable(di, t.Columns, t)
case "polymorphic":
err = addVirtualTable(di, t.Columns, t)
}
if err != nil {
return err
}
}
return nil
}
func addJsonTable(di *psql.DBInfo, cols []Column, t Table) error {
// This is for jsonb columns that want to be tables.
bc, ok := di.GetColumn(t.Table, t.Name)
if !ok {
return fmt.Errorf(
"json table: column '%s' not found on table '%s'",
t.Name, t.Table)
}
if bc.Type != "json" && bc.Type != "jsonb" {
return fmt.Errorf(
"json table: column '%s' in table '%s' is of type '%s'. Only JSON or JSONB is valid",
t.Name, t.Table, bc.Type)
}
table := psql.DBTable{
Name: t.Name,
Key: strings.ToLower(t.Name),
Type: bc.Type,
}
columns := make([]psql.DBColumn, 0, len(cols))
for i := range cols {
c := cols[i]
columns = append(columns, psql.DBColumn{
Name: c.Name,
Key: strings.ToLower(c.Name),
Type: c.Type,
})
}
di.AddTable(table, columns)
bc.FKeyTable = t.Name
return nil
}
func addVirtualTable(di *psql.DBInfo, cols []Column, t Table) error {
if len(cols) == 0 {
return fmt.Errorf("polymorphic table: no id column specified")
}
c := cols[0]
if c.ForeignKey == "" {
return fmt.Errorf("polymorphic table: no 'related_to' specified on id column")
}
s := strings.SplitN(c.ForeignKey, ".", 2)
if len(s) != 2 {
return fmt.Errorf("polymorphic table: foreign key must be <type column>.<foreign key column>")
}
di.VTables = append(di.VTables, psql.VirtualTable{
Name: t.Name,
IDColumn: c.Name,
TypeColumn: s[0],
FKeyColumn: s[1],
})
return nil
}
func addForeignKeys(c *Config, di *psql.DBInfo) error {
for _, t := range c.Tables {
if t.Type != "" {
continue
}
for _, c := range t.Columns {
if c.ForeignKey == "" {
continue
}
if err := addForeignKey(di, c, t); err != nil {
return err
}
}
}
return nil
}
func addForeignKey(di *psql.DBInfo, c Column, t Table) error {
var tn string
if t.Type == "polymorphic" {
tn = t.Table
} else {
tn = t.Name
}
c1, ok := di.GetColumn(tn, c.Name)
if !ok {
return fmt.Errorf(
"config: invalid table '%s' or column '%s' defined",
tn, c.Name)
}
v := strings.SplitN(c.ForeignKey, ".", 2)
if len(v) != 2 {
return fmt.Errorf(
"config: invalid foreign_key defined for table '%s' and column '%s': %s",
tn, c.Name, c.ForeignKey)
}
// check if it's a polymorphic foreign key
if _, ok := di.GetColumn(tn, v[0]); ok {
c2, ok := di.GetColumn(tn, v[1])
if !ok {
return fmt.Errorf(
"config: invalid column '%s' for polymorphic relationship on table '%s' and column '%s'",
v[1], tn, c.Name)
}
c1.FKeyTable = v[0]
c1.FKeyColID = []int16{c2.ID}
return nil
}
fkt, fkc := v[0], v[1]
c3, ok := di.GetColumn(fkt, fkc)
if !ok {
return fmt.Errorf(
"config: foreign_key for table '%s' and column '%s' points to unknown table '%s' and column '%s'",
t.Name, c.Name, v[0], v[1])
}
c1.FKeyTable = fkt
c1.FKeyColID = []int16{c3.ID}
return nil
}
func addRoles(c *Config, qc *qcode.Compiler) error {
for _, r := range c.Roles {
for _, t := range r.Tables {
if err := addRole(qc, r, t, c.DefaultBlock); err != nil {
return err
}
}
}
return nil
}
func addRole(qc *qcode.Compiler, r Role, t RoleTable, defaultBlock bool) error {
ro := false // read-only
if defaultBlock && r.Name == "anon" {
ro = true
}
if t.ReadOnly {
ro = true
}
query := qcode.QueryConfig{Block: false}
insert := qcode.InsertConfig{Block: ro}
update := qcode.UpdateConfig{Block: ro}
del := qcode.DeleteConfig{Block: ro}
if t.Query != nil {
query = qcode.QueryConfig{
Limit: t.Query.Limit,
Filters: t.Query.Filters,
Columns: t.Query.Columns,
DisableFunctions: t.Query.DisableFunctions,
Block: t.Query.Block,
}
}
if t.Insert != nil {
insert = qcode.InsertConfig{
Filters: t.Insert.Filters,
Columns: t.Insert.Columns,
Presets: t.Insert.Presets,
Block: t.Insert.Block,
}
}
if t.Update != nil {
update = qcode.UpdateConfig{
Filters: t.Update.Filters,
Columns: t.Update.Columns,
Presets: t.Update.Presets,
Block: t.Update.Block,
}
}
if t.Delete != nil {
del = qcode.DeleteConfig{
Filters: t.Delete.Filters,
Columns: t.Delete.Columns,
Block: t.Delete.Block,
}
}
return qc.AddRole(r.Name, t.Name, qcode.TRConfig{
Query: query,
Insert: insert,
Update: update,
Delete: del,
})
}
func (r *Role) GetTable(name string) *RoleTable {
return r.tm[name]
}
func sanitize(value string) string {
return strings.ToLower(strings.TrimSpace(value))
}

View File

@ -1,349 +0,0 @@
package allow
import (
"bytes"
"encoding/json"
"errors"
"fmt"
"io/ioutil"
"log"
"os"
"sort"
"strings"
"text/scanner"
"github.com/chirino/graphql/schema"
"github.com/dosco/super-graph/jsn"
)
const (
expComment = iota + 1
expVar
expQuery
)
type Item struct {
Name string
key string
Query string
Vars string
Comment string
}
type List struct {
filepath string
saveChan chan Item
}
type Config struct {
CreateIfNotExists bool
Persist bool
Log *log.Logger
}
func New(filename string, conf Config) (*List, error) {
al := List{}
if filename != "" {
fp := filename
if _, err := os.Stat(fp); err == nil {
al.filepath = fp
} else if !os.IsNotExist(err) {
return nil, err
}
}
if al.filepath == "" {
fp := "./allow.list"
if _, err := os.Stat(fp); err == nil {
al.filepath = fp
} else if !os.IsNotExist(err) {
return nil, err
}
}
if al.filepath == "" {
fp := "./config/allow.list"
if _, err := os.Stat(fp); err == nil {
al.filepath = fp
} else if !os.IsNotExist(err) {
return nil, err
}
}
if al.filepath == "" {
if !conf.CreateIfNotExists {
return nil, errors.New("allow.list not found")
}
if filename == "" {
al.filepath = "./config/allow.list"
} else {
al.filepath = filename
}
if file, err := os.OpenFile(al.filepath, os.O_RDONLY|os.O_CREATE, 0644); err != nil {
return nil, err
} else {
file.Close()
}
}
var err error
if conf.Persist {
al.saveChan = make(chan Item)
go func() {
for v := range al.saveChan {
err := al.save(v)
if err != nil && conf.Log != nil {
conf.Log.Println("WRN allow list save:", err)
}
}
}()
}
if err != nil {
return nil, err
}
return &al, nil
}
func (al *List) IsPersist() bool {
return al.saveChan != nil
}
func (al *List) Set(vars []byte, query, comment string) error {
if al.saveChan == nil {
return errors.New("allow.list is read-only")
}
if query == "" {
return errors.New("empty query")
}
al.saveChan <- Item{
Comment: comment,
Query: query,
Vars: string(vars),
}
return nil
}
func (al *List) Load() ([]Item, error) {
b, err := ioutil.ReadFile(al.filepath)
if err != nil {
return nil, err
}
return parse(string(b), al.filepath)
}
func parse(b, filename string) ([]Item, error) {
var items []Item
var s scanner.Scanner
s.Init(strings.NewReader(b))
s.Filename = filename
s.Mode ^= scanner.SkipComments
var op, sp scanner.Position
var item Item
newComment := false
st := expComment
for tok := s.Scan(); tok != scanner.EOF; tok = s.Scan() {
txt := s.TokenText()
switch {
case strings.HasPrefix(txt, "/*"):
if st == expQuery {
v := b[sp.Offset:s.Pos().Offset]
item.Query = strings.TrimSpace(v[:strings.LastIndexByte(v, '}')+1])
items = append(items, item)
}
item = Item{Comment: strings.TrimSpace(txt[2 : len(txt)-2])}
sp = s.Pos()
st = expComment
newComment = true
case !newComment && strings.HasPrefix(txt, "#"):
if st == expQuery {
v := b[sp.Offset:s.Pos().Offset]
item.Query = strings.TrimSpace(v[:strings.LastIndexByte(v, '}')+1])
items = append(items, item)
}
item = Item{}
sp = s.Pos()
st = expComment
case strings.HasPrefix(txt, "variables"):
if st == expComment {
v := b[sp.Offset:s.Pos().Offset]
item.Comment = strings.TrimSpace(v[:strings.IndexByte(v, '\n')])
}
sp = s.Pos()
st = expVar
case isGraphQL(txt):
if st == expVar {
v := b[sp.Offset:s.Pos().Offset]
item.Vars = strings.TrimSpace(v[:strings.LastIndexByte(v, '}')+1])
}
sp = op
st = expQuery
}
op = s.Pos()
}
if st == expQuery {
v := b[sp.Offset:s.Pos().Offset]
item.Query = strings.TrimSpace(v[:strings.LastIndexByte(v, '}')+1])
items = append(items, item)
}
for i := range items {
items[i].Name = QueryName(items[i].Query)
items[i].key = strings.ToLower(items[i].Name)
}
return items, nil
}
func isGraphQL(s string) bool {
return strings.HasPrefix(s, "query") ||
strings.HasPrefix(s, "mutation") ||
strings.HasPrefix(s, "subscription")
}
func (al *List) save(item Item) error {
var buf bytes.Buffer
qd := &schema.QueryDocument{}
if err := qd.Parse(item.Query); err != nil {
return err
}
qd.WriteTo(&buf)
query := buf.String()
buf.Reset()
item.Name = QueryName(query)
item.key = strings.ToLower(item.Name)
if item.Name == "" {
return nil
}
list, err := al.Load()
if err != nil {
return err
}
index := -1
for i, v := range list {
if strings.EqualFold(v.Name, item.Name) {
index = i
break
}
}
if index != -1 {
if list[index].Comment != "" {
item.Comment = list[index].Comment
}
list[index] = item
} else {
list = append(list, item)
}
f, err := os.Create(al.filepath)
if err != nil {
return err
}
defer f.Close()
sort.Slice(list, func(i, j int) bool {
return strings.Compare(list[i].key, list[j].key) == -1
})
for i, v := range list {
var vars string
if v.Vars != "" {
buf.Reset()
if err := jsn.Clear(&buf, []byte(v.Vars)); err != nil {
continue
}
vj := json.RawMessage(buf.Bytes())
if vj, err = json.MarshalIndent(vj, "", " "); err != nil {
continue
}
vars = string(vj)
}
list[i].Vars = vars
list[i].Comment = strings.TrimSpace(v.Comment)
}
for _, v := range list {
if v.Comment != "" {
_, err = f.WriteString(fmt.Sprintf("/* %s */\n\n", v.Comment))
} else {
_, err = f.WriteString(fmt.Sprintf("/* %s */\n\n", v.Name))
}
if err != nil {
return err
}
if v.Vars != "" {
_, err = f.WriteString(fmt.Sprintf("variables %s\n\n", v.Vars))
if err != nil {
return err
}
}
_, err = f.WriteString(fmt.Sprintf("%s\n\n", v.Query))
if err != nil {
return err
}
}
return nil
}
func QueryName(b string) string {
state, s := 0, 0
for i := 0; i < len(b); i++ {
switch {
case state == 2 && !isValidNameChar(b[i]):
return b[s:i]
case state == 1 && b[i] == '{':
return ""
case state == 1 && isValidNameChar(b[i]):
s = i
state = 2
case i != 0 && b[i] == ' ' && (b[i-1] == 'n' || b[i-1] == 'y'):
state = 1
}
}
return ""
}
func isValidNameChar(c byte) bool {
return (c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z') || (c >= '0' && c <= '9') || c == '_'
}

View File

@ -1,241 +0,0 @@
package allow
import (
"testing"
)
func TestGQLName1(t *testing.T) {
var q = `
query {
products(
distinct: [price]
where: { id: { and: { greater_or_equals: 20, lt: 28 } } }
) { id name } }`
name := QueryName(q)
if name != "" {
t.Fatal("Name should be empty, not ", name)
}
}
func TestGQLName2(t *testing.T) {
var q = `
query hakuna_matata
{
products(
distinct: [price]
where: { id: { and: { greater_or_equals: 20, lt: 28 } } }
) {
id
name
}
}`
name := QueryName(q)
if name != "hakuna_matata" {
t.Fatal("Name should be 'hakuna_matata', not ", name)
}
}
func TestGQLName3(t *testing.T) {
var q = `
mutation means{ users { id } }`
// var v2 = ` { products( limit: 30, order_by: { price: desc }, distinct: [ price ] where: { id: { and: { greater_or_equals: 20, lt: 28 } } }) { id name price user { id email } } } `
name := QueryName(q)
if name != "means" {
t.Fatal("Name should be 'means', not ", name)
}
}
func TestGQLName4(t *testing.T) {
var q = `
query no_worries
users {
id
}
}`
name := QueryName(q)
if name != "no_worries" {
t.Fatal("Name should be 'no_worries', not ", name)
}
}
func TestGQLName5(t *testing.T) {
var q = `
{
users {
id
}
}`
name := QueryName(q)
if len(name) != 0 {
t.Fatal("Name should be empty, not ", name)
}
}
func TestParse1(t *testing.T) {
var al = `
# Hello world
variables {
"data": {
"slug": "",
"body": "",
"post": {
"connect": {
"slug": ""
}
}
}
}
mutation createComment {
comment(insert: $data) {
slug
body
createdAt: created_at
totalVotes: cached_votes_total
totalReplies: cached_replies_total
vote: comment_vote(where: {user_id: {eq: $user_id}}) {
created_at
__typename
}
author: user {
slug
firstName: first_name
lastName: last_name
pictureURL: picture_url
bio
__typename
}
__typename
}
}
# Query named createPost
query createPost {
post(insert: $data) {
slug
body
published
createdAt: created_at
totalVotes: cached_votes_total
totalComments: cached_comments_total
vote: post_vote(where: {user_id: {eq: $user_id}}) {
created_at
__typename
}
author: user {
slug
firstName: first_name
lastName: last_name
pictureURL: picture_url
bio
__typename
}
__typename
}
}`
_, err := parse(al, "allow.list")
if err != nil {
t.Fatal(err)
}
}
func TestParse2(t *testing.T) {
var al = `
/* Hello world */
variables {
"data": {
"slug": "",
"body": "",
"post": {
"connect": {
"slug": ""
}
}
}
}
mutation createComment {
comment(insert: $data) {
slug
body
createdAt: created_at
totalVotes: cached_votes_total
totalReplies: cached_replies_total
vote: comment_vote(where: {user_id: {eq: $user_id}}) {
created_at
__typename
}
author: user {
slug
firstName: first_name
lastName: last_name
pictureURL: picture_url
bio
__typename
}
__typename
}
}
/*
Query named createPost
*/
variables {
"data": {
"thread": {
"connect": {
"slug": ""
}
},
"slug": "",
"published": false,
"body": ""
}
}
query createPost {
post(insert: $data) {
slug
body
published
createdAt: created_at
totalVotes: cached_votes_total
totalComments: cached_comments_total
vote: post_vote(where: {user_id: {eq: $user_id}}) {
created_at
__typename
}
author: user {
slug
firstName: first_name
lastName: last_name
pictureURL: picture_url
bio
__typename
}
__typename
}
}`
_, err := parse(al, "allow.list")
if err != nil {
t.Fatal(err)
}
}

View File

@ -1,15 +0,0 @@
package allow
import "testing"
func TestFuzzCrashers(t *testing.T) {
var crashers = []string{
"query",
"q",
"que",
}
for _, f := range crashers {
_ = QueryName(f)
}
}

View File

@ -1,80 +0,0 @@
// cryptopasta - basic cryptography examples
//
// Written in 2015 by George Tankersley <george.tankersley@gmail.com>
//
// To the extent possible under law, the author(s) have dedicated all copyright
// and related and neighboring rights to this software to the public domain
// worldwide. This software is distributed without any warranty.
//
// You should have received a copy of the CC0 Public Domain Dedication along
// with this software. If not, see // <http://creativecommons.org/publicdomain/zero/1.0/>.
// Provides symmetric authenticated encryption using 256-bit AES-GCM with a random nonce.
package crypto
import (
"crypto/aes"
"crypto/cipher"
"crypto/rand"
"errors"
"io"
)
// NewEncryptionKey generates a random 256-bit key for Encrypt() and
// Decrypt(). It panics if the source of randomness fails.
func NewEncryptionKey() [32]byte {
key := [32]byte{}
_, err := io.ReadFull(rand.Reader, key[:])
if err != nil {
panic(err)
}
return key
}
// Encrypt encrypts data using 256-bit AES-GCM. This both hides the content of
// the data and provides a check that it hasn't been altered. Output takes the
// form nonce|ciphertext|tag where '|' indicates concatenation.
func Encrypt(plaintext []byte, key *[32]byte) (ciphertext []byte, err error) {
block, err := aes.NewCipher(key[:])
if err != nil {
return nil, err
}
gcm, err := cipher.NewGCM(block)
if err != nil {
return nil, err
}
nonce := make([]byte, gcm.NonceSize())
_, err = io.ReadFull(rand.Reader, nonce)
if err != nil {
return nil, err
}
return gcm.Seal(nonce, nonce, plaintext, nil), nil
}
// Decrypt decrypts data using 256-bit AES-GCM. This both hides the content of
// the data and provides a check that it hasn't been altered. Expects input
// form nonce|ciphertext|tag where '|' indicates concatenation.
func Decrypt(ciphertext []byte, key *[32]byte) (plaintext []byte, err error) {
block, err := aes.NewCipher(key[:])
if err != nil {
return nil, err
}
gcm, err := cipher.NewGCM(block)
if err != nil {
return nil, err
}
if len(ciphertext) < gcm.NonceSize() {
return nil, errors.New("malformed ciphertext")
}
return gcm.Open(nil,
ciphertext[:gcm.NonceSize()],
ciphertext[gcm.NonceSize():],
nil,
)
}

View File

@ -1,88 +0,0 @@
package cockraochdb_test
import (
"database/sql"
"fmt"
"io/ioutil"
"log"
"os"
"os/exec"
"regexp"
"sync/atomic"
"testing"
integration_tests "github.com/dosco/super-graph/core/internal/integration_tests"
_ "github.com/jackc/pgx/v4/stdlib"
"github.com/stretchr/testify/require"
)
func TestCockroachDB(t *testing.T) {
dir, err := ioutil.TempDir("", "temp-cockraochdb-")
if err != nil {
log.Fatal(err)
}
cmd := exec.Command("cockroach", "start", "--insecure", "--listen-addr", ":0", "--http-addr", ":0", "--store=path="+dir)
finder := &urlFinder{
c: make(chan bool),
}
cmd.Stdout = finder
cmd.Stderr = ioutil.Discard
err = cmd.Start()
if err != nil {
t.Skip("is CockroachDB installed?: " + err.Error())
return
}
fmt.Println("started temporary cockroach db")
stopped := int32(0)
stopDatabase := func() {
fmt.Println("stopping temporary cockroach db")
if atomic.CompareAndSwapInt32(&stopped, 0, 1) {
if err := cmd.Process.Kill(); err != nil {
log.Fatal(err)
}
if _, err := cmd.Process.Wait(); err != nil {
log.Fatal(err)
}
os.RemoveAll(dir)
}
}
defer stopDatabase()
// Wait till we figure out the URL we should connect to...
<-finder.c
db, err := sql.Open("pgx", finder.URL)
if err != nil {
stopDatabase()
require.NoError(t, err)
}
integration_tests.SetupSchema(t, db)
integration_tests.TestSuperGraph(t, db, func(t *testing.T) {
if t.Name() == "TestCockroachDB/nested_insert" {
t.Skip("nested inserts currently not working yet on cockroach db")
}
})
}
type urlFinder struct {
c chan bool
done bool
URL string
}
func (finder *urlFinder) Write(p []byte) (n int, err error) {
s := string(p)
urlRegex := regexp.MustCompile(`\nsql:\s+(postgresql:[^\s]+)\n`)
if !finder.done {
submatch := urlRegex.FindAllStringSubmatch(s, -1)
if submatch != nil {
finder.URL = submatch[0][1]
finder.done = true
close(finder.c)
}
}
return len(p), nil
}

View File

@ -1,249 +0,0 @@
package integration_tests
import (
"context"
"database/sql"
"encoding/json"
"io/ioutil"
"testing"
"github.com/dosco/super-graph/core"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
func SetupSchema(t *testing.T, db *sql.DB) {
_, err := db.Exec(`
CREATE TABLE users (
id integer PRIMARY KEY,
full_name text
)`)
require.NoError(t, err)
_, err = db.Exec(`CREATE TABLE product (
id integer PRIMARY KEY,
name text,
weight float
)`)
require.NoError(t, err)
_, err = db.Exec(`CREATE TABLE line_item (
id integer PRIMARY KEY,
product integer REFERENCES product(id),
quantity integer,
price float
)`)
require.NoError(t, err)
}
func DropSchema(t *testing.T, db *sql.DB) {
_, err := db.Exec(`DROP TABLE IF EXISTS line_item`)
require.NoError(t, err)
_, err = db.Exec(`DROP TABLE IF EXISTS product`)
require.NoError(t, err)
_, err = db.Exec(`DROP TABLE IF EXISTS users`)
require.NoError(t, err)
}
func TestSuperGraph(t *testing.T, db *sql.DB, before func(t *testing.T)) {
config := core.Config{}
config.UseAllowList = false
config.AllowListFile = "./allow.list"
config.RolesQuery = `SELECT * FROM users WHERE id = $user_id`
sg, err := core.NewSuperGraph(&config, db)
require.NoError(t, err)
ctx := context.Background()
t.Run("seed fixtures", func(t *testing.T) {
before(t)
res, err := sg.GraphQL(ctx,
`mutation { products (insert: $products) { id } }`,
json.RawMessage(`{"products":[
{"id":1, "name":"Charmin Ultra Soft", "weight": 0.5},
{"id":2, "name":"Hand Sanitizer", "weight": 0.2},
{"id":3, "name":"Case of Corona", "weight": 1.2}
]}`))
require.NoError(t, err, res.SQL())
require.Equal(t, `{"products": [{"id": 1}, {"id": 2}, {"id": 3}]}`, string(res.Data))
res, err = sg.GraphQL(ctx,
`mutation { line_items (insert: $line_items) { id } }`,
json.RawMessage(`{"line_items":[
{"id":5001, "product":1, "price":6.95, "quantity":10},
{"id":5002, "product":2, "price":10.99, "quantity":2}
]}`))
require.NoError(t, err, res.SQL())
require.Equal(t, `{"line_items": [{"id": 5001}, {"id": 5002}]}`, string(res.Data))
})
t.Run("get line item", func(t *testing.T) {
before(t)
res, err := sg.GraphQL(ctx,
`query { line_item(id:$id) { id, price, quantity } }`,
json.RawMessage(`{"id":5001}`))
require.NoError(t, err, res.SQL())
require.Equal(t, `{"line_item": {"id": 5001, "price": 6.95, "quantity": 10}}`, string(res.Data))
})
t.Run("get line items", func(t *testing.T) {
before(t)
res, err := sg.GraphQL(ctx,
`query { line_items { id, price, quantity } }`,
json.RawMessage(`{}`))
require.NoError(t, err, res.SQL())
require.Equal(t, `{"line_items": [{"id": 5001, "price": 6.95, "quantity": 10}, {"id": 5002, "price": 10.99, "quantity": 2}]}`, string(res.Data))
})
t.Run("update line item", func(t *testing.T) {
before(t)
res, err := sg.GraphQL(ctx,
`mutation { line_item(update:$update, id:$id) { id } }`,
json.RawMessage(`{"id":5001, "update":{"quantity":20}}`))
require.NoError(t, err, res.SQL())
require.Equal(t, `{"line_item": {"id": 5001}}`, string(res.Data))
res, err = sg.GraphQL(ctx,
`query { line_item(id:$id) { id, price, quantity } }`,
json.RawMessage(`{"id":5001}`))
require.NoError(t, err, res.SQL())
require.Equal(t, `{"line_item": {"id": 5001, "price": 6.95, "quantity": 20}}`, string(res.Data))
})
t.Run("delete line item", func(t *testing.T) {
before(t)
res, err := sg.GraphQL(ctx,
`mutation { line_item(delete:true, id:$id) { id } }`,
json.RawMessage(`{"id":5002}`))
require.NoError(t, err, res.SQL())
require.Equal(t, `{"line_item": {"id": 5002}}`, string(res.Data))
res, err = sg.GraphQL(ctx,
`query { line_items { id, price, quantity } }`,
json.RawMessage(`{}`))
require.NoError(t, err, res.SQL())
require.Equal(t, `{"line_items": [{"id": 5001, "price": 6.95, "quantity": 20}]}`, string(res.Data))
})
t.Run("nested insert", func(t *testing.T) {
before(t)
res, err := sg.GraphQL(ctx,
`mutation { line_items (insert: $line_item) { id, product { name } } }`,
json.RawMessage(`{"line_item":
{"id":5003, "product": { "connect": { "id": 1} }, "price":10.95, "quantity":15}
}`))
require.NoError(t, err, res.SQL())
require.Equal(t, `{"line_items": [{"id": 5003, "product": {"name": "Charmin Ultra Soft"}}]}`, string(res.Data))
})
t.Run("schema introspection", func(t *testing.T) {
before(t)
schema, err := sg.GraphQLSchema()
require.NoError(t, err)
// Uncomment the following line if you need to regenerate the expected schema.
//ioutil.WriteFile("../introspection.graphql", []byte(schema), 0644)
expected, err := ioutil.ReadFile("../introspection.graphql")
require.NoError(t, err)
assert.Equal(t, string(expected), schema)
})
res, err := sg.GraphQL(ctx, introspectionQuery, json.RawMessage(``))
assert.NoError(t, err)
assert.Contains(t, string(res.Data),
`{"queryType":{"name":"Query"},"mutationType":{"name":"Mutation"},"subscriptionType":null,"types":`)
}
const introspectionQuery = `
query IntrospectionQuery {
__schema {
queryType { name }
mutationType { name }
subscriptionType { name }
types {
...FullType
}
directives {
name
description
locations
args {
...InputValue
}
}
}
}
fragment FullType on __Type {
kind
name
description
fields(includeDeprecated: true) {
name
description
args {
...InputValue
}
type {
...TypeRef
}
isDeprecated
deprecationReason
}
inputFields {
...InputValue
}
interfaces {
...TypeRef
}
enumValues(includeDeprecated: true) {
name
description
isDeprecated
deprecationReason
}
possibleTypes {
...TypeRef
}
}
fragment InputValue on __InputValue {
name
description
type { ...TypeRef }
defaultValue
}
fragment TypeRef on __Type {
kind
name
ofType {
kind
name
ofType {
kind
name
ofType {
kind
name
ofType {
kind
name
ofType {
kind
name
ofType {
kind
name
ofType {
kind
name
}
}
}
}
}
}
}
}
`

View File

@ -1,319 +0,0 @@
input FloatExpression {
contained_in:String!
contains:[Float!]!
eq:Float!
equals:Float!
greater_or_equals:Float!
greater_than:Float!
gt:Float!
gte:Float!
has_key:Float!
has_key_all:[Float!]!
has_key_any:[Float!]!
ilike:String!
in:[Float!]!
is_null:Boolean!
lesser_or_equals:Float!
lesser_than:Float!
like:String!
lt:Float!
lte:Float!
neq:Float!
nilike:String!
nin:[Float!]!
nlike:String!
not_equals:Float!
not_ilike:String!
not_in:[Float!]!
not_like:String!
not_similar:String!
nsimilar:String!
similar:String!
}
input IntExpression {
contained_in:String!
contains:[Int!]!
eq:Int!
equals:Int!
greater_or_equals:Int!
greater_than:Int!
gt:Int!
gte:Int!
has_key:Int!
has_key_all:[Int!]!
has_key_any:[Int!]!
ilike:String!
in:[Int!]!
is_null:Boolean!
lesser_or_equals:Int!
lesser_than:Int!
like:String!
lt:Int!
lte:Int!
neq:Int!
nilike:String!
nin:[Int!]!
nlike:String!
not_equals:Int!
not_ilike:String!
not_in:[Int!]!
not_like:String!
not_similar:String!
nsimilar:String!
similar:String!
}
type Mutation {
line_item(
"To sort or ordering results just use the order_by argument. This can be combined with where, search, etc to build complex queries to fit you needs."
order_by:line_itemOrderBy!, where:line_itemExpression!, limit:Int!, offset:Int!, first:Int!, last:Int!, before:String, after:String,
"Finds the record by the primary key"
id:Int!, insert:line_itemInput, update:line_itemInput, upsert:line_itemInput
):line_itemOutput
line_items(
"To sort or ordering results just use the order_by argument. This can be combined with where, search, etc to build complex queries to fit you needs."
order_by:line_itemOrderBy!, where:line_itemExpression!, limit:Int!, offset:Int!, first:Int!, last:Int!, before:String, after:String,
"Finds the record by the primary key"
id:Int!, insert:line_itemInput, update:line_itemInput, upsert:line_itemInput, inserts:[line_itemInput!]!, updates:[line_itemInput!]!, upserts:[line_itemInput!]!
):line_itemOutput
product(
"To sort or ordering results just use the order_by argument. This can be combined with where, search, etc to build complex queries to fit you needs."
order_by:productOrderBy!, where:productExpression!, limit:Int!, offset:Int!, first:Int!, last:Int!, before:String, after:String,
"Finds the record by the primary key"
id:Int!, insert:productInput, update:productInput, upsert:productInput
):productOutput
products(
"To sort or ordering results just use the order_by argument. This can be combined with where, search, etc to build complex queries to fit you needs."
order_by:productOrderBy!, where:productExpression!, limit:Int!, offset:Int!, first:Int!, last:Int!, before:String, after:String,
"Finds the record by the primary key"
id:Int!, insert:productInput, update:productInput, upsert:productInput, inserts:[productInput!]!, updates:[productInput!]!, upserts:[productInput!]!
):productOutput
user(
"To sort or ordering results just use the order_by argument. This can be combined with where, search, etc to build complex queries to fit you needs."
order_by:userOrderBy!, where:userExpression!, limit:Int!, offset:Int!, first:Int!, last:Int!, before:String, after:String,
"Finds the record by the primary key"
id:Int!, insert:userInput, update:userInput, upsert:userInput
):userOutput
users(
"To sort or ordering results just use the order_by argument. This can be combined with where, search, etc to build complex queries to fit you needs."
order_by:userOrderBy!, where:userExpression!, limit:Int!, offset:Int!, first:Int!, last:Int!, before:String, after:String,
"Finds the record by the primary key"
id:Int!, insert:userInput, update:userInput, upsert:userInput, inserts:[userInput!]!, updates:[userInput!]!, upserts:[userInput!]!
):userOutput
}
enum OrderDirection {
asc
desc
}
type Query {
line_item(
"To sort or ordering results just use the order_by argument. This can be combined with where, search, etc to build complex queries to fit you needs."
order_by:line_itemOrderBy!, where:line_itemExpression!, limit:Int!, offset:Int!, first:Int!, last:Int!, before:String, after:String,
"Finds the record by the primary key"
id:Int!
):line_itemOutput
line_items(
"To sort or ordering results just use the order_by argument. This can be combined with where, search, etc to build complex queries to fit you needs."
order_by:line_itemOrderBy!, where:line_itemExpression!, limit:Int!, offset:Int!, first:Int!, last:Int!, before:String, after:String,
"Finds the record by the primary key"
id:Int!
):[line_itemOutput!]!
product(
"To sort or ordering results just use the order_by argument. This can be combined with where, search, etc to build complex queries to fit you needs."
order_by:productOrderBy!, where:productExpression!, limit:Int!, offset:Int!, first:Int!, last:Int!, before:String, after:String,
"Finds the record by the primary key"
id:Int!
):productOutput
products(
"To sort or ordering results just use the order_by argument. This can be combined with where, search, etc to build complex queries to fit you needs."
order_by:productOrderBy!, where:productExpression!, limit:Int!, offset:Int!, first:Int!, last:Int!, before:String, after:String,
"Finds the record by the primary key"
id:Int!
):[productOutput!]!
user(
"To sort or ordering results just use the order_by argument. This can be combined with where, search, etc to build complex queries to fit you needs."
order_by:userOrderBy!, where:userExpression!, limit:Int!, offset:Int!, first:Int!, last:Int!, before:String, after:String,
"Finds the record by the primary key"
id:Int!
):userOutput
users(
"To sort or ordering results just use the order_by argument. This can be combined with where, search, etc to build complex queries to fit you needs."
order_by:userOrderBy!, where:userExpression!, limit:Int!, offset:Int!, first:Int!, last:Int!, before:String, after:String,
"Finds the record by the primary key"
id:Int!
):[userOutput!]!
}
input StringExpression {
contained_in:String!
contains:[String!]!
eq:String!
equals:String!
greater_or_equals:String!
greater_than:String!
gt:String!
gte:String!
has_key:String!
has_key_all:[String!]!
has_key_any:[String!]!
ilike:String!
in:[String!]!
is_null:Boolean!
lesser_or_equals:String!
lesser_than:String!
like:String!
lt:String!
lte:String!
neq:String!
nilike:String!
nin:[String!]!
nlike:String!
not_equals:String!
not_ilike:String!
not_in:[String!]!
not_like:String!
not_similar:String!
nsimilar:String!
similar:String!
}
input line_itemExpression {
and:line_itemExpression!
id:IntExpression!
not:line_itemExpression!
or:line_itemExpression!
price:FloatExpression!
product:IntExpression!
quantity:IntExpression!
}
input line_itemInput {
id:Int!
price:Float
product:Int
quantity:Int
}
input line_itemOrderBy {
id:OrderDirection!
price:OrderDirection!
product:OrderDirection!
quantity:OrderDirection!
}
type line_itemOutput {
avg_id:Int!
avg_price:Float
avg_product:Int
avg_quantity:Int
count_id:Int!
count_price:Float
count_product:Int
count_quantity:Int
id:Int!
max_id:Int!
max_price:Float
max_product:Int
max_quantity:Int
min_id:Int!
min_price:Float
min_product:Int
min_quantity:Int
price:Float
product:Int
quantity:Int
stddev_id:Int!
stddev_pop_id:Int!
stddev_pop_price:Float
stddev_pop_product:Int
stddev_pop_quantity:Int
stddev_price:Float
stddev_product:Int
stddev_quantity:Int
stddev_samp_id:Int!
stddev_samp_price:Float
stddev_samp_product:Int
stddev_samp_quantity:Int
var_pop_id:Int!
var_pop_price:Float
var_pop_product:Int
var_pop_quantity:Int
var_samp_id:Int!
var_samp_price:Float
var_samp_product:Int
var_samp_quantity:Int
variance_id:Int!
variance_price:Float
variance_product:Int
variance_quantity:Int
}
input productExpression {
and:productExpression!
id:IntExpression!
name:StringExpression!
not:productExpression!
or:productExpression!
weight:FloatExpression!
}
input productInput {
id:Int!
name:String
weight:Float
}
input productOrderBy {
id:OrderDirection!
name:OrderDirection!
weight:OrderDirection!
}
type productOutput {
avg_id:Int!
avg_weight:Float
count_id:Int!
count_weight:Float
id:Int!
max_id:Int!
max_weight:Float
min_id:Int!
min_weight:Float
name:String
stddev_id:Int!
stddev_pop_id:Int!
stddev_pop_weight:Float
stddev_samp_id:Int!
stddev_samp_weight:Float
stddev_weight:Float
var_pop_id:Int!
var_pop_weight:Float
var_samp_id:Int!
var_samp_weight:Float
variance_id:Int!
variance_weight:Float
weight:Float
}
input userExpression {
and:userExpression!
full_name:StringExpression!
id:IntExpression!
not:userExpression!
or:userExpression!
}
input userInput {
full_name:String
id:Int!
}
input userOrderBy {
full_name:OrderDirection!
id:OrderDirection!
}
type userOutput {
avg_id:Int!
count_id:Int!
full_name:String
id:Int!
max_id:Int!
min_id:Int!
stddev_id:Int!
stddev_pop_id:Int!
stddev_samp_id:Int!
var_pop_id:Int!
var_samp_id:Int!
variance_id:Int!
}
schema {
mutation: Mutation
query: Query
}

View File

@ -1,27 +0,0 @@
package cockraochdb_test
import (
"database/sql"
"os"
"testing"
integration_tests "github.com/dosco/super-graph/core/internal/integration_tests"
_ "github.com/jackc/pgx/v4/stdlib"
"github.com/stretchr/testify/require"
)
func TestCockroachDB(t *testing.T) {
url, found := os.LookupEnv("SG_POSTGRESQL_TEST_URL")
if !found {
t.Skip("set the SG_POSTGRESQL_TEST_URL env variable if you want to run integration tests against a PostgreSQL database")
} else {
db, err := sql.Open("pgx", url)
require.NoError(t, err)
integration_tests.DropSchema(t, db)
integration_tests.SetupSchema(t, db)
integration_tests.TestSuperGraph(t, db, func(t *testing.T) {
})
}
}

View File

@ -1,214 +0,0 @@
package psql
import (
"errors"
"io"
"strings"
"github.com/dosco/super-graph/core/internal/qcode"
)
func (c *compilerContext) renderBaseColumns(
sel *qcode.Select,
ti *DBTableInfo,
childCols []*qcode.Column) ([]int, bool, error) {
var realColsRendered []int
colcount := (len(sel.Cols) + len(sel.OrderBy) + 1)
colmap := make(map[string]struct{}, colcount)
isSearch := sel.Args["search"] != nil
isCursorPaged := sel.Paging.Type != qcode.PtOffset
isAgg := false
i := 0
for n, col := range sel.Cols {
cn := col.Name
colmap[cn] = struct{}{}
_, isRealCol := ti.ColMap[cn]
if isRealCol {
c.renderComma(i)
realColsRendered = append(realColsRendered, n)
colWithTable(c.w, ti.Name, cn)
} else {
switch {
case isSearch && cn == "search_rank":
if err := c.renderColumnSearchRank(sel, ti, col, i); err != nil {
return nil, false, err
}
case isSearch && strings.HasPrefix(cn, "search_headline_"):
if err := c.renderColumnSearchHeadline(sel, ti, col, i); err != nil {
return nil, false, err
}
case cn == "__typename":
if err := c.renderColumnTypename(sel, ti, col, i); err != nil {
return nil, false, err
}
case strings.HasSuffix(cn, "_cursor"):
continue
default:
if err := c.renderColumnFunction(sel, ti, col, i); err != nil {
return nil, false, err
}
isAgg = true
}
}
i++
}
if isCursorPaged {
if _, ok := colmap[ti.PrimaryCol.Key]; !ok {
colmap[ti.PrimaryCol.Key] = struct{}{}
c.renderComma(i)
colWithTable(c.w, ti.Name, ti.PrimaryCol.Name)
}
i++
}
for _, ob := range sel.OrderBy {
if _, ok := colmap[ob.Col]; ok {
continue
}
colmap[ob.Col] = struct{}{}
c.renderComma(i)
colWithTable(c.w, ti.Name, ob.Col)
i++
}
for _, col := range childCols {
if _, ok := colmap[col.Name]; ok {
continue
}
c.renderComma(i)
colWithTable(c.w, col.Table, col.Name)
i++
}
return realColsRendered, isAgg, nil
}
func (c *compilerContext) renderColumnSearchRank(sel *qcode.Select, ti *DBTableInfo, col qcode.Column, columnsRendered int) error {
if isColumnBlocked(sel, col.Name) {
return nil
}
if ti.TSVCol == nil {
return errors.New("no ts_vector column found")
}
cn := ti.TSVCol.Name
arg := sel.Args["search"]
c.renderComma(columnsRendered)
//fmt.Fprintf(w, `ts_rank("%s"."%s", websearch_to_tsquery('%s')) AS %s`,
//c.sel.Name, cn, arg.Val, col.Name)
_, _ = io.WriteString(c.w, `ts_rank(`)
colWithTable(c.w, ti.Name, cn)
if c.schema.ver >= 110000 {
_, _ = io.WriteString(c.w, `, websearch_to_tsquery(`)
} else {
_, _ = io.WriteString(c.w, `, to_tsquery(`)
}
c.md.renderValueExp(c.w, Param{Name: arg.Val, Type: "string"})
_, _ = io.WriteString(c.w, `))`)
alias(c.w, col.Name)
return nil
}
func (c *compilerContext) renderColumnSearchHeadline(sel *qcode.Select, ti *DBTableInfo, col qcode.Column, columnsRendered int) error {
cn := col.Name[16:]
if isColumnBlocked(sel, cn) {
return nil
}
arg := sel.Args["search"]
c.renderComma(columnsRendered)
//fmt.Fprintf(w, `ts_headline("%s"."%s", websearch_to_tsquery('%s')) AS %s`,
//c.sel.Name, cn, arg.Val, col.Name)
_, _ = io.WriteString(c.w, `ts_headline(`)
colWithTable(c.w, ti.Name, cn)
if c.schema.ver >= 110000 {
_, _ = io.WriteString(c.w, `, websearch_to_tsquery(`)
} else {
_, _ = io.WriteString(c.w, `, to_tsquery(`)
}
c.md.renderValueExp(c.w, Param{Name: arg.Val, Type: "string"})
_, _ = io.WriteString(c.w, `))`)
alias(c.w, col.Name)
return nil
}
func (c *compilerContext) renderColumnTypename(sel *qcode.Select, ti *DBTableInfo, col qcode.Column, columnsRendered int) error {
if isColumnBlocked(sel, col.Name) {
return nil
}
c.renderComma(columnsRendered)
_, _ = io.WriteString(c.w, `(`)
squoted(c.w, ti.Name)
_, _ = io.WriteString(c.w, ` :: text)`)
alias(c.w, col.Name)
return nil
}
func (c *compilerContext) renderColumnFunction(sel *qcode.Select, ti *DBTableInfo, col qcode.Column, columnsRendered int) error {
pl := funcPrefixLen(c.schema.fm, col.Name)
// if pl == 0 {
// //fmt.Fprintf(w, `'%s not defined' AS %s`, cn, col.Name)
// _, _ = io.WriteString(c.w, `'`)
// _, _ = io.WriteString(c.w, col.Name)
// _, _ = io.WriteString(c.w, ` not defined'`)
// alias(c.w, col.Name)
// }
if pl == 0 || !sel.Functions {
return nil
}
cn := col.Name[pl:]
if isColumnBlocked(sel, cn) {
return nil
}
fn := col.Name[:pl-1]
c.renderComma(columnsRendered)
//fmt.Fprintf(w, `%s("%s"."%s") AS %s`, fn, c.sel.Name, cn, col.Name)
_, _ = io.WriteString(c.w, fn)
_, _ = io.WriteString(c.w, `(`)
colWithTable(c.w, ti.Name, cn)
_, _ = io.WriteString(c.w, `)`)
alias(c.w, col.Name)
return nil
}
func (c *compilerContext) renderComma(columnsRendered int) {
if columnsRendered != 0 {
_, _ = io.WriteString(c.w, `, `)
}
}
func isColumnBlocked(sel *qcode.Select, name string) bool {
if len(sel.Allowed) != 0 {
if _, ok := sel.Allowed[name]; !ok {
return true
}
}
return false
}

View File

@ -1,134 +0,0 @@
// +build gofuzz
package psql
import (
"encoding/json"
"github.com/dosco/super-graph/core/internal/qcode"
)
var (
qcompileTest, _ = qcode.NewCompiler(qcode.Config{})
schema, _ = GetTestSchema()
vars = map[string]string{
"admin_account_id": "5",
}
pcompileTest = NewCompiler(Config{
Schema: schema,
Vars: vars,
})
)
// FuzzerEntrypoint for Fuzzbuzz
func Fuzz(data []byte) int {
err1 := query(data)
err2 := insert(data)
err3 := update(data)
err4 := delete(data)
if err1 != nil || err2 != nil || err3 != nil || err4 != nil {
return 0
}
return 1
}
func query(data []byte) error {
gql := data
qc, err1 := qcompileTest.Compile(gql, "user")
vars := map[string]json.RawMessage{
"data": json.RawMessage(data),
}
_, _, err2 := pcompileTest.CompileEx(qc, vars)
if err1 != nil {
return err1
} else {
return err2
}
}
func insert(data []byte) error {
gql := `mutation {
product(insert: $data) {
id
name
user {
id
full_name
email
}
}
}`
qc, err := qcompileTest.Compile([]byte(gql), "user")
if err != nil {
panic("qcompile can't fail")
}
vars := map[string]json.RawMessage{
"data": json.RawMessage(data),
}
_, _, err = pcompileTest.CompileEx(qc, vars)
return err
}
func update(data []byte) error {
gql := `mutation {
product(insert: $data) {
id
name
user {
id
full_name
email
}
}
}`
qc, err := qcompileTest.Compile([]byte(gql), "user")
if err != nil {
panic("qcompile can't fail")
}
vars := map[string]json.RawMessage{
"data": json.RawMessage(data),
}
_, _, err = pcompileTest.CompileEx(qc, vars)
return err
}
func delete(data []byte) error {
gql := `mutation {
product(insert: $data) {
id
name
user {
id
full_name
email
}
}
}`
qc, err := qcompileTest.Compile([]byte(gql), "user")
if err != nil {
panic("qcompile can't fail")
}
vars := map[string]json.RawMessage{
"data": json.RawMessage(data),
}
_, _, err = pcompileTest.CompileEx(qc, vars)
return err
}

View File

@ -1,20 +0,0 @@
// +build gofuzz
package psql
import (
"testing"
)
var ret int
func TestFuzzCrashers(t *testing.T) {
var crashers = []string{
"{\"connect\":{}}",
"q(q{q{q{q{q{q{q{q{",
}
for _, f := range crashers {
ret = Fuzz([]byte(f))
}
}

View File

@ -1,271 +0,0 @@
package psql_test
import (
"encoding/json"
"testing"
)
func simpleInsert(t *testing.T) {
gql := `mutation {
user(insert: $data) {
id
}
}`
vars := map[string]json.RawMessage{
"data": json.RawMessage(`{"email": "reannagreenholt@orn.com", "full_name": "Flo Barton"}`),
}
compileGQLToPSQL(t, gql, vars, "user")
}
func singleInsert(t *testing.T) {
gql := `mutation {
product(id: $id, insert: $insert) {
id
name
}
}`
vars := map[string]json.RawMessage{
"insert": json.RawMessage(` { "name": "my_name", "price": 6.95, "description": "my_desc", "user_id": 5 }`),
}
compileGQLToPSQL(t, gql, vars, "anon")
}
func bulkInsert(t *testing.T) {
gql := `mutation {
product(name: "test", id: $id, insert: $insert) {
id
name
}
}`
vars := map[string]json.RawMessage{
"insert": json.RawMessage(` [{ "name": "my_name", "description": "my_desc" }]`),
}
compileGQLToPSQL(t, gql, vars, "anon")
}
func simpleInsertWithPresets(t *testing.T) {
gql := `mutation {
product(insert: $data) {
id
}
}`
vars := map[string]json.RawMessage{
"data": json.RawMessage(`{"name": "Tomato", "price": 5.76}`),
}
compileGQLToPSQL(t, gql, vars, "user")
}
func nestedInsertManyToMany(t *testing.T) {
gql := `mutation {
purchase(insert: $data) {
sale_type
quantity
due_date
customer {
id
full_name
email
}
product {
id
name
price
}
}
}`
vars := map[string]json.RawMessage{
"data": json.RawMessage(` {
"sale_type": "bought",
"quantity": 5,
"due_date": "now",
"customer": {
"email": "thedude@rug.com",
"full_name": "The Dude"
},
"product": {
"name": "Apple",
"price": 1.25
}
}
`),
}
compileGQLToPSQL(t, gql, vars, "admin")
}
func nestedInsertOneToMany(t *testing.T) {
gql := `mutation {
user(insert: $data) {
id
full_name
email
product {
id
name
price
}
}
}`
vars := map[string]json.RawMessage{
"data": json.RawMessage(`{
"email": "thedude@rug.com",
"full_name": "The Dude",
"created_at": "now",
"updated_at": "now",
"product": {
"name": "Apple",
"price": 1.25,
"created_at": "now",
"updated_at": "now"
}
}`),
}
compileGQLToPSQL(t, gql, vars, "admin")
}
func nestedInsertOneToOne(t *testing.T) {
gql := `mutation {
product(insert: $data) {
id
name
user {
id
full_name
email
}
}
}`
vars := map[string]json.RawMessage{
"data": json.RawMessage(`{
"name": "Apple",
"price": 1.25,
"created_at": "now",
"updated_at": "now",
"user": {
"hey": {
"now": "what's the matter"
},
"email": "thedude@rug.com",
"full_name": "The Dude",
"created_at": "now",
"updated_at": "now"
}
}`),
}
compileGQLToPSQL(t, gql, vars, "admin")
}
func nestedInsertOneToManyWithConnect(t *testing.T) {
gql := `mutation {
user(insert: $data) {
id
full_name
email
product {
id
name
price
}
}
}`
vars := map[string]json.RawMessage{
"data": json.RawMessage(`{
"email": "thedude@rug.com",
"full_name": "The Dude",
"created_at": "now",
"updated_at": "now",
"product": {
"connect": { "id": 5 }
}
}`),
}
compileGQLToPSQL(t, gql, vars, "admin")
}
func nestedInsertOneToOneWithConnect(t *testing.T) {
gql := `mutation {
product(insert: $data) {
id
name
tags {
id
name
}
user {
id
full_name
email
}
}
}`
vars := map[string]json.RawMessage{
"data": json.RawMessage(`{
"name": "Apple",
"price": 1.25,
"created_at": "now",
"updated_at": "now",
"user": {
"connect": { "id": 5 }
}
}`),
}
compileGQLToPSQL(t, gql, vars, "admin")
}
func nestedInsertOneToOneWithConnectArray(t *testing.T) {
gql := `mutation {
product(insert: $data) {
id
name
user {
id
full_name
email
}
}
}`
vars := map[string]json.RawMessage{
"data": json.RawMessage(`{
"name": "Apple",
"price": 1.25,
"created_at": "now",
"updated_at": "now",
"user": {
"connect": { "id": [1,2] }
}
}`),
}
compileGQLToPSQL(t, gql, vars, "admin")
}
func TestCompileInsert(t *testing.T) {
t.Run("simpleInsert", simpleInsert)
t.Run("singleInsert", singleInsert)
t.Run("bulkInsert", bulkInsert)
t.Run("simpleInsertWithPresets", simpleInsertWithPresets)
t.Run("nestedInsertManyToMany", nestedInsertManyToMany)
t.Run("nestedInsertOneToMany", nestedInsertOneToMany)
t.Run("nestedInsertOneToOne", nestedInsertOneToOne)
t.Run("nestedInsertOneToManyWithConnect", nestedInsertOneToManyWithConnect)
t.Run("nestedInsertOneToOneWithConnect", nestedInsertOneToOneWithConnect)
t.Run("nestedInsertOneToOneWithConnectArray", nestedInsertOneToOneWithConnectArray)
}

View File

@ -1,61 +0,0 @@
package psql
import (
"io"
)
func (md *Metadata) RenderVar(w io.Writer, vv string) {
f, s := -1, 0
for i := range vv {
v := vv[i]
switch {
case (i > 0 && vv[i-1] != '\\' && v == '$') || v == '$':
if (i - s) > 0 {
_, _ = io.WriteString(w, vv[s:i])
}
f = i
case (v < 'a' && v > 'z') &&
(v < 'A' && v > 'Z') &&
(v < '0' && v > '9') &&
v != '_' &&
f != -1 &&
(i-f) > 1:
md.renderValueExp(w, Param{Name: vv[f+1 : i]})
s = i
f = -1
}
}
if f != -1 && (len(vv)-f) > 1 {
md.renderValueExp(w, Param{Name: vv[f+1:]})
} else {
_, _ = io.WriteString(w, vv[s:])
}
}
func (md *Metadata) renderValueExp(w io.Writer, p Param) {
_, _ = io.WriteString(w, `$`)
if v, ok := md.pindex[p.Name]; ok {
int32String(w, int32(v))
} else {
md.params = append(md.params, p)
n := len(md.params)
if md.pindex == nil {
md.pindex = make(map[string]int)
}
md.pindex[p.Name] = n
int32String(w, int32(n))
}
}
func (md Metadata) Skipped() uint32 {
return md.skipped
}
func (md Metadata) Params() []Param {
return md.params
}

View File

@ -1,123 +0,0 @@
package psql_test
import (
"encoding/json"
"testing"
)
func singleUpsert(t *testing.T) {
gql := `mutation {
product(upsert: $upsert) {
id
name
}
}`
vars := map[string]json.RawMessage{
"upsert": json.RawMessage(` { "name": "my_name", "description": "my_desc" }`),
}
compileGQLToPSQL(t, gql, vars, "user")
}
func singleUpsertWhere(t *testing.T) {
gql := `mutation {
product(upsert: $upsert, where: { price : { gt: 3 } }) {
id
name
}
}`
vars := map[string]json.RawMessage{
"upsert": json.RawMessage(` { "name": "my_name", "description": "my_desc" }`),
}
compileGQLToPSQL(t, gql, vars, "user")
}
func bulkUpsert(t *testing.T) {
gql := `mutation {
product(upsert: $upsert) {
id
name
}
}`
vars := map[string]json.RawMessage{
"upsert": json.RawMessage(` [{ "name": "my_name", "description": "my_desc" }]`),
}
compileGQLToPSQL(t, gql, vars, "user")
}
func delete(t *testing.T) {
gql := `mutation {
product(delete: true, where: { id: { eq: 1 } }) {
id
name
}
}`
vars := map[string]json.RawMessage{
"update": json.RawMessage(` { "name": "my_name", "description": "my_desc" }`),
}
compileGQLToPSQL(t, gql, vars, "user")
}
// func blockedInsert(t *testing.T) {
// gql := `mutation {
// user(insert: $data) {
// id
// }
// }`
// sql := `WITH "users" AS (WITH "input" AS (SELECT '$1' :: json AS j) INSERT INTO "users" ("full_name", "email") SELECT "full_name", "email" FROM input i, json_populate_record(NULL::users, i.j) t WHERE false RETURNING *) SELECT json_object_agg('user', json_0) FROM (SELECT row_to_json((SELECT "json_row_0" FROM (SELECT "users_0"."id" AS "id") AS "json_row_0")) AS "json_0" FROM (SELECT "users"."id" FROM "users" LIMIT ('1') :: integer) AS "users_0" LIMIT ('1') :: integer) AS "sel_0"`
// vars := map[string]json.RawMessage{
// "data": json.RawMessage(`{"email": "reannagreenholt@orn.com", "full_name": "Flo Barton"}`),
// }
// resSQL, err := compileGQLToPSQL(gql, vars, "bad_dude")
// if err != nil {
// t.Fatal(err)
// }
// fmt.Println(string(resSQL))
// if string(resSQL) != sql {
// t.Fatal(errNotExpected)
// }
// }
// func blockedUpdate(t *testing.T) {
// gql := `mutation {
// user(where: { id: { lt: 5 } }, update: $data) {
// id
// email
// }
// }`
// sql := `WITH "users" AS (WITH "input" AS (SELECT '$1' :: json AS j) UPDATE "users" SET ("full_name", "email") = (SELECT "full_name", "email" FROM input i, json_populate_record(NULL::users, i.j) t) WHERE false RETURNING *) SELECT json_object_agg('user', json_0) FROM (SELECT row_to_json((SELECT "json_row_0" FROM (SELECT "users_0"."id" AS "id", "users_0"."email" AS "email") AS "json_row_0")) AS "json_0" FROM (SELECT "users"."id", "users"."email" FROM "users" LIMIT ('1') :: integer) AS "users_0" LIMIT ('1') :: integer) AS "sel_0"`
// vars := map[string]json.RawMessage{
// "data": json.RawMessage(`{"email": "reannagreenholt@orn.com", "full_name": "Flo Barton"}`),
// }
// resSQL, err := compileGQLToPSQL(gql, vars, "bad_dude")
// if err != nil {
// t.Fatal(err)
// }
// if string(resSQL) != sql {
// t.Fatal(errNotExpected)
// }
// }
func TestCompileMutate(t *testing.T) {
t.Run("singleUpsert", singleUpsert)
t.Run("singleUpsertWhere", singleUpsertWhere)
t.Run("bulkUpsert", bulkUpsert)
t.Run("delete", delete)
// t.Run("blockedInsert", blockedInsert)
// t.Run("blockedUpdate", blockedUpdate)
}

View File

@ -1,241 +0,0 @@
package psql_test
import (
"fmt"
"io/ioutil"
"log"
"os"
"strings"
"testing"
"github.com/dosco/super-graph/core/internal/psql"
"github.com/dosco/super-graph/core/internal/qcode"
)
const (
errNotExpected = "Generated SQL did not match what was expected"
headerMarker = "=== RUN"
commentMarker = "---"
)
var (
qcompile *qcode.Compiler
pcompile *psql.Compiler
expected map[string][]string
)
func TestMain(m *testing.M) {
var err error
qcompile, err = qcode.NewCompiler(qcode.Config{
Blocklist: []string{
"secret",
"password",
"token",
},
})
if err != nil {
log.Fatal(err)
}
err = qcompile.AddRole("user", "product", qcode.TRConfig{
Query: qcode.QueryConfig{
Columns: []string{"id", "name", "price", "users", "customers"},
Filters: []string{
"{ price: { gt: 0 } }",
"{ price: { lt: 8 } }",
},
},
Insert: qcode.InsertConfig{
Presets: map[string]string{
"user_id": "$user_id",
"created_at": "now",
"updated_at": "now",
},
},
Update: qcode.UpdateConfig{
Filters: []string{"{ user_id: { eq: $user_id } }"},
Presets: map[string]string{"updated_at": "now"},
},
Delete: qcode.DeleteConfig{
Filters: []string{
"{ price: { gt: 0 } }",
"{ price: { lt: 8 } }",
},
},
})
if err != nil {
log.Fatal(err)
}
err = qcompile.AddRole("anon", "product", qcode.TRConfig{
Query: qcode.QueryConfig{
Columns: []string{"id", "name"},
},
})
if err != nil {
log.Fatal(err)
}
err = qcompile.AddRole("anon1", "product", qcode.TRConfig{
Query: qcode.QueryConfig{
Columns: []string{"id", "name", "price"},
DisableFunctions: true,
},
})
if err != nil {
log.Fatal(err)
}
err = qcompile.AddRole("user", "users", qcode.TRConfig{
Query: qcode.QueryConfig{
Columns: []string{"id", "full_name", "avatar", "email", "products"},
},
})
if err != nil {
log.Fatal(err)
}
err = qcompile.AddRole("bad_dude", "users", qcode.TRConfig{
Query: qcode.QueryConfig{
Filters: []string{"false"},
DisableFunctions: true,
},
Insert: qcode.InsertConfig{
Filters: []string{"false"},
},
Update: qcode.UpdateConfig{
Filters: []string{"false"},
},
})
if err != nil {
log.Fatal(err)
}
err = qcompile.AddRole("user", "mes", qcode.TRConfig{
Query: qcode.QueryConfig{
Columns: []string{"id", "full_name", "avatar"},
Filters: []string{
"{ id: { eq: $user_id } }",
},
},
})
if err != nil {
log.Fatal(err)
}
err = qcompile.AddRole("user", "customers", qcode.TRConfig{
Query: qcode.QueryConfig{
Columns: []string{"id", "email", "full_name", "products"},
},
})
if err != nil {
log.Fatal(err)
}
schema, err := psql.GetTestSchema()
if err != nil {
log.Fatal(err)
}
vars := map[string]string{
"admin_account_id": "5",
}
pcompile = psql.NewCompiler(psql.Config{
Schema: schema,
Vars: vars,
})
expected = make(map[string][]string)
b, err := ioutil.ReadFile("tests.sql")
if err != nil {
log.Fatal(err)
}
text := string(b)
lines := strings.Split(text, "\n")
var h string
for _, v := range lines {
switch {
case strings.HasPrefix(v, headerMarker):
h = strings.TrimSpace(v[len(headerMarker):])
case strings.HasPrefix(v, commentMarker):
break
default:
v := strings.TrimSpace(v)
if len(v) != 0 {
expected[h] = append(expected[h], v)
}
}
}
os.Exit(m.Run())
}
func compileGQLToPSQL(t *testing.T, gql string, vars psql.Variables, role string) {
generateTestFile := false
if generateTestFile {
var sqlStmts []string
for i := 0; i < 100; i++ {
qc, err := qcompile.Compile([]byte(gql), role)
if err != nil {
t.Fatal(err)
}
_, sqlB, err := pcompile.CompileEx(qc, vars)
if err != nil {
t.Fatal(err)
}
sql := string(sqlB)
match := false
for _, s := range sqlStmts {
if sql == s {
match = true
break
}
}
if !match {
s := string(sql)
sqlStmts = append(sqlStmts, s)
fmt.Println(s)
}
}
return
}
for i := 0; i < 200; i++ {
qc, err := qcompile.Compile([]byte(gql), role)
if err != nil {
t.Fatal(err)
}
_, sqlStmt, err := pcompile.CompileEx(qc, vars)
if err != nil {
t.Fatal(err)
}
failed := true
for _, sql := range expected[t.Name()] {
if string(sqlStmt) == sql {
failed = false
}
}
if failed {
fmt.Println(string(sqlStmt))
t.Fatal(errNotExpected)
}
}
}

View File

@ -1,1370 +0,0 @@
//nolint:errcheck
package psql
import (
"bytes"
"encoding/json"
"errors"
"fmt"
"io"
"strconv"
"strings"
"github.com/dosco/super-graph/core/internal/qcode"
"github.com/dosco/super-graph/core/internal/util"
)
const (
closeBlock = 500
)
type Param struct {
Name string
Type string
IsArray bool
}
type Metadata struct {
skipped uint32
params []Param
pindex map[string]int
}
type compilerContext struct {
md Metadata
w io.Writer
s []qcode.Select
*Compiler
}
type Variables map[string]json.RawMessage
type Config struct {
Schema *DBSchema
Vars map[string]string
}
type Compiler struct {
schema *DBSchema
vars map[string]string
}
func NewCompiler(conf Config) *Compiler {
return &Compiler{
schema: conf.Schema,
vars: conf.Vars,
}
}
func (co *Compiler) AddRelationship(child, parent string, rel *DBRel) error {
return co.schema.SetRel(child, parent, rel)
}
func (co *Compiler) IDColumn(table string) (*DBColumn, error) {
ti, err := co.schema.GetTable(table)
if err != nil {
return nil, err
}
if ti.PrimaryCol == nil {
return nil, fmt.Errorf("no primary key column found")
}
return ti.PrimaryCol, nil
}
func (co *Compiler) CompileEx(qc *qcode.QCode, vars Variables) (Metadata, []byte, error) {
w := &bytes.Buffer{}
metad, err := co.Compile(w, qc, vars)
return metad, w.Bytes(), err
}
func (co *Compiler) Compile(w io.Writer, qc *qcode.QCode, vars Variables) (Metadata, error) {
return co.CompileWithMetadata(w, qc, vars, Metadata{})
}
func (co *Compiler) CompileWithMetadata(w io.Writer, qc *qcode.QCode, vars Variables, md Metadata) (Metadata, error) {
md.skipped = 0
if qc == nil {
return md, fmt.Errorf("qcode is nil")
}
switch qc.Type {
case qcode.QTQuery:
return co.compileQueryWithMetadata(w, qc, vars, md)
case qcode.QTInsert,
qcode.QTUpdate,
qcode.QTDelete,
qcode.QTUpsert:
return co.compileMutation(w, qc, vars)
default:
return Metadata{}, fmt.Errorf("Unknown operation type %d", qc.Type)
}
}
func (co *Compiler) compileQueryWithMetadata(
w io.Writer, qc *qcode.QCode, vars Variables, md Metadata) (Metadata, error) {
if len(qc.Selects) == 0 {
return md, errors.New("empty query")
}
c := &compilerContext{md, w, qc.Selects, co}
st := NewIntStack()
i := 0
io.WriteString(c.w, `SELECT jsonb_build_object(`)
for _, id := range qc.Roots {
if i != 0 {
io.WriteString(c.w, `, `)
}
root := &qc.Selects[id]
if root.SkipRender || len(root.Cols) == 0 {
squoted(c.w, root.FieldName)
io.WriteString(c.w, `, `)
io.WriteString(c.w, `NULL`)
} else {
st.Push(root.ID + closeBlock)
st.Push(root.ID)
c.renderRootSelect(root)
}
i++
}
if st.Len() != 0 {
io.WriteString(c.w, `) as "__root" FROM `)
} else {
io.WriteString(c.w, `) as "__root"`)
return c.md, nil
}
for {
if st.Len() == 0 {
break
}
id := st.Pop()
if id < closeBlock {
sel := &c.s[id]
ti, err := c.schema.GetTable(sel.Name)
if err != nil {
return c.md, err
}
if sel.Type != qcode.STUnion {
if len(sel.Cols) == 0 {
continue
}
if sel.ParentID == -1 {
io.WriteString(c.w, `(`)
} else {
c.renderLateralJoin(sel)
}
if !ti.IsSingular {
c.renderPluralSelect(sel, ti)
}
if err := c.renderSelect(sel, ti, vars); err != nil {
return c.md, err
}
}
for _, cid := range sel.Children {
if hasBit(c.md.skipped, uint32(cid)) {
continue
}
child := &c.s[cid]
if child.SkipRender {
continue
}
st.Push(child.ID + closeBlock)
st.Push(child.ID)
}
} else {
sel := &c.s[(id - closeBlock)]
if sel.Type != qcode.STUnion {
ti, err := c.schema.GetTable(sel.Name)
if err != nil {
return c.md, err
}
io.WriteString(c.w, `)`)
aliasWithID(c.w, "__sr", sel.ID)
io.WriteString(c.w, `)`)
aliasWithID(c.w, "__sj", sel.ID)
if !ti.IsSingular {
io.WriteString(c.w, `)`)
aliasWithID(c.w, "__sj", sel.ID)
}
if sel.ParentID == -1 {
if st.Len() != 0 {
io.WriteString(c.w, `, `)
}
} else {
c.renderLateralJoinClose(sel)
}
}
if sel.Type != qcode.STMember {
if len(sel.Args) != 0 {
for _, v := range sel.Args {
qcode.FreeNode(v)
}
}
}
}
}
return c.md, nil
}
func (c *compilerContext) renderPluralSelect(sel *qcode.Select, ti *DBTableInfo) error {
io.WriteString(c.w, `SELECT coalesce(jsonb_agg("__sj_`)
int32String(c.w, sel.ID)
io.WriteString(c.w, `"."json"), '[]') as "json"`)
if sel.Paging.Type != qcode.PtOffset {
n := 0
// check if primary key already included in order by
// query argument
for _, ob := range sel.OrderBy {
if ob.Col == ti.PrimaryCol.Key {
n = 1
break
}
}
if n == 1 {
n = len(sel.OrderBy)
} else {
n = len(sel.OrderBy) + 1
}
io.WriteString(c.w, `, CONCAT_WS(','`)
for i := 0; i < n; i++ {
io.WriteString(c.w, `, max("__cur_`)
int32String(c.w, int32(i))
io.WriteString(c.w, `")`)
}
io.WriteString(c.w, `) as "cursor"`)
}
io.WriteString(c.w, ` FROM (`)
return nil
}
func (c *compilerContext) renderRootSelect(sel *qcode.Select) error {
io.WriteString(c.w, `'`)
io.WriteString(c.w, sel.FieldName)
io.WriteString(c.w, `', `)
io.WriteString(c.w, `"__sj_`)
int32String(c.w, sel.ID)
io.WriteString(c.w, `"."json"`)
if sel.Paging.Type != qcode.PtOffset {
io.WriteString(c.w, `, '`)
io.WriteString(c.w, sel.FieldName)
io.WriteString(c.w, `_cursor', `)
io.WriteString(c.w, `"__sj_`)
int32String(c.w, sel.ID)
io.WriteString(c.w, `"."cursor"`)
}
return nil
}
func (c *compilerContext) initSelect(sel *qcode.Select, ti *DBTableInfo, vars Variables) ([]*qcode.Column, error) {
cols := make([]*qcode.Column, 0, len(sel.Cols))
colmap := make(map[string]struct{}, len(sel.Cols))
for i := range sel.Cols {
colmap[sel.Cols[i].Name] = struct{}{}
}
for i := range sel.OrderBy {
colmap[sel.OrderBy[i].Col] = struct{}{}
}
if sel.Paging.Type != qcode.PtOffset {
colmap[ti.PrimaryCol.Key] = struct{}{}
addPrimaryKey := true
for _, ob := range sel.OrderBy {
if ob.Col == ti.PrimaryCol.Key {
addPrimaryKey = false
break
}
}
if addPrimaryKey {
ob := &qcode.OrderBy{Col: ti.PrimaryCol.Name, Order: qcode.OrderAsc}
if sel.Paging.Type == qcode.PtBackward {
ob.Order = qcode.OrderDesc
}
sel.OrderBy = append(sel.OrderBy, ob)
}
}
if sel.Paging.Cursor {
c.addSeekPredicate(sel)
}
for _, id := range sel.Children {
child := &c.s[id]
rel, err := c.schema.GetRel(child.Name, ti.Name)
if err != nil {
return nil, err
}
switch rel.Type {
case RelOneToOne, RelOneToMany:
if _, ok := colmap[rel.Right.Col]; !ok {
cols = append(cols, &qcode.Column{Table: ti.Name, Name: rel.Right.Col, FieldName: rel.Right.Col})
colmap[rel.Right.Col] = struct{}{}
}
case RelOneToManyThrough:
if _, ok := colmap[rel.Left.Col]; !ok {
cols = append(cols, &qcode.Column{Table: ti.Name, Name: rel.Left.Col, FieldName: rel.Left.Col})
colmap[rel.Left.Col] = struct{}{}
}
case RelEmbedded:
if _, ok := colmap[rel.Left.Col]; !ok {
cols = append(cols, &qcode.Column{Table: ti.Name, Name: rel.Left.Col, FieldName: rel.Left.Col})
colmap[rel.Left.Col] = struct{}{}
}
case RelRemote:
if _, ok := colmap[rel.Left.Col]; !ok {
cols = append(cols, &qcode.Column{Table: ti.Name, Name: rel.Left.Col, FieldName: rel.Right.Col})
colmap[rel.Left.Col] = struct{}{}
c.md.skipped |= (1 << uint(id))
}
case RelPolymorphic:
if _, ok := colmap[rel.Left.Col]; !ok {
cols = append(cols, &qcode.Column{Table: ti.Name, Name: rel.Left.Col, FieldName: rel.Left.Col})
colmap[rel.Left.Col] = struct{}{}
}
if _, ok := colmap[rel.Right.Table]; !ok {
cols = append(cols, &qcode.Column{Table: ti.Name, Name: rel.Right.Table, FieldName: rel.Right.Table})
colmap[rel.Right.Table] = struct{}{}
}
default:
return nil, fmt.Errorf("unknown relationship %s", rel)
}
}
return cols, nil
}
// This
// (A, B, C) >= (X, Y, Z)
//
// Becomes
// (A > X)
// OR ((A = X) AND (B > Y))
// OR ((A = X) AND (B = Y) AND (C > Z))
// OR ((A = X) AND (B = Y) AND (C = Z))
func (c *compilerContext) addSeekPredicate(sel *qcode.Select) error {
var or, and *qcode.Exp
obLen := len(sel.OrderBy)
if obLen > 1 {
or = qcode.NewFilter()
or.Op = qcode.OpOr
}
for i := 0; i < obLen; i++ {
if i > 0 {
and = qcode.NewFilter()
and.Op = qcode.OpAnd
}
for n, ob := range sel.OrderBy {
f := qcode.NewFilter()
f.Col = ob.Col
f.Type = qcode.ValRef
f.Table = "__cur"
f.Val = ob.Col
if obLen == 1 {
qcode.AddFilter(sel, f)
return nil
}
switch {
case i > 0 && n != i:
f.Op = qcode.OpEquals
case ob.Order == qcode.OrderDesc:
f.Op = qcode.OpLesserThan
default:
f.Op = qcode.OpGreaterThan
}
if and != nil {
and.Children = append(and.Children, f)
} else {
or.Children = append(or.Children, f)
}
if n == i {
break
}
}
if and != nil {
or.Children = append(or.Children, and)
}
}
qcode.AddFilter(sel, or)
return nil
}
func (c *compilerContext) renderSelect(sel *qcode.Select, ti *DBTableInfo, vars Variables) error {
var rel *DBRel
var err error
// Relationships must be between union parents and their parents
if sel.ParentID != -1 {
if sel.Type == qcode.STMember && sel.UParentID != -1 {
cn := c.s[sel.ParentID].Name
pn := c.s[sel.UParentID].Name
rel, err = c.schema.GetRel(cn, pn)
} else {
pn := c.s[sel.ParentID].Name
rel, err = c.schema.GetRel(ti.Name, pn)
}
}
if err != nil {
return err
}
childCols, err := c.initSelect(sel, ti, vars)
if err != nil {
return err
}
// SELECT
// io.WriteString(c.w, `SELECT jsonb_build_object(`)
// if err := c.renderColumns(sel, ti, skipped); err != nil {
// return 0, err
// }
io.WriteString(c.w, `SELECT to_jsonb("__sr_`)
int32String(c.w, sel.ID)
io.WriteString(c.w, `".*) `)
if sel.Paging.Type != qcode.PtOffset {
for i := range sel.OrderBy {
io.WriteString(c.w, `- '__cur_`)
int32String(c.w, int32(i))
io.WriteString(c.w, `' `)
}
}
io.WriteString(c.w, `AS "json"`)
if sel.Paging.Type != qcode.PtOffset {
for i := range sel.OrderBy {
io.WriteString(c.w, `, "__cur_`)
int32String(c.w, int32(i))
io.WriteString(c.w, `"`)
}
}
io.WriteString(c.w, `FROM (SELECT `)
if err := c.renderColumns(sel, ti); err != nil {
return err
}
if sel.Paging.Type != qcode.PtOffset {
for i, ob := range sel.OrderBy {
io.WriteString(c.w, `, LAST_VALUE(`)
colWithTableID(c.w, ti.Name, sel.ID, ob.Col)
io.WriteString(c.w, `) OVER() AS "__cur_`)
int32String(c.w, int32(i))
io.WriteString(c.w, `"`)
}
}
io.WriteString(c.w, ` FROM (`)
// FROM (SELECT .... )
if err = c.renderBaseSelect(sel, ti, rel, childCols); err != nil {
return err
}
//fmt.Fprintf(w, `) AS "%s_%d"`, c.sel.Name, c.sel.ID)
io.WriteString(c.w, `)`)
aliasWithID(c.w, ti.Name, sel.ID)
// END-FROM
return nil
}
func (c *compilerContext) renderLateralJoin(sel *qcode.Select) error {
io.WriteString(c.w, ` LEFT OUTER JOIN LATERAL (`)
return nil
}
func (c *compilerContext) renderLateralJoinClose(sel *qcode.Select) error {
// io.WriteString(c.w, `) `)
// aliasWithID(c.w, "__sj", sel.ID)
io.WriteString(c.w, ` ON ('true')`)
return nil
}
func (c *compilerContext) renderJoin(sel *qcode.Select, ti *DBTableInfo) error {
parent := &c.s[sel.ParentID]
return c.renderJoinByName(ti.Name, parent.Name, parent.ID)
}
func (c *compilerContext) renderJoinByName(table, parent string, id int32) error {
rel, _ := c.schema.GetRel(table, parent)
// This join is only required for one-to-many relations since
// these make use of join tables that need to be pulled in.
if rel == nil || rel.Type != RelOneToManyThrough {
return nil
}
// pt, err := c.schema.GetTable(parent)
// if err != nil {
// return err
// }
//fmt.Fprintf(w, ` LEFT OUTER JOIN "%s" ON (("%s"."%s") = ("%s_%d"."%s"))`,
//rel.Through, rel.Through, rel.ColT, c.parent.Name, c.parent.ID, rel.Left.Col)
io.WriteString(c.w, ` LEFT OUTER JOIN "`)
io.WriteString(c.w, rel.Through.Table)
io.WriteString(c.w, `" ON ((`)
colWithTable(c.w, rel.Through.Table, rel.Through.ColL)
io.WriteString(c.w, `) = (`)
colWithTable(c.w, rel.Left.Table, rel.Left.Col)
io.WriteString(c.w, `))`)
return nil
}
func (c *compilerContext) renderColumns(sel *qcode.Select, ti *DBTableInfo) error {
i := 0
var cn string
for _, col := range sel.Cols {
if n := funcPrefixLen(c.schema.fm, col.Name); n != 0 {
if !sel.Functions {
continue
}
cn = col.Name[n:]
} else {
cn = col.Name
if strings.HasSuffix(cn, "_cursor") {
continue
}
}
if len(sel.Allowed) != 0 {
if _, ok := sel.Allowed[cn]; !ok {
continue
}
}
if i != 0 {
io.WriteString(c.w, ", ")
}
colWithTableID(c.w, ti.Name, sel.ID, col.Name)
alias(c.w, col.FieldName)
i++
}
i += c.renderRemoteRelColumns(sel, ti, i)
return c.renderJoinColumns(sel, ti, i)
}
func (c *compilerContext) renderRemoteRelColumns(sel *qcode.Select, ti *DBTableInfo, colsRendered int) int {
i := colsRendered
for _, id := range sel.Children {
child := &c.s[id]
rel, err := c.schema.GetRel(child.Name, sel.Name)
if err != nil || rel.Type != RelRemote {
continue
}
if i != 0 || len(sel.Cols) != 0 {
io.WriteString(c.w, ", ")
}
colWithTableID(c.w, ti.Name, sel.ID, rel.Left.Col)
alias(c.w, rel.Right.Col)
i++
}
return i
}
func (c *compilerContext) renderJoinColumns(sel *qcode.Select, ti *DBTableInfo, colsRendered int) error {
// columns previously rendered
i := colsRendered
for _, id := range sel.Children {
if hasBit(c.md.skipped, uint32(id)) {
continue
}
childSel := &c.s[id]
if i != 0 {
io.WriteString(c.w, ", ")
}
if childSel.SkipRender {
io.WriteString(c.w, `NULL`)
alias(c.w, childSel.FieldName)
continue
}
if childSel.Type == qcode.STUnion {
rel, err := c.schema.GetRel(childSel.Name, ti.Name)
if err != nil {
return err
}
io.WriteString(c.w, `(CASE `)
for _, uid := range childSel.Children {
unionSel := &c.s[uid]
io.WriteString(c.w, `WHEN `)
colWithTableID(c.w, ti.Name, sel.ID, rel.Right.Table)
io.WriteString(c.w, ` = `)
squoted(c.w, unionSel.Name)
io.WriteString(c.w, ` THEN `)
io.WriteString(c.w, `"__sj_`)
int32String(c.w, unionSel.ID)
io.WriteString(c.w, `"."json"`)
}
io.WriteString(c.w, `END)`)
alias(c.w, childSel.FieldName)
} else {
io.WriteString(c.w, `"__sj_`)
int32String(c.w, childSel.ID)
io.WriteString(c.w, `"."json"`)
alias(c.w, childSel.FieldName)
}
if childSel.Paging.Type != qcode.PtOffset {
io.WriteString(c.w, `, "__sj_`)
int32String(c.w, childSel.ID)
io.WriteString(c.w, `"."cursor" AS "`)
io.WriteString(c.w, childSel.FieldName)
io.WriteString(c.w, `_cursor"`)
}
i++
}
return nil
}
func (c *compilerContext) renderBaseSelect(sel *qcode.Select, ti *DBTableInfo, rel *DBRel,
childCols []*qcode.Column) error {
isRoot := (rel == nil)
isFil := (sel.Where != nil && sel.Where.Op != qcode.OpNop)
hasOrder := len(sel.OrderBy) != 0
if sel.Paging.Cursor {
c.renderCursorCTE(sel)
}
io.WriteString(c.w, `SELECT `)
if len(sel.DistinctOn) != 0 {
c.renderDistinctOn(sel, ti)
}
realColsRendered, isAgg, err := c.renderBaseColumns(sel, ti, childCols)
if err != nil {
return err
}
io.WriteString(c.w, ` FROM `)
c.renderFrom(sel, ti, rel)
if isRoot && isFil {
io.WriteString(c.w, ` WHERE (`)
if err := c.renderWhere(sel, ti); err != nil {
return err
}
io.WriteString(c.w, `)`)
}
if !isRoot {
if err := c.renderJoin(sel, ti); err != nil {
return err
}
io.WriteString(c.w, ` WHERE (`)
if err := c.renderRelationship(sel, rel); err != nil {
return err
}
if isFil {
io.WriteString(c.w, ` AND `)
if err := c.renderWhere(sel, ti); err != nil {
return err
}
}
io.WriteString(c.w, `)`)
}
if isAgg && len(realColsRendered) != 0 {
io.WriteString(c.w, ` GROUP BY `)
for i, id := range realColsRendered {
c.renderComma(i)
//fmt.Fprintf(w, `"%s"."%s"`, c.sel.Name, c.sel.Cols[id].Name)
colWithTable(c.w, ti.Name, sel.Cols[id].Name)
}
}
if hasOrder {
if err := c.renderOrderBy(sel, ti); err != nil {
return err
}
}
switch {
case ti.IsSingular:
io.WriteString(c.w, ` LIMIT ('1') :: integer`)
case sel.Paging.Limit != "":
//fmt.Fprintf(w, ` LIMIT ('%s') :: integer`, c.sel.Paging.Limit)
io.WriteString(c.w, ` LIMIT ('`)
io.WriteString(c.w, sel.Paging.Limit)
io.WriteString(c.w, `') :: integer`)
case sel.Paging.NoLimit:
break
default:
io.WriteString(c.w, ` LIMIT ('20') :: integer`)
}
if sel.Paging.Offset != "" {
//fmt.Fprintf(w, ` OFFSET ('%s') :: integer`, c.sel.Paging.Offset)
io.WriteString(c.w, ` OFFSET ('`)
io.WriteString(c.w, sel.Paging.Offset)
io.WriteString(c.w, `') :: integer`)
}
return nil
}
func (c *compilerContext) renderFrom(sel *qcode.Select, ti *DBTableInfo, rel *DBRel) error {
if rel != nil && rel.Type == RelEmbedded {
// jsonb_to_recordset('[{"a":1,"b":[1,2,3],"c":"bar"}, {"a":2,"b":[1,2,3],"c":"bar"}]') as x(a int, b text, d text);
io.WriteString(c.w, `"`)
io.WriteString(c.w, rel.Left.Table)
io.WriteString(c.w, `", `)
io.WriteString(c.w, ti.Type)
io.WriteString(c.w, `_to_recordset(`)
colWithTable(c.w, rel.Left.Table, rel.Right.Col)
io.WriteString(c.w, `) AS `)
io.WriteString(c.w, `"`)
io.WriteString(c.w, ti.Name)
io.WriteString(c.w, `"`)
io.WriteString(c.w, `(`)
for i, col := range ti.Columns {
if i != 0 {
io.WriteString(c.w, `, `)
}
io.WriteString(c.w, col.Name)
io.WriteString(c.w, ` `)
io.WriteString(c.w, col.Type)
}
io.WriteString(c.w, `)`)
} else {
//fmt.Fprintf(w, ` FROM "%s"`, c.sel.Name)
io.WriteString(c.w, `"`)
io.WriteString(c.w, ti.Name)
io.WriteString(c.w, `"`)
}
if sel.Paging.Cursor {
io.WriteString(c.w, `, "__cur"`)
}
return nil
}
func (c *compilerContext) renderCursorCTE(sel *qcode.Select) error {
io.WriteString(c.w, `WITH "__cur" AS (SELECT `)
for i, ob := range sel.OrderBy {
if i != 0 {
io.WriteString(c.w, `, `)
}
io.WriteString(c.w, `a[`)
int32String(c.w, int32(i+1))
io.WriteString(c.w, `] as `)
quoted(c.w, ob.Col)
}
io.WriteString(c.w, ` FROM string_to_array(`)
c.md.renderValueExp(c.w, Param{Name: "cursor", Type: "json"})
io.WriteString(c.w, `, ',') as a) `)
return nil
}
func (c *compilerContext) renderRelationshipByName(table, parent string) error {
rel, err := c.schema.GetRel(table, parent)
if err != nil {
return err
}
return c.renderRelationship(nil, rel)
}
func (c *compilerContext) renderRelationship(sel *qcode.Select, rel *DBRel) error {
var pid int32
switch {
case sel == nil:
pid = int32(-1)
case sel.Type == qcode.STMember:
pid = sel.UParentID
default:
pid = sel.ParentID
}
io.WriteString(c.w, `((`)
switch rel.Type {
case RelOneToOne, RelOneToMany:
//fmt.Fprintf(w, `(("%s"."%s") = ("%s_%d"."%s"))`,
//c.sel.Name, rel.Left.Col, c.parent.Name, c.parent.ID, rel.Right.Col)
switch {
case !rel.Left.Array && rel.Right.Array:
colWithTable(c.w, rel.Left.Table, rel.Left.Col)
io.WriteString(c.w, `) = any (`)
colWithTableID(c.w, rel.Right.Table, pid, rel.Right.Col)
case rel.Left.Array && !rel.Right.Array:
colWithTableID(c.w, rel.Right.Table, pid, rel.Right.Col)
io.WriteString(c.w, `) = any (`)
colWithTable(c.w, rel.Left.Table, rel.Left.Col)
default:
colWithTable(c.w, rel.Left.Table, rel.Left.Col)
io.WriteString(c.w, `) = (`)
colWithTableID(c.w, rel.Right.Table, pid, rel.Right.Col)
}
case RelOneToManyThrough:
// This requires the through table to be joined onto this select
//fmt.Fprintf(w, `(("%s"."%s") = ("%s"."%s"))`,
//c.sel.Name, rel.Left.Col, rel.Through, rel.Right.Col)
switch {
case !rel.Left.Array && rel.Right.Array:
colWithTable(c.w, rel.Left.Table, rel.Left.Col)
io.WriteString(c.w, `) = any (`)
colWithTable(c.w, rel.Through.Table, rel.Through.ColR)
case rel.Left.Array && !rel.Right.Array:
colWithTable(c.w, rel.Through.Table, rel.Through.ColR)
io.WriteString(c.w, `) = any (`)
colWithTable(c.w, rel.Left.Table, rel.Left.Col)
default:
colWithTable(c.w, rel.Through.Table, rel.Through.ColR)
io.WriteString(c.w, `) = (`)
colWithTable(c.w, rel.Right.Table, rel.Right.Col)
}
case RelEmbedded:
colWithTable(c.w, rel.Left.Table, rel.Left.Col)
io.WriteString(c.w, `) = (`)
colWithTableID(c.w, rel.Left.Table, pid, rel.Left.Col)
case RelPolymorphic:
colWithTable(c.w, sel.Name, rel.Right.Col)
io.WriteString(c.w, `) = (`)
colWithTableID(c.w, rel.Left.Table, pid, rel.Left.Col)
io.WriteString(c.w, `) AND (`)
colWithTableID(c.w, rel.Left.Table, pid, rel.Right.Table)
io.WriteString(c.w, `) = (`)
squoted(c.w, sel.Name)
}
io.WriteString(c.w, `))`)
return nil
}
func (c *compilerContext) renderWhere(sel *qcode.Select, ti *DBTableInfo) error {
if sel.Where != nil {
return c.renderExp(sel.Where, ti, false)
}
return nil
}
func (c *compilerContext) renderExp(ex *qcode.Exp, ti *DBTableInfo, skipNested bool) error {
st := util.NewStack()
st.Push(ex)
for {
if st.Len() == 0 {
break
}
intf := st.Pop()
switch val := intf.(type) {
case int32:
switch val {
case '(':
io.WriteString(c.w, `(`)
case ')':
io.WriteString(c.w, `)`)
}
case qcode.ExpOp:
switch val {
case qcode.OpAnd:
io.WriteString(c.w, ` AND `)
case qcode.OpOr:
io.WriteString(c.w, ` OR `)
case qcode.OpNot:
io.WriteString(c.w, `NOT `)
case qcode.OpFalse:
io.WriteString(c.w, `false`)
default:
return fmt.Errorf("11: unexpected value %v (%t)", intf, intf)
}
case *qcode.Exp:
switch val.Op {
case qcode.OpFalse:
st.Push(val.Op)
case qcode.OpAnd, qcode.OpOr:
st.Push(')')
for i := len(val.Children) - 1; i >= 0; i-- {
st.Push(val.Children[i])
if i > 0 {
st.Push(val.Op)
}
}
st.Push('(')
case qcode.OpNot:
st.Push(val.Children[0])
st.Push(qcode.OpNot)
default:
if !skipNested && len(val.NestedCols) != 0 {
io.WriteString(c.w, `EXISTS `)
if err := c.renderNestedWhere(val, ti); err != nil {
return err
}
} else if err := c.renderOp(val, ti); err != nil {
return err
}
}
//qcode.FreeExp(val)
default:
return fmt.Errorf("12: unexpected value %v (%t)", intf, intf)
}
}
return nil
}
func (c *compilerContext) renderNestedWhere(ex *qcode.Exp, ti *DBTableInfo) error {
for i := 0; i < len(ex.NestedCols)-1; i++ {
cti, err := c.schema.GetTable(ex.NestedCols[i])
if err != nil {
return err
}
if i != 0 {
io.WriteString(c.w, ` AND `)
}
io.WriteString(c.w, `(SELECT 1 FROM `)
io.WriteString(c.w, cti.Name)
if err := c.renderJoinByName(cti.Name, ti.Name, -1); err != nil {
return err
}
io.WriteString(c.w, ` WHERE `)
if err := c.renderRelationshipByName(cti.Name, ti.Name); err != nil {
return err
}
io.WriteString(c.w, ` AND (`)
if err := c.renderExp(ex, cti, true); err != nil {
return err
}
io.WriteString(c.w, `)`)
}
for i := 0; i < len(ex.NestedCols)-1; i++ {
io.WriteString(c.w, `)`)
}
return nil
}
func (c *compilerContext) renderOp(ex *qcode.Exp, ti *DBTableInfo) error {
var col *DBColumn
var ok bool
if ex.Op == qcode.OpNop {
return nil
}
if ex.Col != "" {
if col, ok = ti.ColMap[ex.Col]; !ok {
return fmt.Errorf("no column '%s' found ", ex.Col)
}
io.WriteString(c.w, `((`)
colWithTable(c.w, ti.Name, ex.Col)
io.WriteString(c.w, `) `)
}
switch ex.Op {
case qcode.OpEquals:
io.WriteString(c.w, `=`)
case qcode.OpNotEquals:
io.WriteString(c.w, `!=`)
case qcode.OpNotDistinct:
io.WriteString(c.w, `IS NOT DISTINCT FROM`)
case qcode.OpDistinct:
io.WriteString(c.w, `IS DISTINCT FROM`)
case qcode.OpGreaterOrEquals:
io.WriteString(c.w, `>=`)
case qcode.OpLesserOrEquals:
io.WriteString(c.w, `<=`)
case qcode.OpGreaterThan:
io.WriteString(c.w, `>`)
case qcode.OpLesserThan:
io.WriteString(c.w, `<`)
case qcode.OpIn:
io.WriteString(c.w, `= ANY`)
case qcode.OpNotIn:
io.WriteString(c.w, `!= ANY`)
case qcode.OpLike:
io.WriteString(c.w, `LIKE`)
case qcode.OpNotLike:
io.WriteString(c.w, `NOT LIKE`)
case qcode.OpILike:
io.WriteString(c.w, `ILIKE`)
case qcode.OpNotILike:
io.WriteString(c.w, `NOT ILIKE`)
case qcode.OpSimilar:
io.WriteString(c.w, `SIMILAR TO`)
case qcode.OpNotSimilar:
io.WriteString(c.w, `NOT SIMILAR TO`)
case qcode.OpContains:
io.WriteString(c.w, `@>`)
case qcode.OpContainedIn:
io.WriteString(c.w, `<@`)
case qcode.OpHasKey:
io.WriteString(c.w, `?`)
case qcode.OpHasKeyAny:
io.WriteString(c.w, `?|`)
case qcode.OpHasKeyAll:
io.WriteString(c.w, `?&`)
case qcode.OpIsNull:
if strings.EqualFold(ex.Val, "true") {
io.WriteString(c.w, `IS NULL)`)
} else {
io.WriteString(c.w, `IS NOT NULL)`)
}
return nil
case qcode.OpEqID:
if ti.PrimaryCol == nil {
return fmt.Errorf("no primary key column defined for %s", ti.Name)
}
col = ti.PrimaryCol
//fmt.Fprintf(w, `(("%s") =`, c.ti.PrimaryCol)
io.WriteString(c.w, `((`)
colWithTable(c.w, ti.Name, ti.PrimaryCol.Name)
//io.WriteString(c.w, ti.PrimaryCol)
io.WriteString(c.w, `) =`)
case qcode.OpTsQuery:
if ti.PrimaryCol == nil {
return fmt.Errorf("no tsv column defined for %s", ti.Name)
}
//fmt.Fprintf(w, `(("%s") @@ websearch_to_tsquery('%s'))`, c.ti.TSVCol, val.Val)
io.WriteString(c.w, `((`)
colWithTable(c.w, ti.Name, ti.TSVCol.Name)
if c.schema.ver >= 110000 {
io.WriteString(c.w, `) @@ websearch_to_tsquery(`)
} else {
io.WriteString(c.w, `) @@ to_tsquery(`)
}
c.md.renderValueExp(c.w, Param{Name: ex.Val, Type: "string"})
io.WriteString(c.w, `))`)
return nil
default:
return fmt.Errorf("[Where] unexpected op code %d", ex.Op)
}
switch {
case ex.Type == qcode.ValList:
c.renderList(ex)
case col == nil:
return errors.New("no column found for expression value")
default:
c.renderVal(ex, c.vars, col)
}
io.WriteString(c.w, `)`)
return nil
}
func (c *compilerContext) renderOrderBy(sel *qcode.Select, ti *DBTableInfo) error {
io.WriteString(c.w, ` ORDER BY `)
for i := range sel.OrderBy {
if i != 0 {
io.WriteString(c.w, `, `)
}
ob := sel.OrderBy[i]
colWithTable(c.w, ti.Name, ob.Col)
switch ob.Order {
case qcode.OrderAsc:
io.WriteString(c.w, ` ASC`)
case qcode.OrderDesc:
io.WriteString(c.w, ` DESC`)
case qcode.OrderAscNullsFirst:
io.WriteString(c.w, ` ASC NULLS FIRST`)
case qcode.OrderDescNullsFirst:
io.WriteString(c.w, ` DESC NULLLS FIRST`)
case qcode.OrderAscNullsLast:
io.WriteString(c.w, ` ASC NULLS LAST`)
case qcode.OrderDescNullsLast:
io.WriteString(c.w, ` DESC NULLS LAST`)
default:
return fmt.Errorf("13: unexpected value %v", ob.Order)
}
}
return nil
}
func (c *compilerContext) renderDistinctOn(sel *qcode.Select, ti *DBTableInfo) {
io.WriteString(c.w, `DISTINCT ON (`)
for i := range sel.DistinctOn {
if i != 0 {
io.WriteString(c.w, `, `)
}
colWithTable(c.w, ti.Name, sel.DistinctOn[i])
}
io.WriteString(c.w, `) `)
}
func (c *compilerContext) renderList(ex *qcode.Exp) {
io.WriteString(c.w, ` (`)
for i := range ex.ListVal {
if i != 0 {
io.WriteString(c.w, `, `)
}
switch ex.ListType {
case qcode.ValBool, qcode.ValInt, qcode.ValFloat:
io.WriteString(c.w, ex.ListVal[i])
case qcode.ValStr:
io.WriteString(c.w, `'`)
io.WriteString(c.w, ex.ListVal[i])
io.WriteString(c.w, `'`)
}
}
io.WriteString(c.w, `)`)
}
func (c *compilerContext) renderVal(ex *qcode.Exp, vars map[string]string, col *DBColumn) {
io.WriteString(c.w, ` `)
switch ex.Type {
case qcode.ValVar:
val, ok := vars[ex.Val]
switch {
case ok && strings.HasPrefix(val, "sql:"):
io.WriteString(c.w, `(`)
c.md.RenderVar(c.w, val[4:])
io.WriteString(c.w, `)`)
case ok:
squoted(c.w, val)
case ex.Op == qcode.OpIn || ex.Op == qcode.OpNotIn:
io.WriteString(c.w, `(ARRAY(SELECT json_array_elements_text(`)
c.md.renderValueExp(c.w, Param{Name: ex.Val, Type: col.Type, IsArray: true})
io.WriteString(c.w, `))`)
io.WriteString(c.w, ` :: `)
io.WriteString(c.w, col.Type)
io.WriteString(c.w, `[])`)
return
default:
c.md.renderValueExp(c.w, Param{Name: ex.Val, Type: col.Type, IsArray: false})
}
case qcode.ValRef:
colWithTable(c.w, ex.Table, ex.Col)
default:
squoted(c.w, ex.Val)
}
io.WriteString(c.w, ` :: `)
io.WriteString(c.w, col.Type)
}
func funcPrefixLen(fm map[string]*DBFunction, fn string) int {
switch {
case strings.HasPrefix(fn, "avg_"):
return 4
case strings.HasPrefix(fn, "count_"):
return 6
case strings.HasPrefix(fn, "max_"):
return 4
case strings.HasPrefix(fn, "min_"):
return 4
case strings.HasPrefix(fn, "sum_"):
return 4
case strings.HasPrefix(fn, "stddev_"):
return 7
case strings.HasPrefix(fn, "stddev_pop_"):
return 11
case strings.HasPrefix(fn, "stddev_samp_"):
return 12
case strings.HasPrefix(fn, "variance_"):
return 9
case strings.HasPrefix(fn, "var_pop_"):
return 8
case strings.HasPrefix(fn, "var_samp_"):
return 9
}
fnLen := len(fn)
for k := range fm {
kLen := len(k)
if kLen < fnLen && k[0] == fn[0] && strings.HasPrefix(fn, k) && fn[kLen] == '_' {
return kLen + 1
}
}
return 0
}
func hasBit(n, pos uint32) bool {
val := n & (1 << pos)
return (val > 0)
}
func alias(w io.Writer, alias string) {
io.WriteString(w, ` AS "`)
io.WriteString(w, alias)
io.WriteString(w, `"`)
}
func aliasWithID(w io.Writer, alias string, id int32) {
io.WriteString(w, ` AS "`)
io.WriteString(w, alias)
io.WriteString(w, `_`)
int32String(w, id)
io.WriteString(w, `"`)
}
func colWithTable(w io.Writer, table, col string) {
io.WriteString(w, `"`)
io.WriteString(w, table)
io.WriteString(w, `"."`)
io.WriteString(w, col)
io.WriteString(w, `"`)
}
func colWithTableID(w io.Writer, table string, id int32, col string) {
io.WriteString(w, `"`)
io.WriteString(w, table)
if id >= 0 {
io.WriteString(w, `_`)
int32String(w, id)
}
io.WriteString(w, `"."`)
io.WriteString(w, col)
io.WriteString(w, `"`)
}
func quoted(w io.Writer, identifier string) {
io.WriteString(w, `"`)
io.WriteString(w, identifier)
io.WriteString(w, `"`)
}
func squoted(w io.Writer, identifier string) {
io.WriteString(w, `'`)
io.WriteString(w, identifier)
io.WriteString(w, `'`)
}
func int32String(w io.Writer, val int32) {
io.WriteString(w, strconv.FormatInt(int64(val), 10))
}

View File

@ -1,572 +0,0 @@
package psql_test
import (
"bytes"
"encoding/json"
"testing"
)
func withComplexArgs(t *testing.T) {
gql := `query {
proDUcts(
# returns only 30 items
limit: 30,
# starts from item 10, commented out for now
# offset: 10,
# orders the response items by highest price
order_by: { price: desc },
# no duplicate prices returned
distinct: [ price ]
# only items with an id >= 20 and < 28 are returned
where: { id: { and: { greater_or_equals: 20, lt: 28 } } }) {
id
NAME
price
}
}`
compileGQLToPSQL(t, gql, nil, "user")
}
func withWhereIn(t *testing.T) {
gql := `query {
products(where: { id: { in: $list } }) {
id
}
}`
vars := map[string]json.RawMessage{
"list": json.RawMessage(`[1,2,3]`),
}
compileGQLToPSQL(t, gql, vars, "user")
}
func withWhereAndList(t *testing.T) {
gql := `query {
products(
where: {
and: [
{ not: { id: { is_null: true } } },
{ price: { gt: 10 } },
] } ) {
id
name
price
}
}`
compileGQLToPSQL(t, gql, nil, "user")
}
func withWhereIsNull(t *testing.T) {
gql := `query {
products(
where: {
and: {
not: { id: { is_null: true } },
price: { gt: 10 }
}}) {
id
name
price
}
}`
compileGQLToPSQL(t, gql, nil, "user")
}
func withWhereMultiOr(t *testing.T) {
gql := `query {
products(
where: {
or: {
not: { id: { is_null: true } },
price: { gt: 10 },
price: { lt: 20 }
} }
) {
id
name
price
}
}`
compileGQLToPSQL(t, gql, nil, "user")
}
func fetchByID(t *testing.T) {
gql := `query {
product(id: $id) {
id
name
}
}`
compileGQLToPSQL(t, gql, nil, "user")
}
func searchQuery(t *testing.T) {
gql := `query {
products(search: $query) {
id
name
search_rank
search_headline_description
}
}`
compileGQLToPSQL(t, gql, nil, "admin")
}
func oneToMany(t *testing.T) {
gql := `query {
users {
email
products {
name
price
}
}
}`
compileGQLToPSQL(t, gql, nil, "user")
}
func oneToManyReverse(t *testing.T) {
gql := `query {
products {
name
price
users {
email
}
}
}`
compileGQLToPSQL(t, gql, nil, "user")
}
func oneToManyArray(t *testing.T) {
gql := `
query {
product {
name
price
tags {
id
name
}
}
tags {
name
product {
name
}
}
}`
compileGQLToPSQL(t, gql, nil, "admin")
}
func manyToMany(t *testing.T) {
gql := `query {
products {
name
customers {
email
full_name
}
}
}`
compileGQLToPSQL(t, gql, nil, "user")
}
func manyToManyReverse(t *testing.T) {
gql := `query {
customers {
email
full_name
products {
name
}
}
}`
compileGQLToPSQL(t, gql, nil, "user")
}
func aggFunction(t *testing.T) {
gql := `query {
products {
name
count_price
}
}`
compileGQLToPSQL(t, gql, nil, "user")
}
func aggFunctionBlockedByCol(t *testing.T) {
gql := `query {
products {
name
count_price
}
}`
compileGQLToPSQL(t, gql, nil, "anon")
}
func aggFunctionDisabled(t *testing.T) {
gql := `query {
products {
name
count_price
}
}`
compileGQLToPSQL(t, gql, nil, "anon1")
}
func aggFunctionWithFilter(t *testing.T) {
gql := `query {
products(where: { id: { gt: 10 } }) {
id
max_price
}
}`
compileGQLToPSQL(t, gql, nil, "user")
}
func syntheticTables(t *testing.T) {
gql := `query {
me {
email
}
}`
compileGQLToPSQL(t, gql, nil, "user")
}
func queryWithVariables(t *testing.T) {
gql := `query {
product(id: $PRODUCT_ID, where: { price: { eq: $PRODUCT_PRICE } }) {
id
name
}
}`
compileGQLToPSQL(t, gql, nil, "user")
}
func withWhereOnRelations(t *testing.T) {
gql := `query {
users(where: {
not: {
products: {
price: { gt: 3 }
}
}
}) {
id
email
}
}`
compileGQLToPSQL(t, gql, nil, "user")
}
func multiRoot(t *testing.T) {
gql := `query {
product {
id
name
customer {
email
}
customers {
email
}
}
user {
id
email
}
customer {
id
}
}`
compileGQLToPSQL(t, gql, nil, "user")
}
func withFragment1(t *testing.T) {
gql := `
fragment userFields1 on user {
id
email
}
query {
users {
...userFields2
created_at
...userFields1
}
}
fragment userFields2 on user {
first_name
last_name
}`
compileGQLToPSQL(t, gql, nil, "anon")
}
func withFragment2(t *testing.T) {
gql := `
query {
users {
...userFields2
created_at
...userFields1
}
}
fragment userFields1 on user {
id
email
}
fragment userFields2 on user {
first_name
last_name
}`
compileGQLToPSQL(t, gql, nil, "anon")
}
func withFragment3(t *testing.T) {
gql := `
fragment userFields1 on user {
id
email
}
fragment userFields2 on user {
first_name
last_name
}
query {
users {
...userFields2
created_at
...userFields1
}
}
`
compileGQLToPSQL(t, gql, nil, "anon")
}
// func withInlineFragment(t *testing.T) {
// gql := `
// query {
// users {
// ... on users {
// id
// email
// }
// created_at
// ... on user {
// first_name
// last_name
// }
// }
// }
// `
// compileGQLToPSQL(t, gql, nil, "anon")
// }
func withCursor(t *testing.T) {
gql := `query {
Products(
first: 20
after: $cursor
order_by: { price: desc }) {
Name
}
}`
vars := map[string]json.RawMessage{
"cursor": json.RawMessage(`"0,1"`),
}
compileGQLToPSQL(t, gql, vars, "admin")
}
func jsonColumnAsTable(t *testing.T) {
gql := `query {
products {
id
name
tag_count {
count
tags {
name
}
}
}
}`
compileGQLToPSQL(t, gql, nil, "admin")
}
func nullForAuthRequiredInAnon(t *testing.T) {
gql := `query {
products {
id
name
user(where: { id: { eq: $user_id } }) {
id
email
}
}
}`
compileGQLToPSQL(t, gql, nil, "anon")
}
func blockedQuery(t *testing.T) {
gql := `query {
user(id: $id, where: { id: { gt: 3 } }) {
id
full_name
email
}
}`
compileGQLToPSQL(t, gql, nil, "bad_dude")
}
func blockedFunctions(t *testing.T) {
gql := `query {
users {
count_id
email
}
}`
compileGQLToPSQL(t, gql, nil, "bad_dude")
}
func TestCompileQuery(t *testing.T) {
t.Run("withComplexArgs", withComplexArgs)
t.Run("withWhereIn", withWhereIn)
t.Run("withWhereAndList", withWhereAndList)
t.Run("withWhereIsNull", withWhereIsNull)
t.Run("withWhereMultiOr", withWhereMultiOr)
t.Run("fetchByID", fetchByID)
t.Run("searchQuery", searchQuery)
t.Run("oneToMany", oneToMany)
t.Run("oneToManyReverse", oneToManyReverse)
t.Run("oneToManyArray", oneToManyArray)
t.Run("manyToMany", manyToMany)
t.Run("manyToManyReverse", manyToManyReverse)
t.Run("aggFunction", aggFunction)
t.Run("aggFunctionBlockedByCol", aggFunctionBlockedByCol)
t.Run("aggFunctionDisabled", aggFunctionDisabled)
t.Run("aggFunctionWithFilter", aggFunctionWithFilter)
t.Run("syntheticTables", syntheticTables)
t.Run("queryWithVariables", queryWithVariables)
t.Run("withWhereOnRelations", withWhereOnRelations)
t.Run("multiRoot", multiRoot)
t.Run("withFragment1", withFragment1)
t.Run("withFragment2", withFragment2)
t.Run("withFragment3", withFragment3)
//t.Run("withInlineFragment", withInlineFragment)
t.Run("jsonColumnAsTable", jsonColumnAsTable)
t.Run("withCursor", withCursor)
t.Run("nullForAuthRequiredInAnon", nullForAuthRequiredInAnon)
t.Run("blockedQuery", blockedQuery)
t.Run("blockedFunctions", blockedFunctions)
}
var benchGQL = []byte(`query {
proDUcts(
# returns only 30 items
limit: 30,
# starts from item 10, commented out for now
# offset: 10,
# orders the response items by highest price
order_by: { price: desc },
# only items with an id >= 30 and < 30 are returned
where: { id: { and: { greater_or_equals: 20, lt: 28 } } }) {
id
NAME
price
user {
full_name
picture : avatar
}
}
}`)
func BenchmarkCompile(b *testing.B) {
w := &bytes.Buffer{}
b.ResetTimer()
b.ReportAllocs()
for n := 0; n < b.N; n++ {
w.Reset()
qc, err := qcompile.Compile(benchGQL, "user")
if err != nil {
b.Fatal(err)
}
_, err = pcompile.Compile(w, qc, nil)
if err != nil {
b.Fatal(err)
}
}
}
func BenchmarkCompileParallel(b *testing.B) {
b.ReportAllocs()
b.RunParallel(func(pb *testing.PB) {
w := &bytes.Buffer{}
for pb.Next() {
w.Reset()
qc, err := qcompile.Compile(benchGQL, "user")
if err != nil {
b.Fatal(err)
}
_, err = pcompile.Compile(w, qc, nil)
if err != nil {
b.Fatal(err)
}
}
})
}

View File

@ -1,163 +0,0 @@
=== RUN TestCompileInsert
=== RUN TestCompileInsert/simpleInsert
WITH "_sg_input" AS (SELECT $1 :: json AS j), "users" AS (INSERT INTO "users" ("full_name", "email") SELECT CAST( i.j ->>'full_name' AS character varying), CAST( i.j ->>'email' AS character varying) FROM "_sg_input" i RETURNING *) SELECT jsonb_build_object('user', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "users_0"."id" AS "id" FROM (SELECT "users"."id" FROM "users" LIMIT ('1') :: integer) AS "users_0") AS "__sr_0") AS "__sj_0"
=== RUN TestCompileInsert/singleInsert
WITH "_sg_input" AS (SELECT $1 :: json AS j), "products" AS (INSERT INTO "products" ("name", "description", "price", "user_id") SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'description' AS text), CAST( i.j ->>'price' AS numeric(7,2)), CAST( i.j ->>'user_id' AS bigint) FROM "_sg_input" i RETURNING *) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name" FROM (SELECT "products"."id", "products"."name" FROM "products" LIMIT ('1') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0"
=== RUN TestCompileInsert/bulkInsert
WITH "_sg_input" AS (SELECT $1 :: json AS j), "products" AS (INSERT INTO "products" ("name", "description") SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'description' AS text) FROM "_sg_input" i RETURNING *) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name" FROM (SELECT "products"."id", "products"."name" FROM "products" LIMIT ('1') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0"
=== RUN TestCompileInsert/simpleInsertWithPresets
WITH "_sg_input" AS (SELECT $1 :: json AS j), "products" AS (INSERT INTO "products" ("name", "price", "created_at", "updated_at", "user_id") SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'price' AS numeric(7,2)), 'now' :: timestamp without time zone, 'now' :: timestamp without time zone, $2 :: bigint FROM "_sg_input" i RETURNING *) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id" FROM (SELECT "products"."id" FROM "products" LIMIT ('1') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0"
=== RUN TestCompileInsert/nestedInsertManyToMany
WITH "_sg_input" AS (SELECT $1 :: json AS j), "products" AS (INSERT INTO "products" ("name", "price") SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'price' AS numeric(7,2)) FROM "_sg_input" i RETURNING *), "customers" AS (INSERT INTO "customers" ("full_name", "email") SELECT CAST( i.j ->>'full_name' AS character varying), CAST( i.j ->>'email' AS character varying) FROM "_sg_input" i RETURNING *), "purchases" AS (INSERT INTO "purchases" ("sale_type", "quantity", "due_date", "customer_id", "product_id") SELECT CAST( i.j ->>'sale_type' AS character varying), CAST( i.j ->>'quantity' AS integer), CAST( i.j ->>'due_date' AS timestamp without time zone), "customers"."id", "products"."id" FROM "_sg_input" i, "customers", "products" RETURNING *) SELECT jsonb_build_object('purchase', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "purchases_0"."sale_type" AS "sale_type", "purchases_0"."quantity" AS "quantity", "purchases_0"."due_date" AS "due_date", "__sj_1"."json" AS "product", "__sj_2"."json" AS "customer" FROM (SELECT "purchases"."sale_type", "purchases"."quantity", "purchases"."due_date", "purchases"."product_id", "purchases"."customer_id" FROM "purchases" LIMIT ('1') :: integer) AS "purchases_0" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_2".*) AS "json"FROM (SELECT "customers_2"."id" AS "id", "customers_2"."full_name" AS "full_name", "customers_2"."email" AS "email" FROM (SELECT "customers"."id", "customers"."full_name", "customers"."email" FROM "customers" WHERE ((("customers"."id") = ("purchases_0"."customer_id"))) LIMIT ('1') :: integer) AS "customers_2") AS "__sr_2") AS "__sj_2" ON ('true') LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "products_1"."id" AS "id", "products_1"."name" AS "name", "products_1"."price" AS "price" FROM (SELECT "products"."id", "products"."name", "products"."price" FROM "products" WHERE ((("products"."id") = ("purchases_0"."product_id"))) LIMIT ('1') :: integer) AS "products_1") AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0"
WITH "_sg_input" AS (SELECT $1 :: json AS j), "customers" AS (INSERT INTO "customers" ("full_name", "email") SELECT CAST( i.j ->>'full_name' AS character varying), CAST( i.j ->>'email' AS character varying) FROM "_sg_input" i RETURNING *), "products" AS (INSERT INTO "products" ("name", "price") SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'price' AS numeric(7,2)) FROM "_sg_input" i RETURNING *), "purchases" AS (INSERT INTO "purchases" ("sale_type", "quantity", "due_date", "product_id", "customer_id") SELECT CAST( i.j ->>'sale_type' AS character varying), CAST( i.j ->>'quantity' AS integer), CAST( i.j ->>'due_date' AS timestamp without time zone), "products"."id", "customers"."id" FROM "_sg_input" i, "products", "customers" RETURNING *) SELECT jsonb_build_object('purchase', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "purchases_0"."sale_type" AS "sale_type", "purchases_0"."quantity" AS "quantity", "purchases_0"."due_date" AS "due_date", "__sj_1"."json" AS "product", "__sj_2"."json" AS "customer" FROM (SELECT "purchases"."sale_type", "purchases"."quantity", "purchases"."due_date", "purchases"."product_id", "purchases"."customer_id" FROM "purchases" LIMIT ('1') :: integer) AS "purchases_0" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_2".*) AS "json"FROM (SELECT "customers_2"."id" AS "id", "customers_2"."full_name" AS "full_name", "customers_2"."email" AS "email" FROM (SELECT "customers"."id", "customers"."full_name", "customers"."email" FROM "customers" WHERE ((("customers"."id") = ("purchases_0"."customer_id"))) LIMIT ('1') :: integer) AS "customers_2") AS "__sr_2") AS "__sj_2" ON ('true') LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "products_1"."id" AS "id", "products_1"."name" AS "name", "products_1"."price" AS "price" FROM (SELECT "products"."id", "products"."name", "products"."price" FROM "products" WHERE ((("products"."id") = ("purchases_0"."product_id"))) LIMIT ('1') :: integer) AS "products_1") AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0"
=== RUN TestCompileInsert/nestedInsertOneToMany
WITH "_sg_input" AS (SELECT $1 :: json AS j), "users" AS (INSERT INTO "users" ("full_name", "email", "created_at", "updated_at") SELECT CAST( i.j ->>'full_name' AS character varying), CAST( i.j ->>'email' AS character varying), CAST( i.j ->>'created_at' AS timestamp without time zone), CAST( i.j ->>'updated_at' AS timestamp without time zone) FROM "_sg_input" i RETURNING *), "products" AS (INSERT INTO "products" ("name", "price", "created_at", "updated_at", "user_id") SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'price' AS numeric(7,2)), CAST( i.j ->>'created_at' AS timestamp without time zone), CAST( i.j ->>'updated_at' AS timestamp without time zone), "users"."id" FROM "_sg_input" i, "users" RETURNING *) SELECT jsonb_build_object('user', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "users_0"."id" AS "id", "users_0"."full_name" AS "full_name", "users_0"."email" AS "email", "__sj_1"."json" AS "product" FROM (SELECT "users"."id", "users"."full_name", "users"."email" FROM "users" LIMIT ('1') :: integer) AS "users_0" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "products_1"."id" AS "id", "products_1"."name" AS "name", "products_1"."price" AS "price" FROM (SELECT "products"."id", "products"."name", "products"."price" FROM "products" WHERE ((("products"."user_id") = ("users_0"."id"))) LIMIT ('1') :: integer) AS "products_1") AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0"
=== RUN TestCompileInsert/nestedInsertOneToOne
WITH "_sg_input" AS (SELECT $1 :: json AS j), "users" AS (INSERT INTO "users" ("full_name", "email", "created_at", "updated_at") SELECT CAST( i.j ->>'full_name' AS character varying), CAST( i.j ->>'email' AS character varying), CAST( i.j ->>'created_at' AS timestamp without time zone), CAST( i.j ->>'updated_at' AS timestamp without time zone) FROM "_sg_input" i RETURNING *), "products" AS (INSERT INTO "products" ("name", "price", "created_at", "updated_at", "user_id") SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'price' AS numeric(7,2)), CAST( i.j ->>'created_at' AS timestamp without time zone), CAST( i.j ->>'updated_at' AS timestamp without time zone), "users"."id" FROM "_sg_input" i, "users" RETURNING *) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name", "__sj_1"."json" AS "user" FROM (SELECT "products"."id", "products"."name", "products"."user_id" FROM "products" LIMIT ('1') :: integer) AS "products_0" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "users_1"."id" AS "id", "users_1"."full_name" AS "full_name", "users_1"."email" AS "email" FROM (SELECT "users"."id", "users"."full_name", "users"."email" FROM "users" WHERE ((("users"."id") = ("products_0"."user_id"))) LIMIT ('1') :: integer) AS "users_1") AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0"
=== RUN TestCompileInsert/nestedInsertOneToManyWithConnect
WITH "_sg_input" AS (SELECT $1 :: json AS j), "users" AS (INSERT INTO "users" ("full_name", "email", "created_at", "updated_at") SELECT CAST( i.j ->>'full_name' AS character varying), CAST( i.j ->>'email' AS character varying), CAST( i.j ->>'created_at' AS timestamp without time zone), CAST( i.j ->>'updated_at' AS timestamp without time zone) FROM "_sg_input" i RETURNING *), "products" AS ( UPDATE "products" SET "user_id" = "users"."id" FROM "users" WHERE ("products"."id"= ((i.j->'product'->'connect'->>'id'))::bigint) RETURNING "products".*) SELECT jsonb_build_object('user', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "users_0"."id" AS "id", "users_0"."full_name" AS "full_name", "users_0"."email" AS "email", "__sj_1"."json" AS "product" FROM (SELECT "users"."id", "users"."full_name", "users"."email" FROM "users" LIMIT ('1') :: integer) AS "users_0" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "products_1"."id" AS "id", "products_1"."name" AS "name", "products_1"."price" AS "price" FROM (SELECT "products"."id", "products"."name", "products"."price" FROM "products" WHERE ((("products"."user_id") = ("users_0"."id"))) LIMIT ('1') :: integer) AS "products_1") AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0"
=== RUN TestCompileInsert/nestedInsertOneToOneWithConnect
WITH "_sg_input" AS (SELECT $1 :: json AS j), "_x_users" AS (SELECT "id" FROM "_sg_input" i,"users" WHERE "users"."id"= ((i.j->'user'->'connect'->>'id'))::bigint LIMIT 1), "products" AS (INSERT INTO "products" ("name", "price", "created_at", "updated_at", "user_id") SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'price' AS numeric(7,2)), CAST( i.j ->>'created_at' AS timestamp without time zone), CAST( i.j ->>'updated_at' AS timestamp without time zone), "_x_users"."id" FROM "_sg_input" i, "_x_users" RETURNING *) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name", "__sj_1"."json" AS "user", "__sj_2"."json" AS "tags" FROM (SELECT "products"."id", "products"."name", "products"."user_id", "products"."tags" FROM "products" LIMIT ('1') :: integer) AS "products_0" LEFT OUTER JOIN LATERAL (SELECT coalesce(jsonb_agg("__sj_2"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_2".*) AS "json"FROM (SELECT "tags_2"."id" AS "id", "tags_2"."name" AS "name" FROM (SELECT "tags"."id", "tags"."name" FROM "tags" WHERE ((("tags"."slug") = any ("products_0"."tags"))) LIMIT ('20') :: integer) AS "tags_2") AS "__sr_2") AS "__sj_2") AS "__sj_2" ON ('true') LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "users_1"."id" AS "id", "users_1"."full_name" AS "full_name", "users_1"."email" AS "email" FROM (SELECT "users"."id", "users"."full_name", "users"."email" FROM "users" WHERE ((("users"."id") = ("products_0"."user_id"))) LIMIT ('1') :: integer) AS "users_1") AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0"
=== RUN TestCompileInsert/nestedInsertOneToOneWithConnectArray
WITH "_sg_input" AS (SELECT $1 :: json AS j), "_x_users" AS (SELECT "id" FROM "_sg_input" i,"users" WHERE "users"."id" = ANY((select a::bigint AS list from json_array_elements_text((i.j->'user'->'connect'->>'id')::json) AS a)) LIMIT 1), "products" AS (INSERT INTO "products" ("name", "price", "created_at", "updated_at", "user_id") SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'price' AS numeric(7,2)), CAST( i.j ->>'created_at' AS timestamp without time zone), CAST( i.j ->>'updated_at' AS timestamp without time zone), "_x_users"."id" FROM "_sg_input" i, "_x_users" RETURNING *) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name", "__sj_1"."json" AS "user" FROM (SELECT "products"."id", "products"."name", "products"."user_id" FROM "products" LIMIT ('1') :: integer) AS "products_0" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "users_1"."id" AS "id", "users_1"."full_name" AS "full_name", "users_1"."email" AS "email" FROM (SELECT "users"."id", "users"."full_name", "users"."email" FROM "users" WHERE ((("users"."id") = ("products_0"."user_id"))) LIMIT ('1') :: integer) AS "users_1") AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0"
--- PASS: TestCompileInsert (0.03s)
--- PASS: TestCompileInsert/simpleInsert (0.00s)
--- PASS: TestCompileInsert/singleInsert (0.00s)
--- PASS: TestCompileInsert/bulkInsert (0.00s)
--- PASS: TestCompileInsert/simpleInsertWithPresets (0.00s)
--- PASS: TestCompileInsert/nestedInsertManyToMany (0.00s)
--- PASS: TestCompileInsert/nestedInsertOneToMany (0.00s)
--- PASS: TestCompileInsert/nestedInsertOneToOne (0.00s)
--- PASS: TestCompileInsert/nestedInsertOneToManyWithConnect (0.00s)
--- PASS: TestCompileInsert/nestedInsertOneToOneWithConnect (0.00s)
--- PASS: TestCompileInsert/nestedInsertOneToOneWithConnectArray (0.00s)
=== RUN TestCompileMutate
=== RUN TestCompileMutate/singleUpsert
WITH "_sg_input" AS (SELECT $1 :: json AS j), "products" AS (INSERT INTO "products" ("name", "description") SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'description' AS text) FROM "_sg_input" i RETURNING *) ON CONFLICT (id) DO UPDATE SET name = EXCLUDED.name, description = EXCLUDED.description RETURNING *) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name" FROM (SELECT "products"."id", "products"."name" FROM "products" LIMIT ('1') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0"
=== RUN TestCompileMutate/singleUpsertWhere
WITH "_sg_input" AS (SELECT $1 :: json AS j), "products" AS (INSERT INTO "products" ("name", "description") SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'description' AS text) FROM "_sg_input" i RETURNING *) ON CONFLICT (id) WHERE (("products"."price") > '3' :: numeric(7,2)) DO UPDATE SET name = EXCLUDED.name, description = EXCLUDED.description RETURNING *) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name" FROM (SELECT "products"."id", "products"."name" FROM "products" LIMIT ('1') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0"
=== RUN TestCompileMutate/bulkUpsert
WITH "_sg_input" AS (SELECT $1 :: json AS j), "products" AS (INSERT INTO "products" ("name", "description") SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'description' AS text) FROM "_sg_input" i RETURNING *) ON CONFLICT (id) DO UPDATE SET name = EXCLUDED.name, description = EXCLUDED.description RETURNING *) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name" FROM (SELECT "products"."id", "products"."name" FROM "products" LIMIT ('1') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0"
=== RUN TestCompileMutate/delete
WITH "products" AS (DELETE FROM "products" WHERE (((("products"."price") > '0' :: numeric(7,2)) AND (("products"."price") < '8' :: numeric(7,2))) AND (("products"."id") = '1' :: bigint)) RETURNING "products".*) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name" FROM (SELECT "products"."id", "products"."name" FROM "products" LIMIT ('1') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0"
--- PASS: TestCompileMutate (0.01s)
--- PASS: TestCompileMutate/singleUpsert (0.00s)
--- PASS: TestCompileMutate/singleUpsertWhere (0.00s)
--- PASS: TestCompileMutate/bulkUpsert (0.00s)
--- PASS: TestCompileMutate/delete (0.00s)
=== RUN TestCompileQuery
=== RUN TestCompileQuery/withComplexArgs
SELECT jsonb_build_object('products', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name", "products_0"."price" AS "price" FROM (SELECT DISTINCT ON ("products"."price") "products"."id", "products"."name", "products"."price" FROM "products" WHERE (((("products"."id") < '28' :: bigint) AND (("products"."id") >= '20' :: bigint) AND ((("products"."price") > '0' :: numeric(7,2)) AND (("products"."price") < '8' :: numeric(7,2))))) ORDER BY "products"."price" DESC LIMIT ('30') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/withWhereIn
SELECT jsonb_build_object('products', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id" FROM (SELECT "products"."id" FROM "products" WHERE ((((("products"."price") > '0' :: numeric(7,2)) AND (("products"."price") < '8' :: numeric(7,2))) AND (("products"."id") = ANY (ARRAY(SELECT json_array_elements_text($1)) :: bigint[])))) LIMIT ('20') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/withWhereAndList
SELECT jsonb_build_object('products', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name", "products_0"."price" AS "price" FROM (SELECT "products"."id", "products"."name", "products"."price" FROM "products" WHERE (((("products"."price") > '10' :: numeric(7,2)) AND NOT (("products"."id") IS NULL) AND ((("products"."price") > '0' :: numeric(7,2)) AND (("products"."price") < '8' :: numeric(7,2))))) LIMIT ('20') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/withWhereIsNull
SELECT jsonb_build_object('products', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name", "products_0"."price" AS "price" FROM (SELECT "products"."id", "products"."name", "products"."price" FROM "products" WHERE (((("products"."price") > '10' :: numeric(7,2)) AND NOT (("products"."id") IS NULL) AND ((("products"."price") > '0' :: numeric(7,2)) AND (("products"."price") < '8' :: numeric(7,2))))) LIMIT ('20') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/withWhereMultiOr
SELECT jsonb_build_object('products', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name", "products_0"."price" AS "price" FROM (SELECT "products"."id", "products"."name", "products"."price" FROM "products" WHERE ((((("products"."price") > '0' :: numeric(7,2)) AND (("products"."price") < '8' :: numeric(7,2))) AND ((("products"."price") < '20' :: numeric(7,2)) OR (("products"."price") > '10' :: numeric(7,2)) OR NOT (("products"."id") IS NULL)))) LIMIT ('20') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/fetchByID
SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name" FROM (SELECT "products"."id", "products"."name" FROM "products" WHERE ((((("products"."price") > '0' :: numeric(7,2)) AND (("products"."price") < '8' :: numeric(7,2))) AND (("products"."id") = $1 :: bigint))) LIMIT ('1') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0"
=== RUN TestCompileQuery/searchQuery
SELECT jsonb_build_object('products', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name", "products_0"."search_rank" AS "search_rank", "products_0"."search_headline_description" AS "search_headline_description" FROM (SELECT "products"."id", "products"."name", ts_rank("products"."tsv", websearch_to_tsquery($1)) AS "search_rank", ts_headline("products"."description", websearch_to_tsquery($1)) AS "search_headline_description" FROM "products" WHERE ((("products"."tsv") @@ websearch_to_tsquery($1))) LIMIT ('20') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/oneToMany
SELECT jsonb_build_object('users', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "users_0"."email" AS "email", "__sj_1"."json" AS "products" FROM (SELECT "users"."email", "users"."id" FROM "users" LIMIT ('20') :: integer) AS "users_0" LEFT OUTER JOIN LATERAL (SELECT coalesce(jsonb_agg("__sj_1"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "products_1"."name" AS "name", "products_1"."price" AS "price" FROM (SELECT "products"."name", "products"."price" FROM "products" WHERE ((("products"."user_id") = ("users_0"."id")) AND ((("products"."price") > '0' :: numeric(7,2)) AND (("products"."price") < '8' :: numeric(7,2)))) LIMIT ('20') :: integer) AS "products_1") AS "__sr_1") AS "__sj_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/oneToManyReverse
SELECT jsonb_build_object('products', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."name" AS "name", "products_0"."price" AS "price", "__sj_1"."json" AS "users" FROM (SELECT "products"."name", "products"."price", "products"."user_id" FROM "products" WHERE (((("products"."price") > '0' :: numeric(7,2)) AND (("products"."price") < '8' :: numeric(7,2)))) LIMIT ('20') :: integer) AS "products_0" LEFT OUTER JOIN LATERAL (SELECT coalesce(jsonb_agg("__sj_1"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "users_1"."email" AS "email" FROM (SELECT "users"."email" FROM "users" WHERE ((("users"."id") = ("products_0"."user_id"))) LIMIT ('20') :: integer) AS "users_1") AS "__sr_1") AS "__sj_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/oneToManyArray
SELECT jsonb_build_object('tags', "__sj_0"."json", 'product', "__sj_2"."json") as "__root" FROM (SELECT to_jsonb("__sr_2".*) AS "json"FROM (SELECT "products_2"."name" AS "name", "products_2"."price" AS "price", "__sj_3"."json" AS "tags" FROM (SELECT "products"."name", "products"."price", "products"."tags" FROM "products" LIMIT ('1') :: integer) AS "products_2" LEFT OUTER JOIN LATERAL (SELECT coalesce(jsonb_agg("__sj_3"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_3".*) AS "json"FROM (SELECT "tags_3"."id" AS "id", "tags_3"."name" AS "name" FROM (SELECT "tags"."id", "tags"."name" FROM "tags" WHERE ((("tags"."slug") = any ("products_2"."tags"))) LIMIT ('20') :: integer) AS "tags_3") AS "__sr_3") AS "__sj_3") AS "__sj_3" ON ('true')) AS "__sr_2") AS "__sj_2", (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "tags_0"."name" AS "name", "__sj_1"."json" AS "product" FROM (SELECT "tags"."name", "tags"."slug" FROM "tags" LIMIT ('20') :: integer) AS "tags_0" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "products_1"."name" AS "name" FROM (SELECT "products"."name" FROM "products" WHERE ((("tags_0"."slug") = any ("products"."tags"))) LIMIT ('1') :: integer) AS "products_1") AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/manyToMany
SELECT jsonb_build_object('products', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."name" AS "name", "__sj_1"."json" AS "customers" FROM (SELECT "products"."name", "products"."id" FROM "products" WHERE (((("products"."price") > '0' :: numeric(7,2)) AND (("products"."price") < '8' :: numeric(7,2)))) LIMIT ('20') :: integer) AS "products_0" LEFT OUTER JOIN LATERAL (SELECT coalesce(jsonb_agg("__sj_1"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "customers_1"."email" AS "email", "customers_1"."full_name" AS "full_name" FROM (SELECT "customers"."email", "customers"."full_name" FROM "customers" LEFT OUTER JOIN "purchases" ON (("purchases"."customer_id") = ("customers"."id")) WHERE ((("purchases"."product_id") = ("products"."id"))) LIMIT ('20') :: integer) AS "customers_1") AS "__sr_1") AS "__sj_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/manyToManyReverse
SELECT jsonb_build_object('customers', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "customers_0"."email" AS "email", "customers_0"."full_name" AS "full_name", "__sj_1"."json" AS "products" FROM (SELECT "customers"."email", "customers"."full_name", "customers"."id" FROM "customers" LIMIT ('20') :: integer) AS "customers_0" LEFT OUTER JOIN LATERAL (SELECT coalesce(jsonb_agg("__sj_1"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "products_1"."name" AS "name" FROM (SELECT "products"."name" FROM "products" LEFT OUTER JOIN "purchases" ON (("purchases"."product_id") = ("products"."id")) WHERE ((("purchases"."customer_id") = ("customers"."id")) AND ((("products"."price") > '0' :: numeric(7,2)) AND (("products"."price") < '8' :: numeric(7,2)))) LIMIT ('20') :: integer) AS "products_1") AS "__sr_1") AS "__sj_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/aggFunction
SELECT jsonb_build_object('products', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."name" AS "name", "products_0"."count_price" AS "count_price" FROM (SELECT "products"."name", count("products"."price") AS "count_price" FROM "products" WHERE (((("products"."price") > '0' :: numeric(7,2)) AND (("products"."price") < '8' :: numeric(7,2)))) GROUP BY "products"."name" LIMIT ('20') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/aggFunctionBlockedByCol
SELECT jsonb_build_object('products', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."name" AS "name" FROM (SELECT "products"."name" FROM "products" GROUP BY "products"."name" LIMIT ('20') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/aggFunctionDisabled
SELECT jsonb_build_object('products', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."name" AS "name" FROM (SELECT "products"."name" FROM "products" GROUP BY "products"."name" LIMIT ('20') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/aggFunctionWithFilter
SELECT jsonb_build_object('products', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."max_price" AS "max_price" FROM (SELECT "products"."id", max("products"."price") AS "max_price" FROM "products" WHERE ((((("products"."price") > '0' :: numeric(7,2)) AND (("products"."price") < '8' :: numeric(7,2))) AND (("products"."id") > '10' :: bigint))) GROUP BY "products"."id" LIMIT ('20') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/syntheticTables
SELECT jsonb_build_object('me', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT FROM (SELECT "users"."email" FROM "users" WHERE ((("users"."id") = $1 :: bigint)) LIMIT ('1') :: integer) AS "users_0") AS "__sr_0") AS "__sj_0"
=== RUN TestCompileQuery/queryWithVariables
SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name" FROM (SELECT "products"."id", "products"."name" FROM "products" WHERE (((("products"."price") = $1 :: numeric(7,2)) AND (("products"."id") = $2 :: bigint) AND ((("products"."price") > '0' :: numeric(7,2)) AND (("products"."price") < '8' :: numeric(7,2))))) LIMIT ('1') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0"
=== RUN TestCompileQuery/withWhereOnRelations
SELECT jsonb_build_object('users', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "users_0"."id" AS "id", "users_0"."email" AS "email" FROM (SELECT "users"."id", "users"."email" FROM "users" WHERE (NOT EXISTS (SELECT 1 FROM products WHERE (("products"."user_id") = ("users"."id")) AND ((("products"."price") > '3' :: numeric(7,2))))) LIMIT ('20') :: integer) AS "users_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/multiRoot
SELECT jsonb_build_object('customer', "__sj_0"."json", 'user', "__sj_1"."json", 'product', "__sj_2"."json") as "__root" FROM (SELECT to_jsonb("__sr_2".*) AS "json"FROM (SELECT "products_2"."id" AS "id", "products_2"."name" AS "name", "__sj_3"."json" AS "customers", "__sj_4"."json" AS "customer" FROM (SELECT "products"."id", "products"."name" FROM "products" WHERE (((("products"."price") > '0' :: numeric(7,2)) AND (("products"."price") < '8' :: numeric(7,2)))) LIMIT ('1') :: integer) AS "products_2" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_4".*) AS "json"FROM (SELECT "customers_4"."email" AS "email" FROM (SELECT "customers"."email" FROM "customers" LEFT OUTER JOIN "purchases" ON (("purchases"."customer_id") = ("customers"."id")) WHERE ((("purchases"."product_id") = ("products"."id"))) LIMIT ('1') :: integer) AS "customers_4") AS "__sr_4") AS "__sj_4" ON ('true') LEFT OUTER JOIN LATERAL (SELECT coalesce(jsonb_agg("__sj_3"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_3".*) AS "json"FROM (SELECT "customers_3"."email" AS "email" FROM (SELECT "customers"."email" FROM "customers" LEFT OUTER JOIN "purchases" ON (("purchases"."customer_id") = ("customers"."id")) WHERE ((("purchases"."product_id") = ("products"."id"))) LIMIT ('20') :: integer) AS "customers_3") AS "__sr_3") AS "__sj_3") AS "__sj_3" ON ('true')) AS "__sr_2") AS "__sj_2", (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "users_1"."id" AS "id", "users_1"."email" AS "email" FROM (SELECT "users"."id", "users"."email" FROM "users" LIMIT ('1') :: integer) AS "users_1") AS "__sr_1") AS "__sj_1", (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "customers_0"."id" AS "id" FROM (SELECT "customers"."id" FROM "customers" LIMIT ('1') :: integer) AS "customers_0") AS "__sr_0") AS "__sj_0"
=== RUN TestCompileQuery/withFragment1
SELECT jsonb_build_object('users', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "users_0"."first_name" AS "first_name", "users_0"."last_name" AS "last_name", "users_0"."created_at" AS "created_at", "users_0"."id" AS "id", "users_0"."email" AS "email" FROM (SELECT , "users"."created_at", "users"."id", "users"."email" FROM "users" GROUP BY "users"."created_at", "users"."id", "users"."email" LIMIT ('20') :: integer) AS "users_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/withFragment2
SELECT jsonb_build_object('users', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "users_0"."first_name" AS "first_name", "users_0"."last_name" AS "last_name", "users_0"."created_at" AS "created_at", "users_0"."id" AS "id", "users_0"."email" AS "email" FROM (SELECT , "users"."created_at", "users"."id", "users"."email" FROM "users" GROUP BY "users"."created_at", "users"."id", "users"."email" LIMIT ('20') :: integer) AS "users_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/withFragment3
SELECT jsonb_build_object('users', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "users_0"."first_name" AS "first_name", "users_0"."last_name" AS "last_name", "users_0"."created_at" AS "created_at", "users_0"."id" AS "id", "users_0"."email" AS "email" FROM (SELECT , "users"."created_at", "users"."id", "users"."email" FROM "users" GROUP BY "users"."created_at", "users"."id", "users"."email" LIMIT ('20') :: integer) AS "users_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/jsonColumnAsTable
SELECT jsonb_build_object('products', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name", "__sj_1"."json" AS "tag_count" FROM (SELECT "products"."id", "products"."name" FROM "products" LIMIT ('20') :: integer) AS "products_0" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "tag_count_1"."count" AS "count", "__sj_2"."json" AS "tags" FROM (SELECT "tag_count"."count", "tag_count"."tag_id" FROM "products", json_to_recordset("products"."tag_count") AS "tag_count"(tag_id bigint, count int) WHERE ((("products"."id") = ("products_0"."id"))) LIMIT ('1') :: integer) AS "tag_count_1" LEFT OUTER JOIN LATERAL (SELECT coalesce(jsonb_agg("__sj_2"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_2".*) AS "json"FROM (SELECT "tags_2"."name" AS "name" FROM (SELECT "tags"."name" FROM "tags" WHERE ((("tags"."id") = ("tag_count_1"."tag_id"))) LIMIT ('20') :: integer) AS "tags_2") AS "__sr_2") AS "__sj_2") AS "__sj_2" ON ('true')) AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/withCursor
SELECT jsonb_build_object('products', "__sj_0"."json", 'products_cursor', "__sj_0"."cursor") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json", CONCAT_WS(',', max("__cur_0"), max("__cur_1")) as "cursor" FROM (SELECT to_jsonb("__sr_0".*) - '__cur_0' - '__cur_1' AS "json", "__cur_0", "__cur_1"FROM (SELECT "products_0"."name" AS "name", LAST_VALUE("products_0"."price") OVER() AS "__cur_0", LAST_VALUE("products_0"."id") OVER() AS "__cur_1" FROM (WITH "__cur" AS (SELECT a[1] as "price", a[2] as "id" FROM string_to_array($1, ',') as a) SELECT "products"."name", "products"."id", "products"."price" FROM "products", "__cur" WHERE (((("products"."price") < "__cur"."price" :: numeric(7,2)) OR ((("products"."price") = "__cur"."price" :: numeric(7,2)) AND (("products"."id") > "__cur"."id" :: bigint)))) ORDER BY "products"."price" DESC, "products"."id" ASC LIMIT ('20') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/nullForAuthRequiredInAnon
SELECT jsonb_build_object('products', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name", NULL AS "user" FROM (SELECT "products"."id", "products"."name", "products"."user_id" FROM "products" LIMIT ('20') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
=== RUN TestCompileQuery/blockedQuery
SELECT jsonb_build_object('user', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "users_0"."id" AS "id", "users_0"."full_name" AS "full_name", "users_0"."email" AS "email" FROM (SELECT "users"."id", "users"."full_name", "users"."email" FROM "users" WHERE (false) LIMIT ('1') :: integer) AS "users_0") AS "__sr_0") AS "__sj_0"
=== RUN TestCompileQuery/blockedFunctions
SELECT jsonb_build_object('users', "__sj_0"."json") as "__root" FROM (SELECT coalesce(jsonb_agg("__sj_0"."json"), '[]') as "json" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "users_0"."email" AS "email" FROM (SELECT , "users"."email" FROM "users" WHERE (false) GROUP BY "users"."email" LIMIT ('20') :: integer) AS "users_0") AS "__sr_0") AS "__sj_0") AS "__sj_0"
--- PASS: TestCompileQuery (0.03s)
--- PASS: TestCompileQuery/withComplexArgs (0.00s)
--- PASS: TestCompileQuery/withWhereIn (0.00s)
--- PASS: TestCompileQuery/withWhereAndList (0.00s)
--- PASS: TestCompileQuery/withWhereIsNull (0.00s)
--- PASS: TestCompileQuery/withWhereMultiOr (0.00s)
--- PASS: TestCompileQuery/fetchByID (0.00s)
--- PASS: TestCompileQuery/searchQuery (0.00s)
--- PASS: TestCompileQuery/oneToMany (0.00s)
--- PASS: TestCompileQuery/oneToManyReverse (0.00s)
--- PASS: TestCompileQuery/oneToManyArray (0.00s)
--- PASS: TestCompileQuery/manyToMany (0.00s)
--- PASS: TestCompileQuery/manyToManyReverse (0.00s)
--- PASS: TestCompileQuery/aggFunction (0.00s)
--- PASS: TestCompileQuery/aggFunctionBlockedByCol (0.00s)
--- PASS: TestCompileQuery/aggFunctionDisabled (0.00s)
--- PASS: TestCompileQuery/aggFunctionWithFilter (0.00s)
--- PASS: TestCompileQuery/syntheticTables (0.00s)
--- PASS: TestCompileQuery/queryWithVariables (0.00s)
--- PASS: TestCompileQuery/withWhereOnRelations (0.00s)
--- PASS: TestCompileQuery/multiRoot (0.00s)
--- PASS: TestCompileQuery/withFragment1 (0.00s)
--- PASS: TestCompileQuery/withFragment2 (0.00s)
--- PASS: TestCompileQuery/withFragment3 (0.00s)
--- PASS: TestCompileQuery/jsonColumnAsTable (0.00s)
--- PASS: TestCompileQuery/withCursor (0.00s)
--- PASS: TestCompileQuery/nullForAuthRequiredInAnon (0.00s)
--- PASS: TestCompileQuery/blockedQuery (0.00s)
--- PASS: TestCompileQuery/blockedFunctions (0.00s)
=== RUN TestCompileUpdate
=== RUN TestCompileUpdate/singleUpdate
WITH "_sg_input" AS (SELECT $1 :: json AS j), "products" AS (UPDATE "products" SET ("name", "description") = (SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'description' AS text) FROM "_sg_input" i) WHERE ((("products"."id") = '1' :: bigint) AND (("products"."id") = $2 :: bigint)) RETURNING "products".*) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name" FROM (SELECT "products"."id", "products"."name" FROM "products" LIMIT ('1') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0"
=== RUN TestCompileUpdate/simpleUpdateWithPresets
WITH "_sg_input" AS (SELECT $1 :: json AS j), "products" AS (UPDATE "products" SET ("name", "price", "updated_at") = (SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'price' AS numeric(7,2)), 'now' :: timestamp without time zone FROM "_sg_input" i) WHERE (("products"."user_id") = $2 :: bigint) RETURNING "products".*) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id" FROM (SELECT "products"."id" FROM "products" LIMIT ('1') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0"
=== RUN TestCompileUpdate/nestedUpdateManyToMany
WITH "_sg_input" AS (SELECT $1 :: json AS j), "purchases" AS (UPDATE "purchases" SET ("sale_type", "quantity", "due_date") = (SELECT CAST( i.j ->>'sale_type' AS character varying), CAST( i.j ->>'quantity' AS integer), CAST( i.j ->>'due_date' AS timestamp without time zone) FROM "_sg_input" i) WHERE (("purchases"."id") = $2 :: bigint) RETURNING "purchases".*), "products" AS (UPDATE "products" SET ("name", "price") = (SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'price' AS numeric(7,2)) FROM "_sg_input" i) FROM "purchases" WHERE (("products"."id") = ("purchases"."product_id")) RETURNING "products".*), "customers" AS (UPDATE "customers" SET ("full_name", "email") = (SELECT CAST( i.j ->>'full_name' AS character varying), CAST( i.j ->>'email' AS character varying) FROM "_sg_input" i) FROM "purchases" WHERE (("customers"."id") = ("purchases"."customer_id")) RETURNING "customers".*) SELECT jsonb_build_object('purchase', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "purchases_0"."sale_type" AS "sale_type", "purchases_0"."quantity" AS "quantity", "purchases_0"."due_date" AS "due_date", "__sj_1"."json" AS "product", "__sj_2"."json" AS "customer" FROM (SELECT "purchases"."sale_type", "purchases"."quantity", "purchases"."due_date", "purchases"."product_id", "purchases"."customer_id" FROM "purchases" LIMIT ('1') :: integer) AS "purchases_0" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_2".*) AS "json"FROM (SELECT "customers_2"."id" AS "id", "customers_2"."full_name" AS "full_name", "customers_2"."email" AS "email" FROM (SELECT "customers"."id", "customers"."full_name", "customers"."email" FROM "customers" WHERE ((("customers"."id") = ("purchases_0"."customer_id"))) LIMIT ('1') :: integer) AS "customers_2") AS "__sr_2") AS "__sj_2" ON ('true') LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "products_1"."id" AS "id", "products_1"."name" AS "name", "products_1"."price" AS "price" FROM (SELECT "products"."id", "products"."name", "products"."price" FROM "products" WHERE ((("products"."id") = ("purchases_0"."product_id"))) LIMIT ('1') :: integer) AS "products_1") AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0"
WITH "_sg_input" AS (SELECT $1 :: json AS j), "purchases" AS (UPDATE "purchases" SET ("sale_type", "quantity", "due_date") = (SELECT CAST( i.j ->>'sale_type' AS character varying), CAST( i.j ->>'quantity' AS integer), CAST( i.j ->>'due_date' AS timestamp without time zone) FROM "_sg_input" i) WHERE (("purchases"."id") = $2 :: bigint) RETURNING "purchases".*), "customers" AS (UPDATE "customers" SET ("full_name", "email") = (SELECT CAST( i.j ->>'full_name' AS character varying), CAST( i.j ->>'email' AS character varying) FROM "_sg_input" i) FROM "purchases" WHERE (("customers"."id") = ("purchases"."customer_id")) RETURNING "customers".*), "products" AS (UPDATE "products" SET ("name", "price") = (SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'price' AS numeric(7,2)) FROM "_sg_input" i) FROM "purchases" WHERE (("products"."id") = ("purchases"."product_id")) RETURNING "products".*) SELECT jsonb_build_object('purchase', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "purchases_0"."sale_type" AS "sale_type", "purchases_0"."quantity" AS "quantity", "purchases_0"."due_date" AS "due_date", "__sj_1"."json" AS "product", "__sj_2"."json" AS "customer" FROM (SELECT "purchases"."sale_type", "purchases"."quantity", "purchases"."due_date", "purchases"."product_id", "purchases"."customer_id" FROM "purchases" LIMIT ('1') :: integer) AS "purchases_0" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_2".*) AS "json"FROM (SELECT "customers_2"."id" AS "id", "customers_2"."full_name" AS "full_name", "customers_2"."email" AS "email" FROM (SELECT "customers"."id", "customers"."full_name", "customers"."email" FROM "customers" WHERE ((("customers"."id") = ("purchases_0"."customer_id"))) LIMIT ('1') :: integer) AS "customers_2") AS "__sr_2") AS "__sj_2" ON ('true') LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "products_1"."id" AS "id", "products_1"."name" AS "name", "products_1"."price" AS "price" FROM (SELECT "products"."id", "products"."name", "products"."price" FROM "products" WHERE ((("products"."id") = ("purchases_0"."product_id"))) LIMIT ('1') :: integer) AS "products_1") AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0"
=== RUN TestCompileUpdate/nestedUpdateOneToMany
WITH "_sg_input" AS (SELECT $1 :: json AS j), "users" AS (UPDATE "users" SET ("full_name", "email", "created_at", "updated_at") = (SELECT CAST( i.j ->>'full_name' AS character varying), CAST( i.j ->>'email' AS character varying), CAST( i.j ->>'created_at' AS timestamp without time zone), CAST( i.j ->>'updated_at' AS timestamp without time zone) FROM "_sg_input" i) WHERE (("users"."id") = '8' :: bigint) RETURNING "users".*), "products" AS (UPDATE "products" SET ("name", "price", "created_at", "updated_at") = (SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'price' AS numeric(7,2)), CAST( i.j ->>'created_at' AS timestamp without time zone), CAST( i.j ->>'updated_at' AS timestamp without time zone) FROM "_sg_input" i) FROM "users" WHERE (("products"."user_id") = ("users"."id") AND "products"."id"= ((i.j->'product'->'where'->>'id'))::bigint) RETURNING "products".*) SELECT jsonb_build_object('user', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "users_0"."id" AS "id", "users_0"."full_name" AS "full_name", "users_0"."email" AS "email", "__sj_1"."json" AS "product" FROM (SELECT "users"."id", "users"."full_name", "users"."email" FROM "users" LIMIT ('1') :: integer) AS "users_0" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "products_1"."id" AS "id", "products_1"."name" AS "name", "products_1"."price" AS "price" FROM (SELECT "products"."id", "products"."name", "products"."price" FROM "products" WHERE ((("products"."user_id") = ("users_0"."id"))) LIMIT ('1') :: integer) AS "products_1") AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0"
=== RUN TestCompileUpdate/nestedUpdateOneToOne
WITH "_sg_input" AS (SELECT $1 :: json AS j), "products" AS (UPDATE "products" SET ("name", "price", "created_at", "updated_at") = (SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'price' AS numeric(7,2)), CAST( i.j ->>'created_at' AS timestamp without time zone), CAST( i.j ->>'updated_at' AS timestamp without time zone) FROM "_sg_input" i) WHERE (("products"."id") = $2 :: bigint) RETURNING "products".*), "users" AS (UPDATE "users" SET ("email") = (SELECT CAST( i.j ->>'email' AS character varying) FROM "_sg_input" i) FROM "products" WHERE (("users"."id") = ("products"."user_id")) RETURNING "users".*) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name", "__sj_1"."json" AS "user" FROM (SELECT "products"."id", "products"."name", "products"."user_id" FROM "products" LIMIT ('1') :: integer) AS "products_0" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "users_1"."id" AS "id", "users_1"."full_name" AS "full_name", "users_1"."email" AS "email" FROM (SELECT "users"."id", "users"."full_name", "users"."email" FROM "users" WHERE ((("users"."id") = ("products_0"."user_id"))) LIMIT ('1') :: integer) AS "users_1") AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0"
=== RUN TestCompileUpdate/nestedUpdateOneToManyWithConnect
WITH "_sg_input" AS (SELECT $1 :: json AS j), "users" AS (UPDATE "users" SET ("full_name", "email", "created_at", "updated_at") = (SELECT CAST( i.j ->>'full_name' AS character varying), CAST( i.j ->>'email' AS character varying), CAST( i.j ->>'created_at' AS timestamp without time zone), CAST( i.j ->>'updated_at' AS timestamp without time zone) FROM "_sg_input" i) WHERE (("users"."id") = $2 :: bigint) RETURNING "users".*), "products_c" AS ( UPDATE "products" SET "user_id" = "users"."id" FROM "users" WHERE ("products"."id"= ((i.j->'product'->'connect'->>'id'))::bigint) RETURNING "products".*), "products_d" AS ( UPDATE "products" SET "user_id" = NULL FROM "users" WHERE ("products"."id"= ((i.j->'product'->'disconnect'->>'id'))::bigint) RETURNING "products".*), "products" AS (SELECT * FROM "products_c" UNION ALL SELECT * FROM "products_d") SELECT jsonb_build_object('user', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "users_0"."id" AS "id", "users_0"."full_name" AS "full_name", "users_0"."email" AS "email", "__sj_1"."json" AS "product" FROM (SELECT "users"."id", "users"."full_name", "users"."email" FROM "users" LIMIT ('1') :: integer) AS "users_0" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "products_1"."id" AS "id", "products_1"."name" AS "name", "products_1"."price" AS "price" FROM (SELECT "products"."id", "products"."name", "products"."price" FROM "products" WHERE ((("products"."user_id") = ("users_0"."id"))) LIMIT ('1') :: integer) AS "products_1") AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0"
=== RUN TestCompileUpdate/nestedUpdateOneToOneWithConnect
WITH "_sg_input" AS (SELECT $1 :: json AS j), "_x_users" AS (SELECT "id" FROM "_sg_input" i,"users" WHERE "users"."email"= ((i.j->'user'->'connect'->>'email'))::character varying AND "users"."id"= ((i.j->'user'->'connect'->>'id'))::bigint LIMIT 1), "products" AS (UPDATE "products" SET ("name", "price", "user_id") = (SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'price' AS numeric(7,2)), "_x_users"."id" FROM "_sg_input" i, "_x_users") WHERE (("products"."id") = $2 :: bigint) RETURNING "products".*) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name", "__sj_1"."json" AS "user" FROM (SELECT "products"."id", "products"."name", "products"."user_id" FROM "products" LIMIT ('1') :: integer) AS "products_0" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "users_1"."id" AS "id", "users_1"."full_name" AS "full_name", "users_1"."email" AS "email" FROM (SELECT "users"."id", "users"."full_name", "users"."email" FROM "users" WHERE ((("users"."id") = ("products_0"."user_id"))) LIMIT ('1') :: integer) AS "users_1") AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0"
WITH "_sg_input" AS (SELECT $1 :: json AS j), "_x_users" AS (SELECT "id" FROM "_sg_input" i,"users" WHERE "users"."id"= ((i.j->'user'->'connect'->>'id'))::bigint AND "users"."email"= ((i.j->'user'->'connect'->>'email'))::character varying LIMIT 1), "products" AS (UPDATE "products" SET ("name", "price", "user_id") = (SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'price' AS numeric(7,2)), "_x_users"."id" FROM "_sg_input" i, "_x_users") WHERE (("products"."id") = $2 :: bigint) RETURNING "products".*) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name", "__sj_1"."json" AS "user" FROM (SELECT "products"."id", "products"."name", "products"."user_id" FROM "products" LIMIT ('1') :: integer) AS "products_0" LEFT OUTER JOIN LATERAL (SELECT to_jsonb("__sr_1".*) AS "json"FROM (SELECT "users_1"."id" AS "id", "users_1"."full_name" AS "full_name", "users_1"."email" AS "email" FROM (SELECT "users"."id", "users"."full_name", "users"."email" FROM "users" WHERE ((("users"."id") = ("products_0"."user_id"))) LIMIT ('1') :: integer) AS "users_1") AS "__sr_1") AS "__sj_1" ON ('true')) AS "__sr_0") AS "__sj_0"
=== RUN TestCompileUpdate/nestedUpdateOneToOneWithDisconnect
WITH "_sg_input" AS (SELECT $1 :: json AS j), "_x_users" AS (SELECT * FROM (VALUES(NULL::bigint)) AS LOOKUP("id")), "products" AS (UPDATE "products" SET ("name", "price", "user_id") = (SELECT CAST( i.j ->>'name' AS character varying), CAST( i.j ->>'price' AS numeric(7,2)), "_x_users"."id" FROM "_sg_input" i, "_x_users") WHERE (("products"."id") = $2 :: bigint) RETURNING "products".*) SELECT jsonb_build_object('product', "__sj_0"."json") as "__root" FROM (SELECT to_jsonb("__sr_0".*) AS "json"FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name", "products_0"."user_id" AS "user_id" FROM (SELECT "products"."id", "products"."name", "products"."user_id" FROM "products" LIMIT ('1') :: integer) AS "products_0") AS "__sr_0") AS "__sj_0"
--- PASS: TestCompileUpdate (0.02s)
--- PASS: TestCompileUpdate/singleUpdate (0.00s)
--- PASS: TestCompileUpdate/simpleUpdateWithPresets (0.00s)
--- PASS: TestCompileUpdate/nestedUpdateManyToMany (0.00s)
--- PASS: TestCompileUpdate/nestedUpdateOneToMany (0.00s)
--- PASS: TestCompileUpdate/nestedUpdateOneToOne (0.00s)
--- PASS: TestCompileUpdate/nestedUpdateOneToManyWithConnect (0.00s)
--- PASS: TestCompileUpdate/nestedUpdateOneToOneWithConnect (0.00s)
--- PASS: TestCompileUpdate/nestedUpdateOneToOneWithDisconnect (0.00s)
PASS
ok github.com/dosco/super-graph/core/internal/psql 0.323s

View File

@ -1,258 +0,0 @@
package psql_test
import (
"encoding/json"
"testing"
)
func singleUpdate(t *testing.T) {
gql := `mutation {
product(id: $id, update: $update, where: { id: { eq: 1 } }) {
id
name
}
}`
vars := map[string]json.RawMessage{
"update": json.RawMessage(` { "name": "my_name", "description": "my_desc" }`),
}
compileGQLToPSQL(t, gql, vars, "anon")
}
func simpleUpdateWithPresets(t *testing.T) {
gql := `mutation {
product(update: $data) {
id
}
}`
vars := map[string]json.RawMessage{
"data": json.RawMessage(`{"name": "Apple", "price": 1.25}`),
}
compileGQLToPSQL(t, gql, vars, "user")
}
func nestedUpdateManyToMany(t *testing.T) {
gql := `mutation {
purchase(update: $data, id: $id) {
sale_type
quantity
due_date
customer {
id
full_name
email
}
product {
id
name
price
}
}
}`
vars := map[string]json.RawMessage{
"data": json.RawMessage(` {
"sale_type": "bought",
"quantity": 5,
"due_date": "now",
"customer": {
"email": "thedude@rug.com",
"full_name": "The Dude"
},
"product": {
"name": "Apple",
"price": 1.25
}
}
`),
}
compileGQLToPSQL(t, gql, vars, "admin")
}
func nestedUpdateOneToMany(t *testing.T) {
gql := `mutation {
user(update: $data, where: { id: { eq: 8 } }) {
id
full_name
email
product {
id
name
price
}
}
}`
vars := map[string]json.RawMessage{
"data": json.RawMessage(`{
"email": "thedude@rug.com",
"full_name": "The Dude",
"created_at": "now",
"updated_at": "now",
"product": {
"where": {
"id": 2
},
"name": "Apple",
"price": 1.25,
"created_at": "now",
"updated_at": "now"
}
}`),
}
compileGQLToPSQL(t, gql, vars, "admin")
}
func nestedUpdateOneToOne(t *testing.T) {
gql := `mutation {
product(update: $data, id: $id) {
id
name
user {
id
full_name
email
}
}
}`
vars := map[string]json.RawMessage{
"data": json.RawMessage(`{
"name": "Apple",
"price": 1.25,
"created_at": "now",
"updated_at": "now",
"user": {
"email": "thedude@rug.com"
}
}`),
}
compileGQLToPSQL(t, gql, vars, "admin")
}
func nestedUpdateOneToManyWithConnect(t *testing.T) {
gql := `mutation {
user(update: $data, id: $id) {
id
full_name
email
product {
id
name
price
}
}
}`
vars := map[string]json.RawMessage{
"data": json.RawMessage(`{
"email": "thedude@rug.com",
"full_name": "The Dude",
"created_at": "now",
"updated_at": "now",
"product": {
"connect": { "id": 7 },
"disconnect": { "id": 8 }
}
}`),
}
compileGQLToPSQL(t, gql, vars, "admin")
}
func nestedUpdateOneToOneWithConnect(t *testing.T) {
gql := `mutation {
product(update: $data, id: $product_id) {
id
name
user {
id
full_name
email
}
}
}`
vars := map[string]json.RawMessage{
"data": json.RawMessage(`{
"name": "Apple",
"price": 1.25,
"user": {
"connect": { "id": 5, "email": "test@test.com" }
}
}`),
}
compileGQLToPSQL(t, gql, vars, "admin")
}
func nestedUpdateOneToOneWithDisconnect(t *testing.T) {
gql := `mutation {
product(update: $data, id: $id) {
id
name
user_id
}
}`
vars := map[string]json.RawMessage{
"data": json.RawMessage(`{
"name": "Apple",
"price": 1.25,
"user": {
"disconnect": { "id": 5 }
}
}`),
}
compileGQLToPSQL(t, gql, vars, "admin")
}
// func nestedUpdateOneToOneWithDisconnectArray(t *testing.T) {
// gql := `mutation {
// product(update: $data, id: 2) {
// id
// name
// user_id
// }
// }`
// sql := `WITH "_sg_input" AS (SELECT $1 :: json AS j), "users" AS (SELECT * FROM (VALUES(NULL::bigint)) AS LOOKUP("id")), "products" AS (UPDATE "products" SET ("name", "price", "user_id") = (SELECT "t"."name", "t"."price", "users"."id" FROM "_sg_input" i, "users", json_populate_record(NULL::products, i.j) t) WHERE (("products"."id") = 2) RETURNING "products".*) SELECT json_object_agg('product', json_0) FROM (SELECT row_to_json((SELECT "json_row_0" FROM (SELECT "products_0"."id" AS "id", "products_0"."name" AS "name", "products_0"."user_id" AS "user_id") AS "json_row_0")) AS "json_0" FROM (SELECT "products"."id", "products"."name", "products"."user_id" FROM "products" LIMIT ('1') :: integer) AS "products_0" LIMIT ('1') :: integer) AS "sel_0"`
// vars := map[string]json.RawMessage{
// "data": json.RawMessage(`{
// "name": "Apple",
// "price": 1.25,
// "user": {
// "disconnect": { "id": 5 }
// }
// }`),
// }
// resSQL, err := compileGQLToPSQL(gql, vars, "admin")
// if err != nil {
// t.Fatal(err)
// }
// if string(resSQL) != sql {
// t.Fatal(errNotExpected)
// }
// }
func TestCompileUpdate(t *testing.T) {
t.Run("singleUpdate", singleUpdate)
t.Run("simpleUpdateWithPresets", simpleUpdateWithPresets)
t.Run("nestedUpdateManyToMany", nestedUpdateManyToMany)
t.Run("nestedUpdateOneToMany", nestedUpdateOneToMany)
t.Run("nestedUpdateOneToOne", nestedUpdateOneToOne)
t.Run("nestedUpdateOneToManyWithConnect", nestedUpdateOneToManyWithConnect)
t.Run("nestedUpdateOneToOneWithConnect", nestedUpdateOneToOneWithConnect)
t.Run("nestedUpdateOneToOneWithDisconnect", nestedUpdateOneToOneWithDisconnect)
//t.Run("nestedUpdateOneToOneWithDisconnectArray", nestedUpdateOneToOneWithDisconnectArray)
}

View File

@ -1,11 +0,0 @@
goos: darwin
goarch: amd64
pkg: github.com/dosco/super-graph/core/internal/qcode
BenchmarkQCompile-16 120888 9236 ns/op 3755 B/op 28 allocs/op
BenchmarkQCompileP-16 502248 2620 ns/op 3795 B/op 28 allocs/op
BenchmarkParse-16 128370 9294 ns/op 3902 B/op 18 allocs/op
BenchmarkParseP-16 575752 2340 ns/op 3903 B/op 18 allocs/op
BenchmarkSchemaParse-16 212048 5779 ns/op 3968 B/op 57 allocs/op
BenchmarkSchemaParseP-16 630918 1686 ns/op 3968 B/op 57 allocs/op
PASS
ok github.com/dosco/super-graph/core/internal/qcode 7.710s

View File

@ -1,13 +0,0 @@
goos: darwin
goarch: amd64
pkg: github.com/dosco/super-graph/core/internal/qcode
BenchmarkQCompile-16 118282 9686 ns/op 4031 B/op 30 allocs/op
BenchmarkQCompileP-16 427531 2710 ns/op 4077 B/op 30 allocs/op
BenchmarkQCompileFragment-16 140588 8328 ns/op 8903 B/op 13 allocs/op
BenchmarkParse-16 131396 9212 ns/op 4175 B/op 18 allocs/op
BenchmarkParseP-16 503778 2310 ns/op 4176 B/op 18 allocs/op
BenchmarkParseFragment-16 143725 8158 ns/op 10193 B/op 9 allocs/op
BenchmarkSchemaParse-16 240609 5060 ns/op 3968 B/op 57 allocs/op
BenchmarkSchemaParseP-16 785116 1534 ns/op 3968 B/op 57 allocs/op
PASS
ok github.com/dosco/super-graph/core/internal/qcode 11.092s

View File

@ -1,17 +0,0 @@
goos: darwin
goarch: amd64
pkg: github.com/dosco/super-graph/core/internal/qcode
BenchmarkQCompile
BenchmarkQCompile-16 129614 8649 ns/op 3756 B/op 28 allocs/op
BenchmarkQCompileP
BenchmarkQCompileP-16 487488 2525 ns/op 3792 B/op 28 allocs/op
BenchmarkParse
BenchmarkParse-16 127582 8731 ns/op 3902 B/op 18 allocs/op
BenchmarkParseP
BenchmarkParseP-16 561373 2223 ns/op 3903 B/op 18 allocs/op
BenchmarkSchemaParse
BenchmarkSchemaParse-16 209142 5523 ns/op 3968 B/op 57 allocs/op
BenchmarkSchemaParseP
BenchmarkSchemaParseP-16 716437 1734 ns/op 3968 B/op 57 allocs/op
PASS
ok github.com/dosco/super-graph/core/internal/qcode 8.483s

View File

@ -1,377 +0,0 @@
package qcode
import (
"errors"
"testing"
"github.com/chirino/graphql/schema"
)
func TestCompile1(t *testing.T) {
qc, _ := NewCompiler(Config{})
err := qc.AddRole("user", "product", TRConfig{
Query: QueryConfig{
Columns: []string{"id", "Name"},
},
})
if err != nil {
t.Error(err)
}
_, err = qc.Compile([]byte(`
query { product(id: 15) {
id
name
} }`), "user")
if err == nil {
t.Fatal(errors.New("this should be an error id must be a variable"))
}
}
func TestCompile2(t *testing.T) {
qc, _ := NewCompiler(Config{})
err := qc.AddRole("user", "product", TRConfig{
Query: QueryConfig{
Columns: []string{"ID"},
},
})
if err != nil {
t.Error(err)
}
_, err = qc.Compile([]byte(`
query { product(id: $id) {
id
name
} }`), "user")
if err != nil {
t.Fatal(err)
}
}
func TestCompile3(t *testing.T) {
qc, _ := NewCompiler(Config{})
err := qc.AddRole("user", "product", TRConfig{
Query: QueryConfig{
Columns: []string{"ID"},
},
})
if err != nil {
t.Error(err)
}
_, err = qc.Compile([]byte(`
mutation {
product(id: $test, name: "Test") {
id
name
}
}`), "user")
if err != nil {
t.Fatal(err)
}
}
func TestInvalidCompile1(t *testing.T) {
qcompile, _ := NewCompiler(Config{})
_, err := qcompile.Compile([]byte(`#`), "user")
if err == nil {
t.Fatal(errors.New("expecting an error"))
}
}
func TestInvalidCompile2(t *testing.T) {
qcompile, _ := NewCompiler(Config{})
_, err := qcompile.Compile([]byte(`{u(where:{not:0})}`), "user")
if err == nil {
t.Fatal(errors.New("expecting an error"))
}
}
func TestEmptyCompile(t *testing.T) {
qcompile, _ := NewCompiler(Config{})
_, err := qcompile.Compile([]byte(``), "user")
if err == nil {
t.Fatal(errors.New("expecting an error"))
}
}
func TestInvalidPostfixCompile(t *testing.T) {
gql := `mutation
updateThread {
thread(update: $data, where: { slug: { eq: $slug } }) {
slug
title
published
createdAt : created_at
totalVotes : cached_votes_total
totalPosts : cached_posts_total
vote : thread_vote(where: { user_id: { eq: $user_id } }) {
id
}
topics {
slug
name
}
}
}
}}`
qcompile, _ := NewCompiler(Config{})
_, err := qcompile.Compile([]byte(gql), "anon")
if err == nil {
t.Fatal(errors.New("expecting an error"))
}
}
func TestFragmentsCompile1(t *testing.T) {
gql := `
fragment userFields1 on user {
id
email
}
query {
users {
...userFields2
created_at
...userFields1
}
}
fragment userFields2 on user {
first_name
last_name
}
`
qcompile, _ := NewCompiler(Config{})
_, err := qcompile.Compile([]byte(gql), "user")
if err != nil {
t.Fatal(err)
}
}
func TestFragmentsCompile2(t *testing.T) {
gql := `
query {
users {
...userFields2
created_at
...userFields1
}
}
fragment userFields1 on user {
id
email
}
fragment userFields2 on user {
first_name
last_name
}`
qcompile, _ := NewCompiler(Config{})
_, err := qcompile.Compile([]byte(gql), "user")
if err != nil {
t.Fatal(err)
}
}
func TestFragmentsCompile3(t *testing.T) {
gql := `
fragment userFields1 on user {
id
email
}
fragment userFields2 on user {
first_name
last_name
}
query {
users {
...userFields2
created_at
...userFields1
}
}
`
qcompile, _ := NewCompiler(Config{})
_, err := qcompile.Compile([]byte(gql), "user")
if err != nil {
t.Fatal(err)
}
}
var gql = []byte(`
{products(
# returns only 30 items
limit: 30,
# starts from item 10, commented out for now
# offset: 10,
# orders the response items by highest price
order_by: { price: desc },
# no duplicate prices returned
distinct: [ price ]
# only items with an id >= 30 and < 30 are returned
where: { id: { AND: { greater_or_equals: 20, lt: 28 } } }) {
id
name
price
}}`)
var gqlWithFragments = []byte(`
fragment userFields1 on user {
id
email
__typename
}
query {
users {
...userFields2
created_at
...userFields1
__typename
}
}
fragment userFields2 on user {
first_name
last_name
__typename
}`)
func BenchmarkQCompile(b *testing.B) {
qcompile, _ := NewCompiler(Config{})
b.ResetTimer()
b.ReportAllocs()
for n := 0; n < b.N; n++ {
_, err := qcompile.Compile(gql, "user")
if err != nil {
b.Fatal(err)
}
}
}
func BenchmarkQCompileP(b *testing.B) {
qcompile, _ := NewCompiler(Config{})
b.ResetTimer()
b.ReportAllocs()
b.RunParallel(func(pb *testing.PB) {
for pb.Next() {
_, err := qcompile.Compile(gql, "user")
if err != nil {
b.Fatal(err)
}
}
})
}
func BenchmarkQCompileFragment(b *testing.B) {
qcompile, _ := NewCompiler(Config{})
b.ResetTimer()
b.ReportAllocs()
for n := 0; n < b.N; n++ {
_, err := qcompile.Compile(gqlWithFragments, "user")
if err != nil {
b.Fatal(err)
}
}
}
func BenchmarkParse(b *testing.B) {
b.ResetTimer()
b.ReportAllocs()
for n := 0; n < b.N; n++ {
_, err := Parse(gql)
if err != nil {
b.Fatal(err)
}
}
}
func BenchmarkParseP(b *testing.B) {
b.ResetTimer()
b.ReportAllocs()
b.RunParallel(func(pb *testing.PB) {
for pb.Next() {
_, err := Parse(gql)
if err != nil {
b.Fatal(err)
}
}
})
}
func BenchmarkParseFragment(b *testing.B) {
b.ResetTimer()
b.ReportAllocs()
for n := 0; n < b.N; n++ {
_, err := Parse(gqlWithFragments)
if err != nil {
b.Fatal(err)
}
}
}
func BenchmarkSchemaParse(b *testing.B) {
b.ResetTimer()
b.ReportAllocs()
for n := 0; n < b.N; n++ {
doc := schema.QueryDocument{}
err := doc.Parse(string(gql))
if err != nil {
b.Fatal(err)
}
}
}
func BenchmarkSchemaParseP(b *testing.B) {
b.ResetTimer()
b.ReportAllocs()
b.RunParallel(func(pb *testing.PB) {
for pb.Next() {
doc := schema.QueryDocument{}
err := doc.Parse(string(gql))
if err != nil {
b.Fatal(err)
}
}
})
}

View File

@ -1,49 +0,0 @@
package qcode
func GetQType(gql string) QType {
ic := false
for i := range gql {
b := gql[i]
switch {
case b == '#':
ic = true
case b == '\n':
ic = false
case !ic && b == '{':
return QTQuery
case !ic && al(b):
switch b {
case 'm', 'M':
return QTMutation
case 'q', 'Q':
return QTQuery
}
}
}
return -1
}
func al(b byte) bool {
return (b >= 'a' && b <= 'z') || (b >= 'A' && b <= 'Z') || (b >= '0' && b <= '9')
}
func (qt QType) String() string {
switch qt {
case QTUnknown:
return "unknown"
case QTQuery:
return "query"
case QTMutation:
return "mutation"
case QTInsert:
return "insert"
case QTUpdate:
return "update"
case QTDelete:
return "delete"
case QTUpsert:
return "upsert"
}
return ""
}

View File

@ -1,50 +0,0 @@
package qcode
import "testing"
func TestGetQType(t *testing.T) {
type args struct {
gql string
}
type ts struct {
name string
args args
want QType
}
tests := []ts{
ts{
name: "query",
args: args{gql: " query {"},
want: QTQuery,
},
ts{
name: "mutation",
args: args{gql: " mutation {"},
want: QTMutation,
},
ts{
name: "default query",
args: args{gql: " {"},
want: QTQuery,
},
ts{
name: "default query with comment",
args: args{gql: `# query is good
{`},
want: QTQuery,
},
ts{
name: "failed query with comment",
args: args{gql: `# query is good query {`},
want: -1,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
if got := GetQType(tt.args.gql); got != tt.want {
t.Errorf("GetQType() = %v, want %v", got, tt.want)
}
})
}
}

View File

@ -1,490 +0,0 @@
package core
import (
"strings"
"github.com/chirino/graphql"
"github.com/chirino/graphql/resolvers"
"github.com/chirino/graphql/schema"
"github.com/dosco/super-graph/core/internal/psql"
)
var typeMap map[string]string = map[string]string{
"smallint": "Int",
"integer": "Int",
"bigint": "Int",
"smallserial": "Int",
"serial": "Int",
"bigserial": "Int",
"decimal": "Float",
"numeric": "Float",
"real": "Float",
"double precision": "Float",
"money": "Float",
"boolean": "Boolean",
}
func (sg *SuperGraph) initGraphQLEgine() error {
engine := graphql.New()
engineSchema := engine.Schema
dbSchema := sg.schema
if err := engineSchema.Parse(`enum OrderDirection { asc desc }`); err != nil {
return err
}
gqltype := func(col psql.DBColumn) schema.Type {
typeName := typeMap[strings.ToLower(col.Type)]
if typeName == "" {
typeName = "String"
}
var t schema.Type = &schema.TypeName{Name: typeName}
if col.NotNull {
t = &schema.NonNull{OfType: t}
}
return t
}
query := &schema.Object{
Name: "Query",
Fields: schema.FieldList{},
}
mutation := &schema.Object{
Name: "Mutation",
Fields: schema.FieldList{},
}
engineSchema.Types[query.Name] = query
engineSchema.Types[mutation.Name] = mutation
engineSchema.EntryPoints[schema.Query] = query
engineSchema.EntryPoints[schema.Mutation] = mutation
//validGraphQLIdentifierRegex := regexp.MustCompile(`^[A-Za-z_][A-Za-z_0-9]*$`)
scalarExpressionTypesNeeded := map[string]bool{}
tableNames := dbSchema.GetTableNames()
funcs := dbSchema.GetFunctions()
for _, table := range tableNames {
ti, err := dbSchema.GetTable(table)
if err != nil {
return err
}
if !ti.IsSingular {
continue
}
singularName := ti.Singular
// if !validGraphQLIdentifierRegex.MatchString(singularName) {
// return errors.New("table name is not a valid GraphQL identifier: " + singularName)
// }
pluralName := ti.Plural
// if !validGraphQLIdentifierRegex.MatchString(pluralName) {
// return errors.New("table name is not a valid GraphQL identifier: " + pluralName)
// }
outputType := &schema.Object{
Name: singularName + "Output",
Fields: schema.FieldList{},
}
engineSchema.Types[outputType.Name] = outputType
inputType := &schema.InputObject{
Name: singularName + "Input",
Fields: schema.InputValueList{},
}
engineSchema.Types[inputType.Name] = inputType
orderByType := &schema.InputObject{
Name: singularName + "OrderBy",
Fields: schema.InputValueList{},
}
engineSchema.Types[orderByType.Name] = orderByType
expressionTypeName := singularName + "Expression"
expressionType := &schema.InputObject{
Name: expressionTypeName,
Fields: schema.InputValueList{
&schema.InputValue{
Name: "and",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: expressionTypeName}},
},
&schema.InputValue{
Name: "or",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: expressionTypeName}},
},
&schema.InputValue{
Name: "not",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: expressionTypeName}},
},
},
}
engineSchema.Types[expressionType.Name] = expressionType
for _, col := range ti.Columns {
colName := col.Name
// if !validGraphQLIdentifierRegex.MatchString(colName) {
// return errors.New("column name is not a valid GraphQL identifier: " + colName)
// }
colType := gqltype(col)
nullableColType := ""
if x, ok := colType.(*schema.NonNull); ok {
nullableColType = x.OfType.(*schema.TypeName).Name
} else {
nullableColType = colType.(*schema.TypeName).Name
}
outputType.Fields = append(outputType.Fields, &schema.Field{
Name: colName,
Type: colType,
})
for _, f := range funcs {
if col.Type != f.Params[0].Type {
continue
}
outputType.Fields = append(outputType.Fields, &schema.Field{
Name: f.Name + "_" + colName,
Type: colType,
})
}
// If it's a numeric type...
if nullableColType == "Float" || nullableColType == "Int" {
outputType.Fields = append(outputType.Fields, &schema.Field{
Name: "avg_" + colName,
Type: colType,
})
outputType.Fields = append(outputType.Fields, &schema.Field{
Name: "count_" + colName,
Type: colType,
})
outputType.Fields = append(outputType.Fields, &schema.Field{
Name: "max_" + colName,
Type: colType,
})
outputType.Fields = append(outputType.Fields, &schema.Field{
Name: "min_" + colName,
Type: colType,
})
outputType.Fields = append(outputType.Fields, &schema.Field{
Name: "stddev_" + colName,
Type: colType,
})
outputType.Fields = append(outputType.Fields, &schema.Field{
Name: "stddev_pop_" + colName,
Type: colType,
})
outputType.Fields = append(outputType.Fields, &schema.Field{
Name: "stddev_samp_" + colName,
Type: colType,
})
outputType.Fields = append(outputType.Fields, &schema.Field{
Name: "variance_" + colName,
Type: colType,
})
outputType.Fields = append(outputType.Fields, &schema.Field{
Name: "var_pop_" + colName,
Type: colType,
})
outputType.Fields = append(outputType.Fields, &schema.Field{
Name: "var_samp_" + colName,
Type: colType,
})
}
inputType.Fields = append(inputType.Fields, &schema.InputValue{
Name: colName,
Type: colType,
})
orderByType.Fields = append(orderByType.Fields, &schema.InputValue{
Name: colName,
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "OrderDirection"}},
})
scalarExpressionTypesNeeded[nullableColType] = true
expressionType.Fields = append(expressionType.Fields, &schema.InputValue{
Name: colName,
Type: &schema.NonNull{OfType: &schema.TypeName{Name: nullableColType + "Expression"}},
})
}
outputTypeName := &schema.TypeName{Name: outputType.Name}
inputTypeName := &schema.TypeName{Name: inputType.Name}
pluralOutputTypeName := &schema.NonNull{OfType: &schema.List{OfType: &schema.NonNull{OfType: &schema.TypeName{Name: outputType.Name}}}}
pluralInputTypeName := &schema.NonNull{OfType: &schema.List{OfType: &schema.NonNull{OfType: &schema.TypeName{Name: inputType.Name}}}}
args := schema.InputValueList{
&schema.InputValue{
Desc: schema.Description{Text: "To sort or ordering results just use the order_by argument. This can be combined with where, search, etc to build complex queries to fit you needs."},
Name: "order_by",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: orderByType.Name}},
},
&schema.InputValue{
Desc: schema.Description{Text: ""},
Name: "where",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: expressionType.Name}},
},
&schema.InputValue{
Desc: schema.Description{Text: ""},
Name: "limit",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "Int"}},
},
&schema.InputValue{
Desc: schema.Description{Text: ""},
Name: "offset",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "Int"}},
},
&schema.InputValue{
Desc: schema.Description{Text: ""},
Name: "first",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "Int"}},
},
&schema.InputValue{
Desc: schema.Description{Text: ""},
Name: "last",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "Int"}},
},
&schema.InputValue{
Desc: schema.Description{Text: ""},
Name: "before",
Type: &schema.TypeName{Name: "String"},
},
&schema.InputValue{
Desc: schema.Description{Text: ""},
Name: "after",
Type: &schema.TypeName{Name: "String"},
},
}
if ti.PrimaryCol != nil {
t := gqltype(*ti.PrimaryCol)
if _, ok := t.(*schema.NonNull); !ok {
t = &schema.NonNull{OfType: t}
}
args = append(args, &schema.InputValue{
Desc: schema.Description{Text: "Finds the record by the primary key"},
Name: "id",
Type: t,
})
}
if ti.TSVCol != nil {
args = append(args, &schema.InputValue{
Desc: schema.Description{Text: "Performs full text search using a TSV index"},
Name: "search",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "String"}},
})
}
query.Fields = append(query.Fields, &schema.Field{
Desc: schema.Description{Text: ""},
Name: singularName,
Type: outputTypeName,
Args: args,
})
query.Fields = append(query.Fields, &schema.Field{
Desc: schema.Description{Text: ""},
Name: pluralName,
Type: pluralOutputTypeName,
Args: args,
})
mutationArgs := append(args, schema.InputValueList{
&schema.InputValue{
Desc: schema.Description{Text: ""},
Name: "insert",
Type: inputTypeName,
},
&schema.InputValue{
Desc: schema.Description{Text: ""},
Name: "update",
Type: inputTypeName,
},
&schema.InputValue{
Desc: schema.Description{Text: ""},
Name: "upsert",
Type: inputTypeName,
},
}...)
mutation.Fields = append(mutation.Fields, &schema.Field{
Name: singularName,
Args: mutationArgs,
Type: outputType,
})
mutation.Fields = append(mutation.Fields, &schema.Field{
Name: pluralName,
Args: append(mutationArgs, schema.InputValueList{
&schema.InputValue{
Desc: schema.Description{Text: ""},
Name: "inserts",
Type: pluralInputTypeName,
},
&schema.InputValue{
Desc: schema.Description{Text: ""},
Name: "updates",
Type: pluralInputTypeName,
},
&schema.InputValue{
Desc: schema.Description{Text: ""},
Name: "upserts",
Type: pluralInputTypeName,
},
}...),
Type: outputType,
})
}
for typeName := range scalarExpressionTypesNeeded {
expressionType := &schema.InputObject{
Name: typeName + "Expression",
Fields: schema.InputValueList{
&schema.InputValue{
Name: "eq",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}},
},
&schema.InputValue{
Name: "equals",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}},
},
&schema.InputValue{
Name: "neq",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}},
},
&schema.InputValue{
Name: "not_equals",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}},
},
&schema.InputValue{
Name: "gt",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}},
},
&schema.InputValue{
Name: "greater_than",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}},
},
&schema.InputValue{
Name: "lt",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}},
},
&schema.InputValue{
Name: "lesser_than",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}},
},
&schema.InputValue{
Name: "gte",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}},
},
&schema.InputValue{
Name: "greater_or_equals",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}},
},
&schema.InputValue{
Name: "lte",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}},
},
&schema.InputValue{
Name: "lesser_or_equals",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}},
},
&schema.InputValue{
Name: "in",
Type: &schema.NonNull{OfType: &schema.List{OfType: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}}}},
},
&schema.InputValue{
Name: "nin",
Type: &schema.NonNull{OfType: &schema.List{OfType: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}}}},
},
&schema.InputValue{
Name: "not_in",
Type: &schema.NonNull{OfType: &schema.List{OfType: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}}}},
},
&schema.InputValue{
Name: "like",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "String"}},
},
&schema.InputValue{
Name: "nlike",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "String"}},
},
&schema.InputValue{
Name: "not_like",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "String"}},
},
&schema.InputValue{
Name: "ilike",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "String"}},
},
&schema.InputValue{
Name: "nilike",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "String"}},
},
&schema.InputValue{
Name: "not_ilike",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "String"}},
},
&schema.InputValue{
Name: "similar",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "String"}},
},
&schema.InputValue{
Name: "nsimilar",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "String"}},
},
&schema.InputValue{
Name: "not_similar",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "String"}},
},
&schema.InputValue{
Name: "has_key",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}},
},
&schema.InputValue{
Name: "has_key_any",
Type: &schema.NonNull{OfType: &schema.List{OfType: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}}}},
},
&schema.InputValue{
Name: "has_key_all",
Type: &schema.NonNull{OfType: &schema.List{OfType: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}}}},
},
&schema.InputValue{
Name: "contains",
Type: &schema.NonNull{OfType: &schema.List{OfType: &schema.NonNull{OfType: &schema.TypeName{Name: typeName}}}},
},
&schema.InputValue{
Name: "contained_in",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "String"}},
},
&schema.InputValue{
Name: "is_null",
Type: &schema.NonNull{OfType: &schema.TypeName{Name: "Boolean"}},
},
},
}
engineSchema.Types[expressionType.Name] = expressionType
}
if err := engineSchema.ResolveTypes(); err != nil {
return err
}
engine.Resolver = resolvers.Func(func(request *resolvers.ResolveRequest, next resolvers.Resolution) resolvers.Resolution {
resolver := resolvers.MetadataResolver.Resolve(request, next)
if resolver != nil {
return resolver
}
resolver = resolvers.MethodResolver.Resolve(request, next) // needed by the MetadataResolver
if resolver != nil {
return resolver
}
return nil
})
sg.ge = engine
return nil
}

View File

@ -1,169 +0,0 @@
package core
import (
"bytes"
"database/sql"
"fmt"
"hash/maphash"
"io"
"strings"
"sync"
"github.com/dosco/super-graph/core/internal/allow"
"github.com/dosco/super-graph/core/internal/qcode"
)
type query struct {
sync.Once
sd *sql.Stmt
ai allow.Item
qt qcode.QType
err error
st stmt
roleArg bool
}
func (sg *SuperGraph) prepare(q *query, role string) {
var stmts []stmt
var err error
qb := []byte(q.ai.Query)
vars := []byte(q.ai.Vars)
switch q.qt {
case qcode.QTQuery:
if sg.abacEnabled {
stmts, err = sg.buildMultiStmt(qb, vars)
} else {
stmts, err = sg.buildRoleStmt(qb, vars, role)
}
case qcode.QTMutation:
stmts, err = sg.buildRoleStmt(qb, vars, role)
}
if err != nil {
sg.log.Printf("WRN %s %s: %v", q.qt, q.ai.Name, err)
}
q.st = stmts[0]
q.roleArg = len(stmts) > 1
q.sd, err = sg.db.Prepare(q.st.sql)
if err != nil {
q.err = fmt.Errorf("prepare failed: %v: %s", err, q.st.sql)
}
}
func (sg *SuperGraph) initPrepared() error {
if sg.allowList.IsPersist() {
return nil
}
if err := sg.prepareRoleStmt(); err != nil {
return fmt.Errorf("role query: %w", err)
}
sg.queries = make(map[uint64]*query)
list, err := sg.allowList.Load()
if err != nil {
return err
}
h := maphash.Hash{}
h.SetSeed(sg.hashSeed)
for _, v := range list {
if v.Query == "" {
continue
}
qt := qcode.GetQType(v.Query)
switch qt {
case qcode.QTQuery:
sg.queries[queryID(&h, v.Name, "user")] = &query{ai: v, qt: qt}
sg.queries[queryID(&h, v.Name, "anon")] = &query{ai: v, qt: qt}
case qcode.QTMutation:
for _, role := range sg.conf.Roles {
sg.queries[queryID(&h, v.Name, role.Name)] = &query{ai: v, qt: qt}
}
}
}
return nil
}
// nolint: errcheck
func (sg *SuperGraph) prepareRoleStmt() error {
var err error
if !sg.abacEnabled {
return nil
}
rq := strings.ReplaceAll(sg.conf.RolesQuery, "$user_id", "$1")
w := &bytes.Buffer{}
io.WriteString(w, `SELECT (CASE WHEN EXISTS (`)
io.WriteString(w, rq)
io.WriteString(w, `) THEN `)
io.WriteString(w, `(SELECT (CASE`)
for _, role := range sg.conf.Roles {
if role.Match == "" {
continue
}
io.WriteString(w, ` WHEN `)
io.WriteString(w, role.Match)
io.WriteString(w, ` THEN '`)
io.WriteString(w, role.Name)
io.WriteString(w, `'`)
}
io.WriteString(w, ` ELSE $2 END) FROM (`)
io.WriteString(w, rq)
io.WriteString(w, `) AS "_sg_auth_roles_query" LIMIT 1) `)
io.WriteString(w, `ELSE 'anon' END) FROM (VALUES (1)) AS "_sg_auth_filler" LIMIT 1; `)
sg.getRole, err = sg.db.Prepare(w.String())
if err != nil {
return err
}
return nil
}
func (sg *SuperGraph) initAllowList() error {
var ac allow.Config
var err error
if sg.conf.AllowListFile == "" {
sg.conf.AllowListFile = "allow.list"
}
// When list is not eabled it is still created and
// and new queries are saved to it.
if !sg.conf.UseAllowList {
ac = allow.Config{CreateIfNotExists: true, Persist: true, Log: sg.log}
}
sg.allowList, err = allow.New(sg.conf.AllowListFile, ac)
if err != nil {
return fmt.Errorf("failed to initialize allow list: %w", err)
}
return nil
}
// nolint: errcheck
func queryID(h *maphash.Hash, name, role string) uint64 {
h.WriteString(name)
h.WriteString(role)
v := h.Sum64()
h.Reset()
return v
}

View File

@ -1,13 +0,0 @@
package core
import "hash/maphash"
// nolint: errcheck
func mkkey(h *maphash.Hash, k1, k2 string) uint64 {
h.WriteString(k1)
h.WriteString(k2)
v := h.Sum64()
h.Reset()
return v
}

View File

@ -1,10 +1,7 @@
version: '3.4'
services:
db:
image: postgres:12
environment:
POSTGRES_USER: postgres
POSTGRES_PASSWORD: postgres
image: postgres
ports:
- "5432:5432"

View File

@ -1,5 +1,5 @@
<template>
<div class="shadow p-4 flex items-start" :class="className">
<div class="shadow bg-white p-4 flex items-start" :class="className">
<slot name="image"></slot>
<div class="pl-4">
<h2 class="p-0">

View File

@ -0,0 +1,267 @@
<template>
<div>
<main aria-labelledby="main-title" >
<Navbar />
<div class="container mx-auto">
<div class="flex flex-col md:flex-row justify-between px-10 md:px-20">
<div class="bg-bottom bg-no-repeat bg-cover">
<div class="text-center md:text-left pt-24">
<h1 v-if="data.heroText !== null" class="text-5xl font-bold text-black pb-0 uppercase">
<img src="/super-graph.png" width="250" />
</h1>
<p class="text-4xl text-gray-800 leading-tight mt-1">
Build web products faster. Secure high performance GraphQL
</p>
<NavLink
class="inline-block px-4 py-3 my-8 bg-blue-600 text-blue-100 font-bold rounded"
:item="actionLink"
/>
<a
class="px-4 py-3 my-8 border-2 border-gray-500 text-gray-600 font-bold rounded"
href="https://github.com/dosco/super-graph"
target="_blank"
>Github</a>
</div>
</div>
<div class="py-10 md:p-20">
<img src="/hologram.svg" class="h-64">
</div>
</div>
</div>
<div>
<div
class="flex flex-wrap mx-2 md:mx-20"
v-if="data.features && data.features.length"
>
<div
class="w-2/4 md:w-1/3 shadow"
v-for="(feature, index) in data.features"
:key="index"
>
<div class="p-8">
<h2 class="md:text-xl text-blue-800 font-medium border-0 mb-1">{{ feature.title }}</h2>
<p class="md:text-xl text-gray-700 leading-snug">{{ feature.details }}</p>
</div>
</div>
</div>
</div>
<div class="bg-gray-100 mt-10">
<div class="container mx-auto px-10 md:px-0 py-32">
<div class="pb-8 hidden md:block ">
<img src="arch-basic.svg">
</div>
<h1 class="uppercase font-semibold text-xl text-blue-800 mb-4">
What is {{ data.heroText }}?
</h1>
<div class="text-2xl md:text-3xl">
Super Graph can automatically learn a Postgres database and instantly serve it as a fast and secured GraphQL API. It comes with tools to create a new app and manage it's database. You get it all, a very productive developer and a highly scalable app backend. It's designed to work well on serverless platforms by Google, AWS, Microsoft, etc. The goal is to save you a ton of time and money so you can focus on you're apps core value.
</div>
</div>
</div>
<div class="flex flex-wrap">
<div class="md:w-2/4">
<img src="/graphql.png">
</div>
<div class="md:w-2/4">
<img src="/json.png">
</div>
</div>
<div class="mt-10 py-10 md:py-20">
<div class="container mx-auto px-10 md:px-0">
<h1 class="uppercase font-semibold text-xl text-blue-800 mb-4">
How to use {{ data.heroText }}?
</h1>
<div class="text-2xl md:text-3xl">
<small class="text-sm">Use the below command to download and install Super Graph. You will need Go 1.13 or above</small>
<pre>&#8227; GO111MODULE=on go get -u github.com/dosco/super-graph</pre>
<small class="text-sm">Create a new app and change to it's directory</small>
<pre>&#8227; super-graph new blog; cd blog</pre>
<small class="text-sm">Setup the app database and seed it with fake data. Docker compose will start a Postgres database for your app</small>
<pre>&#8227; docker-compose run blog_api ./super-graph db:setup</pre>
<small class="text-sm">And finally launch Super Graph configured for your app</small>
<pre>&#8227; docker-compose up</pre>
</div>
</div>
</div>
<div class="bg-gray-100 mt-10">
<div class="container mx-auto px-10 md:px-0 py-32">
<h1 class="uppercase font-semibold text-xl text-blue-800 mb-4">
The story of {{ data.heroText }}
</h1>
<div class="text-2xl md:text-3xl">
After working on several products through my career I find that we spend way too much time on building API backends. Most APIs also require constant updating, this costs real time and money.<br><br>
It's always the same thing, figure out what the UI needs then build an endpoint for it. Most API code involves struggling with an ORM to query a database and mangle the data into a shape that the UI expects to see.<br><br>
I didn't want to write this code anymore, I wanted the computer to do it. Enter GraphQL, to me it sounded great, but it still required me to write all the same database query code.<br><br>
Having worked with compilers before I saw this as a compiler problem. Why not build a compiler that converts GraphQL to highly efficient SQL.<br><br>
This compiler is what sits at the heart of Super Graph with layers of useful functionality around it like authentication, remote joins, rails integration, database migrations and everything else needed for you to build production ready apps with it.
</div>
</div>
</div>
<div class="overflow-hidden bg-indigo-900">
<div class="container mx-auto py-20">
<img src="/super-graph-web-ui.png">
</div>
</div>
<div class="py-10 md:py-20">
<div class="container mx-auto px-10 md:px-0">
<h1 class="uppercase font-semibold text-xl text-blue-800 mb-4">
Try it with a demo Rails app
</h1>
<div class="text-2xl md:text-3xl">
<small class="text-sm">Download the Docker compose config for the demo</small>
<pre>&#8227; curl -L -o demo.yml https://bit.ly/2mq05lW</pre>
<small class="text-sm">Setup the demo database</small>
<pre>&#8227; docker-compose -f demo.yml run rails_app rake db:create db:migrate db:seed</pre>
<small class="text-sm">Run the demo</small>
<pre>&#8227; docker-compose -f demo.yml up</pre>
<small class="text-sm">Signin to the demo app (user1@demo.com / 123456)</small>
<pre>&#8227; open http://localhost:3000</pre>
<small class="text-sm">Try the super graph web ui</small>
<pre>&#8227; open http://localhost:8080</pre>
</div>
</div>
</div>
<div class="border-t py-10">
<div class="block md:hidden w-100">
<iframe src='https://www.youtube.com/embed/MfPL2A-DAJk' frameborder='0' allowfullscreen style="width: 100%; height: 250px;">
</iframe>
</div>
<div class="container mx-auto flex flex-col md:flex-row items-center">
<div class="w-100 md:w-1/2 p-8">
<h1 class="text-2xl font-bold">GraphQL the future of APIs</h1>
<p class="text-xl text-gray-600">Keeping a tight and fast development loop helps you iterate quickly. Leveraging technology like Super Graph focuses your team on building the core product and not reinventing wheels. GraphQL eliminate the dependency on the backend engineering and keeps the things moving fast</p>
</div>
<div class="hidden md:block md:w-1/2">
<style>.embed-container { position: relative; padding-bottom: 56.25%; height: 0; overflow: hidden; max-width: 100%; } .embed-container iframe, .embed-container object, .embed-container embed { position: absolute; top: 0; left: 0; width: 100%; height: 100%; }</style>
<div class="embed-container shadow">
<iframe src='https://www.youtube.com/embed/MfPL2A-DAJk' frameborder='0' allowfullscreen >
</iframe>
</div>
</div>
</div>
</div>
<div class="bg-gray-200 mt-10">
<div class="container mx-auto px-10 md:px-0 py-32">
<h1 class="uppercase font-semibold text-xl text-blue-800 mb-4">
Build Secure Apps
</h1>
<div class="flex flex-col text-2xl md:text-3xl">
<card className="mb-1 p-8">
<template #image><font-awesome-icon icon="portrait" class="text-red-500" /></template>
<template #title>Role Based Access Control</template>
<template #body>Dynamically assign roles like admin, manager or anon to specific users. Generate role specific queries at runtime. For example admins can get all users while others can only fetch their own user.</template>
</card>
<card className="mb-1 p-8">
<template #image><font-awesome-icon icon="shield-alt" class="text-blue-500" /></template>
<template #title>Prepared Statements</template>
<template #body>An additional layer of protection from a variety of security issues like SQL injection. In production mode all queries are precompiled into prepared statements so only those can be executed. This also significantly speeds up all queries.</template>
</card>
<card className="p-8">
<template #image><font-awesome-icon icon="lock" class="text-green-500"/></template>
<template #title>Fuzz Tested Code</template>
<template #body>Fuzzing is done by complex software that generates massives amounts of random input to detect if code is free of security bugs. Google uses fuzzing to protects everything from their cloud infrastructure to the Chrome browser.</template>
</card>
</div>
</div>
</div>
<div class="">
<div class="container mx-auto px-10 md:px-0 py-32">
<h1 class="uppercase font-semibold text-xl text-blue-800 mb-4">
More Features
</h1>
<div class="flex flex-col md:flex-row text-2xl md:text-3xl">
<card className="mr-0 md:mr-1 mb-1 flex-col w-100 md:w-1/3 items-center">
<template #image><img src="/arch-remote-join.svg" class="h-64"></template>
<template #title>Remote Joins</template>
<template #body>A powerful feature that allows you to query your database and remote REST APIs at the same time. For example fetch a user from the DB, his tweets from Twitter and his payments from Stripe with a single GraphQL query.</template>
</card>
<card className="mr-0 md:mr-1 mb-1 flex-col w-100 md:w-1/3">
<template #image><img src="/arch-search.svg" class="h-64"></template>
<template #title>Full Text Search</template>
<template #body>Postgres has excellent full-text search built-in. You don't need another expensive service. Super Graph makes it super easy to use with keyword ranking and highlighting also supported.</template>
</card>
<card className="mb-1 flex-col w-100 md:w-1/3">
<template #image><img src="/arch-bulk.svg" class="h-64"></template>
<template #title>Bulk Inserts</template>
<template #body>Efficiently insert, update and delete multiple items with a single query. Upserts are also supported</template>
</card>
</div>
</div>
</div>
<div
class="mx-auto text-center py-8"
v-if="data.footer"
>
{{ data.footer }}
</div>
</main>
</div>
</template>
<script>
import NavLink from '@theme/components/NavLink.vue'
import Navbar from '@theme/components/Navbar.vue'
import Card from './Card.vue'
import { library } from '@fortawesome/fontawesome-svg-core'
import { faPortrait, faShieldAlt, faLock } from '@fortawesome/free-solid-svg-icons'
import { FontAwesomeIcon } from '@fortawesome/vue-fontawesome'
library.add(faPortrait, faShieldAlt, faLock)
export default {
components: { NavLink, Navbar, FontAwesomeIcon, Card },
computed: {
data () {
return this.$page.frontmatter
},
actionLink () {
return {
link: this.data.actionLink,
text: this.data.actionText
}
}
}
}
</script>

View File

@ -1,6 +1,6 @@
let ogprefix = 'og: http://ogp.me/ns#'
let title = 'Super Graph'
let description = 'Fetch data without code'
let description = 'An instant GraphQL API for your app. No code needed.'
let color = '#f42525'
module.exports = {
@ -15,7 +15,7 @@ module.exports = {
{ text: 'Internals', link: '/internals' },
{ text: 'Github', link: 'https://github.com/dosco/super-graph' },
{ text: 'Docker', link: 'https://hub.docker.com/r/dosco/super-graph/builds' },
{ text: 'Join Chat', link: 'https://discord.com/invite/23Wh7c' },
{ text: 'Join Chat', link: 'https://discord.gg/NKdXBc' },
],
serviceWorker: {
updatePopup: true

View File

Before

Width:  |  Height:  |  Size: 5.4 KiB

After

Width:  |  Height:  |  Size: 5.4 KiB

View File

Before

Width:  |  Height:  |  Size: 2.3 KiB

After

Width:  |  Height:  |  Size: 2.3 KiB

View File

Before

Width:  |  Height:  |  Size: 5.4 KiB

After

Width:  |  Height:  |  Size: 5.4 KiB

View File

Before

Width:  |  Height:  |  Size: 1.8 KiB

After

Width:  |  Height:  |  Size: 1.8 KiB

View File

Before

Width:  |  Height:  |  Size: 21 KiB

After

Width:  |  Height:  |  Size: 21 KiB

View File

Before

Width:  |  Height:  |  Size: 66 KiB

After

Width:  |  Height:  |  Size: 66 KiB

View File

Before

Width:  |  Height:  |  Size: 7.4 KiB

After

Width:  |  Height:  |  Size: 7.4 KiB

View File

Before

Width:  |  Height:  |  Size: 149 KiB

After

Width:  |  Height:  |  Size: 149 KiB

View File

Before

Width:  |  Height:  |  Size: 1.8 KiB

After

Width:  |  Height:  |  Size: 1.8 KiB

View File

Before

Width:  |  Height:  |  Size: 146 KiB

After

Width:  |  Height:  |  Size: 146 KiB

View File

Before

Width:  |  Height:  |  Size: 911 KiB

After

Width:  |  Height:  |  Size: 911 KiB

View File

Before

Width:  |  Height:  |  Size: 14 KiB

After

Width:  |  Height:  |  Size: 14 KiB

View File

Before

Width:  |  Height:  |  Size: 42 KiB

After

Width:  |  Height:  |  Size: 42 KiB

View File

@ -1,9 +1,6 @@
@tailwind base;
@css {
body, .navbar, .navbar .links {
@apply bg-white text-black border-0 !important;
}
h1 {
@apply font-semibold text-3xl border-0 py-4
}

View File

@ -10,7 +10,7 @@ longTagline: Get an instant high performance GraphQL API for Postgres. No code n
actionText: Get Started, Free, Open Source →
actionLink: /guide
description: Super Graph can automatically learn a Postgres database and instantly serve it as a fast and secured GraphQL API. It comes with tools to create a new app and manage it's database. You get it all, a very productive developer and a highly scalable app backend. It's designed to work well on serverless platforms by Google, AWS, Microsoft, etc. The goal is to save you a ton of time and money so you can focus on your apps core value.
description: Super Graph can automatically learn a Postgres database and instantly serve it as a fast and secured GraphQL API. It comes with tools to create a new app and manage it's database. You get it all, a very productive developer and a highly scalable app backend. It's designed to work well on serverless platforms by Google, AWS, Microsoft, etc. The goal is to save you a ton of time and money so you can focus on you're apps core value.
features:
- title: Simple
@ -19,8 +19,8 @@ features:
details: Compiles your GraphQL into a fast SQL query in realtime.
- title: Built in GO
details: Built in Go is a language created at Google to build fast and secure web services.
- title: In your own Code
details: Use as a library in your own GO code. Build faster save time and money.
- title: Ruby-on-Rails
details: Can read Rails cookies and supports rails database conventions.
- title: Serverless
details: Instant startup for scale to zero environments like Google Cloud Run, App Engine, AWS Lambda
- title: Free and Open Source

View File

@ -32,7 +32,7 @@ For this to work you have to ensure that the option `:domain => :all` is added t
### With an NGINX loadbalancer
If your infrastructure is fronted by NGINX then it should be configured so that all requests to your GraphQL API path are proxyed to Super Graph. In the example NGINX config below all requests to the path `/api/v1/graphql` are routed to wherever you have Super Graph installed within your architecture. This example is derived from the config file example at [/microservices-nginx-gateway/nginx.conf](https://github.com/launchany/microservices-nginx-gateway/blob/master/nginx.conf)
If you're infrastructure is fronted by NGINX then it should be configured so that all requests to your GraphQL API path are proxyed to Super Graph. In the example NGINX config below all requests to the path `/api/v1/graphql` are routed to wherever you have Super Graph installed within your architecture. This example is derived from the config file example at [/microservices-nginx-gateway/nginx.conf](https://github.com/launchany/microservices-nginx-gateway/blob/master/nginx.conf)
::: tip NGINX with sub-domain
Yes, NGINX is very flexible and you can configure it to keep Super Graph a subdomain instead of on the same top level domain. I'm sure a little Googleing will get you some great example configs for that.

View File

@ -4,9 +4,9 @@ sidebar: auto
# Guide to Super Graph
Super Graph is a service that instantly and without code gives you a high performance and secure GraphQL API. Your GraphQL queries are auto translated into a single fast SQL query. No more spending weeks or months writing backend API code. Just make the query you need and Super Graph will do the rest.
Super Graph is a micro-service that instantly and without code gives you a high performance and secure GraphQL API. Your GraphQL queries are auto translated into a single fast SQL query. No more writing API code as you develop your web frontend just make the query you need and Super Graph will do the rest.
Super Graph has a rich feature set like integrating with your existing Ruby on Rails apps, joining your DB with data from remote APIs, Role and Attribute based access control, Support for JWT tokens, DB migrations, seeding and a lot more.
Super Graph has a rich feature set like integrating with your existing Ruby on Rails apps, joining your DB with data from remote APIs, Role and Attribute based access control, Supoport for JWT tokens, DB migrations, seeding and a lot more.
## Features
@ -31,20 +31,14 @@ Super Graph has a rich feature set like integrating with your existing Ruby on R
## Try the demo app
```bash
# clone the repository
git clone https://github.com/dosco/super-graph
# run db in background
docker-compose up -d db
# see logs and wait until DB is really UP
docker-compose logs db
# download the Docker compose config for the demo
curl -L -o demo.yml https://bit.ly/2mq05lW
# setup the demo rails app & database and run it
docker-compose run rails_app rake db:create db:migrate db:seed
docker-compose -f demo.yml run rails_app rake db:create db:migrate db:seed
# run the demo
docker-compose up
docker-compose -f demo.yml up
# signin to the demo app (user1@demo.com / 123456)
open http://localhost:3000
@ -53,14 +47,14 @@ open http://localhost:3000
open http://localhost:8080
```
::: tip DEMO REQUIREMENTS
::: warning DEMO REQUIREMENTS
This demo requires `docker` you can either install it using `brew` or from the
docker website [https://docs.docker.com/docker-for-mac/install/](https://docs.docker.com/docker-for-mac/install/)
:::
#### Trying out GraphQL
We fully support queries and mutations. For example the below GraphQL query would fetch two products that belong to the current user where the price is greater than 10.
We currently fully support queries and mutations. Support for `subscriptions` is work in progress. For example the below GraphQL query would fetch two products that belong to the current user where the price is greater than 10.
#### GQL Query
@ -82,6 +76,32 @@ query {
}
```
In another example the below GraphQL mutation would insert a product into the database. The first part of the below example is the variable data and the second half is the GraphQL mutation. For mutations data has to always ben passed as a variable.
```json
{
"data": {
"name": "Art of Computer Programming",
"description": "The Art of Computer Programming (TAOCP) is a comprehensive monograph written by computer scientist Donald Knuth",
"price": 30.5
}
}
```
```graphql
mutation {
product(insert: $data) {
id
name
}
}
```
The above GraphQL query returns the JSON result below. It handles all
kinds of complexity without you having to writing a line of code.
For example there is a while greater than `gt` and a limit clause on a child field. And the `avatar` field is renamed to `picture`. The `password` field is blocked and not returned. Finally the relationship between the `users` table and the `products` table is auto discovered and used.
#### JSON Result
```json
@ -108,107 +128,19 @@ query {
}
```
::: tip Testing with a user
In development mode you can use the `X-User-ID: 4` header to set a user id so you don't have to worries about cookies etc. This can be set using the *HTTP Headers* tab at the bottom of the web UI.
:::
#### Try with an authenticated user
In another example the below GraphQL mutation would insert a product into the database. The first part of the below example is the variable data and the second half is the GraphQL mutation. For mutations data has to always ben passed as a variable.
In development mode you can use the `X-User-ID: 4` header to set a user id so you don't have to worries about cookies etc. This can be set using the *HTTP Headers* tab at the bottom of the web UI you'll see when you visit the above link. You can also directly run queries from the commandline like below.
```json
{
"data": {
"name": "Art of Computer Programming",
"description": "The Art of Computer Programming (TAOCP) is a comprehensive monograph written by computer scientist Donald Knuth",
"price": 30.5
}
}
```
#### Querying the GQL endpoint
```graphql
mutation {
product(insert: $data) {
id
name
}
}
```
```bash
## Why Super Graph
Let's take a simple example say you want to fetch 5 products priced over 12 dollars along with the photos of the products and the users that owns them. Additionally also fetch the last 10 of your own purchases along with the name and ID of the product you purchased. This is a common type of query to render a view in say an ecommerce app. Lets be honest it's not very exciting write and maintain. Keep in mind the data needed will only continue to grow and change as your app evolves. Developers might find that most ORMs will not be able to do all of this in a single SQL query and will require n+1 queries to fetch all the data and assembly it into the right JSON response.
What if I told you Super Graph will fetch all this data with a single SQL query and without you having to write a single line of code. Also as your app evolves feel free to evolve the query as you like. In our experience Super Graph saves us hundreds or thousands of man hours that we can put towards the more exciting parts of our app.
#### GraphQL Query
```graphql
query {
products(limit: 5, where: { price: { gt: 12 } }) {
id
name
description
price
photos {
url
}
user {
id
email
picture : avatar
full_name
}
}
purchases(
limit: 10,
order_by: { created_at: desc } ,
where: { user_id: { eq: $user_id } }
) {
id
created_at
product {
id
name
}
}
}
```
#### JSON Result
```json
"data": {
"products": [
{
"id": 1,
"name": "Oaked Arrogant Bastard Ale",
"description": "Coors lite, European Amber Lager, Perle, 1272 - American Ale II, 38 IBU, 6.4%, 9.7°Blg",
"price": 20,
"photos: [{
"url": "https://www.scienceworld.ca/wp-content/uploads/science-world-beer-flavours.jpg"
}],
"user": {
"id": 1,
"email": "user0@demo.com",
"picture": "https://robohash.org/sitaliquamquaerat.png?size=300x300&set=set1",
"full_name": "Mrs. Wilhemina Hilpert"
}
},
...
]
},
"purchases": [
{
"id": 5,
"created_at": "2020-01-24T05:34:39.880599",
"product": {
"id": 45,
"name": "Brooklyn Black",
}
},
...
]
}
# fetch the response json directly from the endpoint using user id 5
curl 'http://localhost:8080/api/v1/graphql' \
-H 'content-type: application/json' \
-H 'X-User-ID: 5' \
--data-binary '{"query":"{ products { name price users { email }}}"}'
```
## Get Started
@ -222,7 +154,7 @@ You can then add your database schema to the migrations, maybe create some seed
git clone https://github.com/dosco/super-graph && cd super-graph && make install
```
And then create and launch your new app
And then create and launch you're new app
```bash
# create a new app and change to it's directory
@ -292,12 +224,6 @@ for (i = 0; i < 10; i++) {
}
```
If you want to import a lot of data using a CSV file is the best and fastest option. The `import_csv` command uses the `COPY FROM` Postgres method to load massive amounts of data into tables. The first line of the CSV file must be the header with column names.
```javascript
var post_count = import_csv("posts", "posts.csv")
```
You can generate the following fake data for your seeding purposes. Below is the list of fake data functions supported by the built-in fake data library. For example `fake.image_url()` will generate a fake image url or `fake.shuffle_strings(['hello', 'world', 'cool'])` will generate a randomly shuffled version of that array of strings or `fake.rand_string(['hello', 'world', 'cool'])` will return a random string from the array provided.
```
@ -347,10 +273,12 @@ beer_style
beer_yeast
// Cars
car
car_type
vehicle
vehicle_type
car_maker
car_model
fuel_type
transmission_gear_type
// Text
word
@ -436,8 +364,8 @@ hipster_paragraph
hipster_sentence
// File
file_extension
file_mine_type
extension
mine_type
// Numbers
number
@ -461,18 +389,11 @@ mac_address
digit
letter
lexify
rand_string
shuffle_strings
numerify
```
Other utility functions
```
shuffle_strings(string_array)
make_slug(text)
make_slug_lang(text, lang)
```
### Migrations
Easy database migrations is the most important thing when building products backend by a relational database. We make it super easy to manage and migrate your database.
@ -730,31 +651,7 @@ query {
}
```
### Custom Functions
Any function defined in the database like the below `add_five` that adds 5 to any number given to it can be used
within your query. The one limitation is that it should be a function that only accepts a single argument. The function is used within you're GraphQL in similar way to how aggregrations are used above. Example below
```grahql
query {
thread(id: 5) {
id
total_votes
add_five_total_votes
}
}
```
Postgres user-defined function `add_five`
```
CREATE OR REPLACE FUNCTION add_five(a integer) RETURNS integer AS $$
BEGIN
RETURN a + 5;
END;
$$ LANGUAGE plpgsql;
```
## Mutations
In GraphQL mutations is the operation type for when you need to modify data. Super Graph supports the `insert`, `update`, `upsert` and `delete`. You can also do complex nested inserts and updates.
@ -939,6 +836,8 @@ mutation {
}
```
## Nested Mutations
Often you will need to create or update multiple related items at the same time. This can be done using nested mutations. For example you might need to create a product and assign it to a user, or create a user and his products at the same time. You just have to use simple json to define you mutation and Super Graph takes care of the rest.
### Nested Insert
@ -1007,7 +906,7 @@ mutation {
}
```
### Nested Update
### Nested Updates
Update a product item first and then assign it to a user
@ -1067,116 +966,9 @@ mutation {
}
```
### Pagination
## Using variables
This is a must have feature of any API. When you want your users to go through a list page by page or implement some fancy infinite scroll you're going to need pagination. There are two ways to paginate in Super Graph.
Limit-Offset
This is simple enough but also inefficient when working with a large number of total items. Limit, limits the number of items fetched and offset is the point you want to fetch from. The below query will fetch 10 results at a time starting with the 100th item. You will have to keep updating offset (110, 120, 130, etc ) to walk thought the results so make offset a variable.
```graphql
query {
products(limit: 10, offset: 100) {
id
slug
name
}
}
```
#### Cursor
This is a powerful and highly efficient way to paginate a large number of results. Infact it does not matter how many total results there are this will always be lighting fast. You can use a cursor to walk forward or backward through the results. If you plan to implement infinite scroll this is the option you should choose.
When going this route the results will contain a cursor value this is an encrypted string that you don't have to worry about just pass this back in to the next API call and you'll received the next set of results. The cursor value is encrypted since its contents should only matter to Super Graph and not the client. Also since the primary key is used for this feature it's possible you might not want to leak it's value to clients.
You will need to set this config value to ensure the encrypted cursor data is secure. If not set a random value is used which will change with each deployment breaking older cursor values that clients might be using so best to set it.
```yaml
# Secret key for general encryption operations like
# encrypting the cursor data
secret_key: supercalifajalistics
```
Paginating forward through your results
```json
{
"variables": {
"cursor": "MJoTLbQF4l0GuoDsYmCrpjPeaaIlNpfm4uFU4PQ="
}
}
```
```graphql
query {
products(first: 10, after: $cursor) {
slug
name
}
}
```
Paginating backward through your results
```graphql
query {
products(last: 10, before: $cursor) {
slug
name
}
}
```
```graphql
"data": {
"products": [
{
"slug": "eius-nulla-et-8",
"name" "Pale Ale"
},
{
"slug": "sapiente-ut-alias-12",
"name" "Brown Ale"
}
...
],
"products_cursor": "dJwHassm5+d82rGydH2xQnwNxJ1dcj4/cxkh5Cer"
}
```
Nested tables can also have cursors. Requesting multiple cursors are supported on a single request but when paginating using a cursor only one table is currently supported. To explain this better, you can only use a `before` or `after` argument with a cursor value to paginate a single table in a query.
```graphql
query {
products(last: 10) {
slug
name
customers(last: 5) {
email
full_name
}
}
}
```
Multiple order-by arguments are supported. Super Graph is smart enough to allow cursor based pagination when you also need complex sort order like below.
```graphql
query {
products(
last: 10
before: $cursor
order_by: [ price: desc, total_customers: asc ]) {
slug
name
}
}
```
## Using Variables
Variables (`$product_id`) and their values (`"product_id": 5`) can be passed along side the GraphQL query. Using variables makes for better client side code as well as improved server side SQL query caching. The built-in web-ui also supports setting variables. Not having to manipulate your GraphQL query string to insert values into it makes for cleaner
Variables (`$product_id`) and their values (`"product_id": 5`) can be passed along side the GraphQL query. Using variables makes for better client side code as well as improved server side SQL query caching. The build-in web-ui also supports setting variables. Not having to manipulate your GraphQL query string to insert values into it makes for cleaner
and better client side code.
```javascript
@ -1196,104 +988,6 @@ fetch('http://localhost:8080/api/v1/graphql', {
.then(res => console.log(res.data));
```
## GraphQL with React
This is a quick simple example using `graphql.js` [https://github.com/f/graphql.js/](https://github.com/f/graphql.js/)
```js
import React, { useState, useEffect } from 'react'
import graphql from 'graphql.js'
// Create a GraphQL client pointing to Super Graph
var graph = graphql("http://localhost:3000/api/v1/graphql", { asJSON: true })
const App = () => {
const [user, setUser] = useState(null)
useEffect(() => {
async function action() {
// Use the GraphQL client to execute a graphQL query
// The second argument to the client are the variables you need to pass
const result = await graph(`{ user { id first_name last_name picture_url } }`)()
setUser(result)
}
action()
}, []);
return (
<div className="App">
<h1>{ JSON.stringify(user) }</h1>
</div>
);
}
```
export default App;
## Advanced Columns
The ablity to have `JSON/JSONB` and `Array` columns is often considered in the top most useful features of Postgres. There are many cases where using an array or a json column saves space and reduces complexity in your app. The only issue with these columns is the really that your SQL queries can get harder to write and maintain.
Super Graph steps in here to help you by supporting these columns right out of the box. It allows you to work with these columns just like you would with tables. Joining data against or modifying array columns using the `connect` or `disconnect` keywords in mutations is fully supported. Another very useful feature is the ability to treat `json` or `binary json (jsonb)` columns as seperate tables, even using them in nested queries joining against related tables. To replicate these features on your own will take a lot of complex SQL. Using Super Graph means you don't have to deal with any of this it just works.
### Array Columns
Configure a relationship between an array column `tag_ids` which contains integer id's for tags and the column `id` in the table `tags`.
```yaml
tables:
- name: posts
columns:
- name: tag_ids
related_to: tags.id
```
```graphql
query {
posts {
title
tags {
name
image
}
}
}
```
### JSON Column
Configure a JSON column called `tag_count` in the table `products` into a seperate table. This JSON column contains a json array of objects each with a tag id and a count of the number of times the tag was used. As a seperate table you can nest it into your GraphQL query and treat it like table using any of the standard features like `order_by`, `limit`, `where clauses`, etc.
The configuration below tells Super Graph to create a synthetic table called `tag_count` using the column `tag_count` from the `products` table. And that this new table has two columns `tag_id` and `count` of the listed types and with the defined relationships.
```yaml
tables:
- name: tag_count
table: products
columns:
- name: tag_id
type: bigint
related_to: tags.id
- name: count
type: int
```
```graphql
query {
products {
name
tag_counts {
count
tag {
name
}
}
}
}
```
## Full text search
Every app these days needs search. Enought his often means reaching for something heavy like Solr. While this will work why add complexity to your infrastructure when Postgres has really great
@ -1379,45 +1073,45 @@ class AddSearchColumn < ActiveRecord::Migration[5.1]
end
```
## API Security
## GraphQL with React
One of the the most common questions I get asked is what happens if a user out on the internet sends queries
that we don't want run. For example how do we stop him from fetching all users or the emails of users. Our answer to this is that it is not an issue as this cannot happen, let me explain.
This is a quick simple example using `graphql.js` [https://github.com/f/graphql.js/](https://github.com/f/graphql.js/)
Super Graph runs in one of two modes `development` or `production`, this is controlled via the config value `production: false` when it's false it's running in development mode and when true, production. In development mode all the **named** queries (including mutations) are saved to the allow list `./config/allow.list`. While in production mode when Super Graph starts only the queries from this allow list file are registered with the database as [prepared statements](https://stackoverflow.com/questions/8263371/how-can-prepared-statements-protect-from-sql-injection-attacks).
```js
import React, { useState, useEffect } from 'react'
import graphql from 'graphql.js'
Prepared statements are designed by databases to be fast and secure. They protect against all kinds of sql injection attacks and since they are pre-processed and pre-planned they are much faster to run then raw sql queries. Also there's no GraphQL to SQL compiling happening in production mode which makes your queries lighting fast as they are directly sent to the database with almost no overhead.
// Create a GraphQL client pointing to Super Graph
var graph = graphql("http://localhost:3000/api/v1/graphql", { asJSON: true })
In short in production only queries listed in the allow list file `./config/allow.list` can be used, all other queries will be blocked.
::: tip How to think about the allow list?
The allow list file is essentially a list of all your exposed API calls and the data that passes within them. It's very easy to build tooling to do things like parsing this file within your tests to ensure fields like `credit_card_no` are not accidently leaked. It's a great way to build compliance tooling and ensure your user data is always safe.
:::
This is an example of a named query, `getUserWithProducts` is the name you've given to this query it can be anything you like but should be unique across all you're queries. Only named queries are saved in the allow list in development mode.
```graphql
query getUserWithProducts {
users {
id
name
products {
id
name
price
const App = () => {
const [user, setUser] = useState(null)
useEffect(() => {
async function action() {
// Use the GraphQL client to execute a graphQL query
// The second argument to the client are the variables you need to pass
const result = await graph(`{ user { id first_name last_name picture_url } }`)()
setUser(result)
}
}
action()
}, []);
return (
<div className="App">
<h1>{ JSON.stringify(user) }</h1>
</div>
);
}
```
export default App;
## Authentication
You can only have one type of auth enabled either Rails or JWT.
You can only have one type of auth enabled. You can either pick Rails or JWT.
### Ruby on Rails
### Rails Auth (Devise / Warden)
Almost all Rails apps use Devise or Warden for authentication. Once the user is
authenticated a session is created with the users ID. The session can either be
@ -1469,7 +1163,7 @@ auth:
max_active: 12000
```
### JWT Tokens
### JWT Token Auth
```yaml
auth:
@ -1483,67 +1177,14 @@ auth:
public_key_type: ecdsa #rsa
```
For JWT tokens we currently support tokens from a provider like Auth0 or if you have a custom solution then we look for the `user_id` in the `subject` claim of of the `id token`. If you pick Auth0 then we derive two variables from the token `user_id` and `user_id_provider` for to use in your filters.
For JWT tokens we currently support tokens from a provider like Auth0
or if you have a custom solution then we look for the `user_id` in the
`subject` claim of of the `id token`. If you pick Auth0 then we derive two variables from the token `user_id` and `user_id_provider` for to use in your filters.
We can get the JWT token either from the `authorization` header where we expect it to be a `bearer` token or if `cookie` is specified then we look there.
For validation a `secret` or a public key (ecdsa or rsa) is required. When using public keys they have to be in a PEM format file.
### HTTP Headers
```yaml
header:
name: X-AppEngine-QueueName
exists: true
#value: default
```
Header auth is usually the best option to authenticate requests to the action endpoints. For example you
might want to use an action to refresh a materalized view every hour and only want a cron service like the Google AppEngine Cron service to make that request in this case a config similar to the one above will do.
The `exists: true` parameter ensures that only the existance of the header is checked not its value. The `value` parameter lets you confirm that the value matches the one assgined to the parameter. This helps in the case you are using a shared secret to protect the endpoint.
### Named Auth
```yaml
# You can add additional named auths to use with actions
# In this example actions using this auth can only be
# called from the Google Appengine Cron service that
# sets a special header to all it's requests
auths:
- name: from_taskqueue
type: header
header:
name: X-Appengine-Cron
exists: true
```
In addition to the default auth configuration you can create additional named auth configurations to be used
with features like `actions`. For example while your main GraphQL endpoint uses JWT for authentication you may want to use a header value to ensure your actions can only be called by clients having access to a shared secret
or security header.
## Actions
Actions is a very useful feature that is currently work in progress. For now the best use case for actions is to
refresh database tables like materialized views or call a database procedure to refresh a cache table, etc. An action creates an http endpoint that anyone can call to have the SQL query executed. The below example will create an endpoint `/api/v1/actions/refresh_leaderboard_users` any request send to that endpoint will cause the sql query to be executed. the `auth_name` points to a named auth that should be used to secure this endpoint. In future we have big plans to allow your own custom code to run using actions.
```yaml
actions:
- name: refresh_leaderboard_users
sql: REFRESH MATERIALIZED VIEW CONCURRENTLY "leaderboard_users"
auth_name: from_taskqueue
```
#### Using CURL to test a query
```bash
# fetch the response json directly from the endpoint using user id 5
curl 'http://localhost:8080/api/v1/graphql' \
-H 'content-type: application/json' \
-H 'X-User-ID: 5' \
--data-binary '{"query":"{ products { name price users { email }}}"}'
```
## Access Control
It's common for APIs to control what information they return or insert based on the role of the user. In Super Graph we have two primary roles `user` and `anon` the first for users where a `user_id` is available the latter for users where it's not.
@ -1556,6 +1197,7 @@ The `user` role can be divided up into further roles based on attributes in the
Super Graph allows you to create roles dynamically using a `roles_query` and ` match` config values.
### Configure RBAC
```yaml
@ -1595,7 +1237,7 @@ roles:
This configuration is relatively simple to follow the `roles_query` parameter is the query that must be run to help figure out a users role. This query can be as complex as you like and include joins with other tables.
The individual roles are defined under the `roles` parameter and this includes each table the role has a custom setting for. The role is dynamically matched using the `match` parameter for example in the above case `users.id = 1` means that when the `roles_query` is executed a user with the id `1` will be assigned the admin role and those that don't match get the `user` role if authenticated successfully or the `anon` role.
The individual roles are defined under the `roles` parameter and this includes each table the role has a custom setting for. The role is dynamically matched using the `match` parameter for example in the above case `users.id = 1` means that when the `roles_query` is executed a user with the id `1` willbe assigned the admin role and those that don't match get the `user` role if authenticated successfully or the `anon` role.
## Remote Joins
@ -1692,7 +1334,7 @@ tables:
```
## Configuration
## Configuration files
Configuration files can either be in YAML or JSON their names are derived from the `GO_ENV` variable, for example `GO_ENV=prod` will cause the `prod.yaml` config file to be used. or `GO_ENV=dev` will use the `dev.yaml`. A path to look for the config files in can be specified using the `-path <folder>` command line argument.
@ -1707,7 +1349,7 @@ app_name: "Super Graph Development"
host_port: 0.0.0.0:8080
web_ui: true
# debug, error, warn, info
# debug, info, warn, error, fatal, panic
log_level: "debug"
# enable or disable http compression (uses gzip)
@ -1735,7 +1377,7 @@ reload_on_config_change: true
# seed_file: seed.js
# Path pointing to where the migrations can be found
migrations_path: ./migrations
migrations_path: ./config/migrations
# Postgres related environment Variables
# SG_DATABASE_HOST
@ -1788,29 +1430,13 @@ auth:
# public_key_file: /secrets/public_key.pem
# public_key_type: ecdsa #rsa
# header:
# name: dnt
# exists: true
# value: localhost:8080
# You can add additional named auths to use with actions
# In this example actions using this auth can only be
# called from the Google Appengine Cron service that
# sets a special header to all it's requests
auths:
- name: from_taskqueue
type: header
header:
name: X-Appengine-Cron
exists: true
database:
type: postgres
host: db
port: 5432
dbname: app_development
user: postgres
password: postgres
password: ''
#schema: "public"
#pool_size: 10
@ -1821,48 +1447,18 @@ database:
# Enable this if you need the user id in triggers, etc
set_user_id: false
# database ping timeout is used for db health checking
ping_timeout: 1m
# Set up an secure tls encrypted db connection
enable_tls: false
# Required for tls. For example with Google Cloud SQL it's
# <gcp-project-id>:<cloud-sql-instance>"
# server_name: blah
# Required for tls. Can be a file path or the contents of the pem file
# server_cert: ./server-ca.pem
# Required for tls. Can be a file path or the contents of the pem file
# client_cert: ./client-cert.pem
# Required for tls. Can be a file path or the contents of the pem file
# client_key: ./client-key.pem
# Define additional variables here to be used with filters
variables:
admin_account_id: "5"
# Field and table names that you wish to block
blocklist:
- ar_internal_metadata
- schema_migrations
- secret
- password
- encrypted
- token
# Create custom actions with their own api endpoints
# For example the below action will be available at /api/v1/actions/refresh_leaderboard_users
# A request to this url will execute the configured SQL query
# which in this case refreshes a materialized view in the database.
# The auth_name is from one of the configured auths
actions:
- name: refresh_leaderboard_users
sql: REFRESH MATERIALIZED VIEW CONCURRENTLY "leaderboard_users"
auth_name: from_taskqueue
# Define additional variables here to be used with filters
variables:
admin_account_id: "5"
# Field and table names that you wish to block
blocklist:
- ar_internal_metadata
- schema_migrations
- secret
- password
- encrypted
- token
tables:
- name: customers
@ -1964,74 +1560,9 @@ SG_AUTH_RAILS_REDIS_PASSWORD
SG_AUTH_JWT_PUBLIC_KEY_FILE
```
## YugabyteDB
Yugabyte is an open-source, geo-distrubuted cloud-native relational DB that scales horizontally. Super Graph works with Yugabyte right out of the box. If you think you're data needs will outgrow Postgres and you don't really want to deal with sharding then Yugabyte is the way to go. Just point Super Graph to your Yugabyte DB and everything will just work including running migrations, seeding, querying, mutations, etc.
To use Yugabyte in your local development flow just uncomment the following lines in the `docker-compose.yml` file that is part of your Super Graph app. Also remember to comment out the originl postgres `db` config.
```yaml
# Postgres DB
# db:
# image: postgres:latest
# ports:
# - "5432:5432"
#Standard config to run a single node of Yugabyte
yb-master:
image: yugabytedb/yugabyte:latest
container_name: yb-master-n1
command: [ "/home/yugabyte/bin/yb-master",
"--fs_data_dirs=/mnt/disk0,/mnt/disk1",
"--master_addresses=yb-master-n1:7100",
"--replication_factor=1",
"--enable_ysql=true"]
ports:
- "7000:7000"
environment:
SERVICE_7000_NAME: yb-master
db:
image: yugabytedb/yugabyte:latest
container_name: yb-tserver-n1
command: [ "/home/yugabyte/bin/yb-tserver",
"--fs_data_dirs=/mnt/disk0,/mnt/disk1",
"--start_pgsql_proxy",
"--tserver_master_addrs=yb-master-n1:7100"]
ports:
- "9042:9042"
- "6379:6379"
- "5433:5433"
- "9000:9000"
environment:
SERVICE_5433_NAME: ysql
SERVICE_9042_NAME: ycql
SERVICE_6379_NAME: yedis
SERVICE_9000_NAME: yb-tserver
depends_on:
- yb-master
# Environment variables to point Super Graph to Yugabyte
# This is required since it uses a different user and port number
yourapp_api:
image: dosco/super-graph:latest
environment:
GO_ENV: "development"
Uncomment below for Yugabyte DB
SG_DATABASE_PORT: 5433
SG_DATABASE_USER: yugabyte
SG_DATABASE_PASSWORD: yugabyte
volumes:
- ./config:/config
ports:
- "8080:8080"
depends_on:
- db
```
## Developing Super Graph
If you want to build and run Super Graph from code then the below commands will build the web ui and launch Super Graph in developer mode with a watcher to rebuild on code changes. And the demo rails app is also launched to make it easier to test changes.
If you want to build and run Super Graph from code then the below commands will build the web ui and launch Super Graph in developer mode with a watcher to rebuild on code changes. And the demo rails app is also launched to make it essier to test changes.
```bash

View File

@ -1,431 +0,0 @@
<template>
<div>
<main aria-labelledby="main-title" >
<Navbar />
<div style="height: 3.6rem"></div>
<div class="container mx-auto pt-4">
<div class="text-center">
<div class="text-center text-3xl md:text-4xl text-black leading-tight font-semibold">
Fetch data without code
</div>
<NavLink
class="inline-block px-4 py-3 my-8 bg-blue-600 text-white font-bold rounded"
:item="actionLink"
/>
<a
class="px-4 py-3 my-8 border-2 border-blue-600 text-blue-600 font-bold rounded"
href="https://github.com/dosco/super-graph"
target="_blank"
>Github</a>
</div>
</div>
<div class="container mx-auto mb-8 mt-0 md:mt-20 bg-green-100">
<div class="flex flex-wrap">
<div class="w-100 md:w-1/2 border border-green-500 text-gray-6 00 text-sm md:text-lg p-6">
<div class="text-xl font-bold pb-4">Before, struggle with SQL</div>
<pre>
type User struct {
gorm.Model
Profile Profile
ProfileID int
}
type Profile struct {
gorm.Model
Name string
}
db.Model(&user).
Related(&profile).
Association("Languages").
Where("name in (?)", []string{"test"}).
Joins("left join emails on emails.user_id = users.id")
Find(&users)
and more ...
</pre>
</div>
<div class="w-100 md:w-1/2 border border-l md:border-l-0 border-green-500 text-blue-900 text-sm md:text-lg p-6">
<div class="text-xl font-bold pb-4">With Super Graph, just ask.</div>
<pre>
query {
user(id: 5) {
id
first_name
last_name
picture_url
posts(first: 20, order_by: { score: desc }) {
slug
title
created_at
votes_total
votes { created_at }
author { id name }
tags { id name }
}
posts_cursor
}
}
</pre>
</div>
</div>
</div>
<div class="mt-0 md:mt-20">
<div
class="flex flex-wrap mx-2 md:mx-20"
v-if="data.features && data.features.length"
>
<div
class="w-2/4 md:w-1/3 shadow"
v-for="(feature, index) in data.features"
:key="index"
>
<div class="p-8">
<h2 class="text-lg uppercase border-0">{{ feature.title }}</h2>
<div class="text-xl text-gray-900 leading-snug">{{ feature.details }}</div>
</div>
</div>
</div>
</div>
<div class="pt-0 md:pt-20">
<div class="container mx-auto p-10">
<div class="flex justify-center pb-20">
<img src="arch-basic.svg">
</div>
<div class="text-2xl md:text-3xl">
Super Graph is a library and service that fetches data from any Postgres database using just GraphQL. No more struggling with ORMs and SQL to wrangle data out of the database. No more having to figure out the right joins or making inefficient queries. However complex the GraphQL, Super Graph will always generate just one single efficient SQL query. The goal is to save you time and money so you can focus on you're apps core value.
</div>
</div>
</div>
<div class="pt-20">
<div class="container mx-auto px-10 md:px-0">
<h1 class="uppercase font-semibold text-2xl text-blue-800 text-center">
Try Super Graph
</h1>
<h1 class="uppercase font-semibold text-lg text-gray-800">
Deploy as a service using docker
</h1>
<div class="p-4 rounded bg-black text-white">
<pre>$ git clone https://github.com/dosco/super-graph && cd super-graph && make install</pre>
<pre>$ super-graph new blog; cd blog</pre>
<pre>$ docker-compose run blog_api ./super-graph db:setup</pre>
<pre>$ docker-compose up</pre>
</div>
<h1 class="uppercase font-semibold text-lg text-gray-800">
Or use it with your own code
</h1>
<div class="text-md">
<pre class="p-4 rounded bg-black text-white">
package main
import (
"database/sql"
"fmt"
"time"
"github.com/dosco/super-graph/config"
"github.com/dosco/super-graph/core"
_ "github.com/jackc/pgx/v4/stdlib"
)
func main() {
db, err := sql.Open("pgx", "postgres://postgrs:@localhost:5432/example_db")
if err != nil {
log.Fatal(err)
}
sg, err := core.NewSuperGraph(nil, db)
if err != nil {
log.Fatal(err)
}
graphqlQuery := `
query {
posts {
id
title
}
}`
res, err := sg.GraphQL(context.Background(), graphqlQuery, nil)
if err != nil {
log.Fatal(err)
}
fmt.Println(string(res.Data))
}
</pre>
</div>
</div>
</div>
<div class="pt-0 md:pt-20">
<div class="container mx-auto px-10 md:px-0 py-32">
<h1 class="uppercase font-semibold text-xl text-blue-800 mb-4">
The story of {{ data.heroText }}
</h1>
<div class="text-2xl md:text-3xl">
After working on several products through my career I find that we spend way too much time on building API backends. Most APIs also require constant updating, this costs real time and money.<br><br>
It's always the same thing, figure out what the UI needs then build an endpoint for it. Most API code involves struggling with an ORM to query a database and mangle the data into a shape that the UI expects to see.<br><br>
I didn't want to write this code anymore, I wanted the computer to do it. Enter GraphQL, to me it sounded great, but it still required me to write all the same database query code.<br><br>
Having worked with compilers before I saw this as a compiler problem. Why not build a compiler that converts GraphQL to highly efficient SQL.<br><br>
This compiler is what sits at the heart of Super Graph with layers of useful functionality around it like authentication, remote joins, rails integration, database migrations and everything else needed for you to build production ready apps with it.
</div>
</div>
</div>
<div class="overflow-hidden bg-indigo-900">
<div class="container mx-auto py-20">
<img src="/super-graph-web-ui.png">
</div>
</div>
<!--
<div class="py-10 md:py-20">
<div class="container mx-auto px-10 md:px-0">
<h1 class="uppercase font-semibold text-xl text-blue-800 mb-4">
Try it with a demo Rails app
</h1>
<div class="text-2xl md:text-3xl">
<small class="text-sm">Download the Docker compose config for the demo</small>
<pre>&#8227; curl -L -o demo.yml https://bit.ly/2FZS0uw</pre>
<small class="text-sm">Setup the demo database</small>
<pre>&#8227; docker-compose -f demo.yml run rails_app rake db:create db:migrate db:seed</pre>
<small class="text-sm">Run the demo</small>
<pre>&#8227; docker-compose -f demo.yml up</pre>
<small class="text-sm">Signin to the demo app (user1@demo.com / 123456)</small>
<pre>&#8227; open http://localhost:3000</pre>
<small class="text-sm">Try the super graph web ui</small>
<pre>&#8227; open http://localhost:8080</pre>
</div>
</div>
</div>
-->
<div class="pt-0 md:pt-20">
<div class="block md:hidden w-100">
<iframe src='https://www.youtube.com/embed/MfPL2A-DAJk' frameborder='0' allowfullscreen style="width: 100%; height: 250px;">
</iframe>
</div>
<div class="container mx-auto flex flex-col md:flex-row items-center">
<div class="w-100 md:w-1/2 p-8">
<h1 class="text-2xl font-bold">GraphQL the future of APIs</h1>
<p class="text-xl text-gray-600">Keeping a tight and fast development loop helps you iterate quickly. Leveraging technology like Super Graph focuses your team on building the core product and not reinventing wheels. GraphQL eliminate the dependency on the backend engineering and keeps the things moving fast</p>
</div>
<div class="hidden md:block md:w-1/2">
<style>.embed-container { position: relative; padding-bottom: 56.25%; height: 0; overflow: hidden; max-width: 100%; } .embed-container iframe, .embed-container object, .embed-container embed { position: absolute; top: 0; left: 0; width: 100%; height: 100%; }</style>
<div class="embed-container shadow">
<iframe src='https://www.youtube.com/embed/MfPL2A-DAJk' frameborder='0' allowfullscreen >
</iframe>
</div>
</div>
</div>
</div>
<div class="container mx-auto pt-0 md:pt-20">
<div class="flex flex-wrap bg-green-100">
<div class="w-100 md:w-1/2 border border-green-500 text-gray-6 00 text-sm md:text-lg p-6">
<div class="text-xl font-bold pb-4">No more joins joins, json, orms, just use GraphQL. Fetch all the data want in the structure you need.</div>
<pre>
query {
thread {
slug
title
published
createdAt : created_at
totalVotes : cached_votes_total
totalPosts : cached_posts_total
vote : thread_vote(where: { user_id: { eq: $user_id } }) {
created_at
}
topics {
slug
name
}
author : me {
slug
}
posts(first: 1, order_by: { score: desc }) {
slug
body
published
createdAt : created_at
totalVotes : cached_votes_total
totalComments : cached_comments_total
vote {
created_at
}
author : user {
slug
firstName : first_name
lastName : last_name
}
}
posts_cursor
}
}
</pre>
</div>
<div class="w-100 md:w-1/2 border border-l md:border-l-0 border-green-500 text-blue-900 text-sm md:text-lg p-6">
<div class="text-xl font-bold pb-4">Instant results using a single highly optimized SQL. It's just that simple.</div>
<pre>
{
"data": {
"thread": {
"slug": "eveniet-ex-24",
"vote": null,
"posts": [
{
"body": "Dolor laborum harum sed sit est ducimus temporibus velit non nobis repudiandae nobis suscipit commodi voluptatem debitis sed voluptas sequi officia.",
"slug": "illum-in-voluptas-1418",
"vote": null,
"author": {
"slug": "sigurd-kemmer",
"lastName": "Effertz",
"firstName": "Brandt"
},
"createdAt": "2020-04-07T04:22:42.115874+00:00",
"published": true,
"totalVotes": 0,
"totalComments": 2
}
],
"title": "In aut qui deleniti quia dolore quasi porro tenetur voluptatem ut adita alias fugit explicabo.",
"author": null,
"topics": [
{
"name": "CloudRun",
"slug": "cloud-run"
},
{
"name": "Postgres",
"slug": "postgres"
}
],
"createdAt": "2020-04-07T04:22:38.099482+00:00",
"published": true,
"totalPosts": 24,
"totalVotes": 0,
"posts_cursor": "mpeBl6L+QfJHc3cmLkLDj9pOdEZYTt5KQtLsazG3TLITB3hJhg=="
}
}
}
</pre>
</div>
</div>
</div>
<div class="pt-0 md:pt-20">
<div class="container mx-auto px-10 md:px-0 py-32">
<h1 class="uppercase font-semibold text-xl text-blue-800 mb-4">
Build Secure Apps
</h1>
<div class="flex flex-col text-2xl md:text-3xl">
<card className="mb-1 p-8">
<template #image><font-awesome-icon icon="portrait" class="text-red-500" /></template>
<template #title>Role Based Access Control</template>
<template #body>Dynamically assign roles like admin, manager or anon to specific users. Generate role specific queries at runtime. For example admins can get all users while others can only fetch their own user.</template>
</card>
<card className="mb-1 p-8">
<template #image><font-awesome-icon icon="shield-alt" class="text-blue-500" /></template>
<template #title>Prepared Statements</template>
<template #body>An additional layer of protection from a variety of security issues like SQL injection. In production mode all queries are precompiled into prepared statements so only those can be executed. This also significantly speeds up all queries.</template>
</card>
<card className="p-8">
<template #image><font-awesome-icon icon="lock" class="text-green-500"/></template>
<template #title>Fuzz Tested Code</template>
<template #body>Fuzzing is done by complex software that generates massives amounts of random input to detect if code is free of security bugs. Google uses fuzzing to protects everything from their cloud infrastructure to the Chrome browser.</template>
</card>
</div>
</div>
</div>
<div class="pt-0 md:py -20">
<div class="container mx-auto">
<h1 class="uppercase font-semibold text-xl text-blue-800 mb-4">
More Features
</h1>
<div class="flex flex-col md:flex-row text-2xl md:text-3xl">
<card className="mr-0 md:mr-1 mb-1 flex-col w-100 md:w-1/3 items-center">
<!-- <template #image><img src="/arch-remote-join.svg" class="h-64"></template> -->
<template #title>Remote Joins</template>
<template #body>A powerful feature that allows you to query your database and remote REST APIs at the same time. For example fetch a user from the DB, his tweets from Twitter and his payments from Stripe with a single GraphQL query.</template>
</card>
<card className="mr-0 md:mr-1 mb-1 flex-col w-100 md:w-1/3">
<!-- <template #image><img src="/arch-search.svg" class="h-64"></template> -->
<template #title>Full Text Search</template>
<template #body>Postgres has excellent full-text search built-in. You don't need another expensive service. Super Graph makes it super easy to use with keyword ranking and highlighting also supported.</template>
</card>
<card className="mb-1 flex-col w-100 md:w-1/3">
<!-- <template #image><img src="/arch-bulk.svg" class="h-64"></template> -->
<template #title>Bulk Inserts</template>
<template #body>Efficiently insert, update and delete multiple items with a single query. Upserts are also supported</template>
</card>
</div>
</div>
</div>
<div
class="mx-auto text-center py-8"
v-if="data.footer"
>
{{ data.footer }}
</div>
</main>
</div>
</template>
<script>
import NavLink from '@theme/components/NavLink.vue'
import Navbar from '@theme/components/Navbar.vue'
import Card from './Card.vue'
import { library } from '@fortawesome/fontawesome-svg-core'
import { faPortrait, faShieldAlt, faLock } from '@fortawesome/free-solid-svg-icons'
import { FontAwesomeIcon } from '@fortawesome/vue-fontawesome'
library.add(faPortrait, faShieldAlt, faLock)
export default {
components: { NavLink, Navbar, FontAwesomeIcon, Card },
computed: {
data () {
return this.$page.frontmatter
},
actionLink () {
return {
link: this.data.actionLink,
text: this.data.actionText
}
}
}
}
</script>

View File

@ -13,7 +13,7 @@ Super Graph code is made up of a number of packages. We have done our best to ke
## QCODE
This package contains the core of the GraphQL compiler it handling the lexing and parsing of the GraphQL query transforming it into an internal representation called
This package contains the core of the GraphQL conpiler it handling the lexing and parsing of the GraphQL query transforming it into an internal representation called
`QCode`.
This is the first step of the compiling process the `func NewCompiler(c Config)` function creates a new instance of this compiler which has it's own config.
@ -71,7 +71,7 @@ item{itemObjOpen, 16, 20} // {
...
```
These tokens are then fed into the parser `parse.go` the parser does the work of generating an abstract syntax tree (AST) from the tokens. This AST is an internal representation (data structure) and is not exposed outside the package. Since the AST is a tree a stack `stack.go` is used to walk the tree and generate the QCode AST. The QCode data structure is also a tree (represented as an array). This is then returned to the caller of the compile function.
These tokens are then fed into the parser `parse.go` the parser does the work of generating an abstract syntax tree (AST) from the tokens. This AST is an internal representation (data structure) and is not exposed outside the package. Sinc the AST is a tree a stack `stack.go` is used to walk the tree and generate the QCode AST. The QCode data structure is also a tree (represented as an array). This is then returned to the caller of the compile function.
```go
type Operation struct {
@ -238,4 +238,4 @@ ok github.com/dosco/super-graph/psql 2.530s
## Reach out
If you'd like me to explain other parts of the code please reach out over Twitter or Discord. I'll keep adding to this doc as I get time.
If you'd like me to explain other parts of the code please reach out over Twitter or Discord. I'll keep adding to this doc as I get time.

View File

@ -1,20 +0,0 @@
# Dependencies
/node_modules
# Production
/build
# Generated files
.docusaurus
.cache-loader
# Misc
.DS_Store
.env.local
.env.development.local
.env.test.local
.env.production.local
npm-debug.log*
yarn-debug.log*
yarn-error.log*

View File

@ -1,33 +0,0 @@
# Website
This website is built using [Docusaurus 2](https://v2.docusaurus.io/), a modern static website generator.
### Installation
```
$ yarn
```
### Local Development
```
$ yarn start
```
This command starts a local development server and open up a browser window. Most changes are reflected live without having to restart the server.
### Build
```
$ yarn build
```
This command generates static content into the `build` directory and can be served using any static contents hosting service.
### Deployment
```
$ GIT_USER=<Your GitHub username> USE_SSH=true yarn deploy
```
If you are using GitHub pages for hosting, this command is a convenient way to build the website and push to the `gh-pages` branch.

View File

@ -1,207 +0,0 @@
---
id: advanced
title: Full Text Search & More
sidebar_label: More Features
---
## Database Relationships
In most cases you don't need this configuration, Super Graph will discover and learn
the relationship graph within your database automatically. It does this using `Foreign Key` relationships that you have defined in your database schema.
The below configs are only needed in special cases such as when you don't use foreign keys or when you want to create a relationship between two tables where a foreign key is not defined or cannot be defined.
For example in the sample below a relationship is defined between the `tags` column on the `posts` table with the `slug` column on the `tags` table. This cannot be defined as using foreign keys since the `tags` column is of type array `text[]` and Postgres for one does not allow foreign keys with array columns.
```yaml
tables:
- name: posts
columns:
- name: tags
related_to: tags.slug
```
## Advanced Columns
The ablity to have `JSON/JSONB` and `Array` columns is often considered in the top most useful features of Postgres. There are many cases where using an array or a json column saves space and reduces complexity in your app. The only issue with these columns is the really that your SQL queries can get harder to write and maintain.
Super Graph steps in here to help you by supporting these columns right out of the box. It allows you to work with these columns just like you would with tables. Joining data against or modifying array columns using the `connect` or `disconnect` keywords in mutations is fully supported. Another very useful feature is the ability to treat `json` or `binary json (jsonb)` columns as seperate tables, even using them in nested queries joining against related tables. To replicate these features on your own will take a lot of complex SQL. Using Super Graph means you don't have to deal with any of this it just works.
### Array Columns
Configure a relationship between an array column `tag_ids` which contains integer id's for tags and the column `id` in the table `tags`.
```yaml
tables:
- name: posts
columns:
- name: tag_ids
related_to: tags.id
```
```graphql
query {
posts {
title
tags {
name
image
}
}
}
```
### JSON Column
Configure a JSON column called `tag_count` in the table `products` into a seperate table. This JSON column contains a json array of objects each with a tag id and a count of the number of times the tag was used. As a seperate table you can nest it into your GraphQL query and treat it like table using any of the standard features like `order_by`, `limit`, `where clauses`, etc.
The configuration below tells Super Graph to create a synthetic table called `tag_count` using the column `tag_count` from the `products` table. And that this new table has two columns `tag_id` and `count` of the listed types and with the defined relationships.
```yaml
tables:
- name: tag_count
table: products
columns:
- name: tag_id
type: bigint
related_to: tags.id
- name: count
type: int
```
```graphql
query {
products {
name
tag_counts {
count
tag {
name
}
}
}
}
```
## Remote Joins
It often happens that after fetching some data from the DB we need to call another API to fetch some more data and all this combined into a single JSON response. For example along with a list of users you need their last 5 payments from Stripe. This requires you to query your DB for the users and Stripe for the payments. Super Graph handles all this for you also only the fields you requested from the Stripe API are returned.
:::info Is this fast?
Super Graph is able fetch remote data and merge it with the DB response in an efficient manner. Several optimizations such as parallel HTTP requests and a zero-allocation JSON merge algorithm makes this very fast. All of this without you having to write a line of code.
:::
For example you need to list the last 3 payments made by a user. You will first need to look up the user in the database and then call the Stripe API to fetch his last 3 payments. For this to work your user table in the db has a `customer_id` column that contains his Stripe customer ID.
Similiarly you could also fetch the users last tweet, lead info from Salesforce or whatever else you need. It's fine to mix up several different `remote joins` into a single GraphQL query.
### Stripe Example
The configuration is self explanatory. A `payments` field has been added under the `customers` table. This field is added to the `remotes` subsection that defines fields associated with `customers` that are remote and not real database columns.
The `id` parameter maps a column from the `customers` table to the `$id` variable. In this case it maps `$id` to the `customer_id` column.
```yaml
tables:
- name: customers
remotes:
- name: payments
id: stripe_id
url: http://rails_app:3000/stripe/$id
path: data
# debug: true
# pass_headers:
# - cookie
# - host
set_headers:
- name: Authorization
value: Bearer <stripe_api_key>
```
#### How do I make use of this?
Just include `payments` like you would any other GraphQL selector under the `customers` selector. Super Graph will call the configured API for you and stitch (merge) the JSON the API sends back with the JSON generated from the database query. GraphQL features like aliases and fields all work.
```graphql
query {
customers {
id
email
payments {
customer_id
amount
billing_details
}
}
}
```
And voila here is the result. You get all of this advanced and honestly complex querying capability without writing a single line of code.
```json
"data": {
"customers": [
{
"id": 1,
"email": "linseymertz@reilly.co",
"payments": [
{
"customer_id": "cus_YCj3ndB5Mz",
"amount": 100,
"billing_details": {
"address": "1 Infinity Drive",
"zipcode": "94024"
}
},
...
```
Even tracing data is availble in the Super Graph web UI if tracing is enabled in the config. By default it is enabled in development. Additionally there you can set `debug: true` to enable http request / response dumping to help with debugging.
![Query Tracing](/tracing.png "Super Graph Web UI Query Tracing")
## Full text search
Every app these days needs search. Enought his often means reaching for something heavy like Solr. While this will work why add complexity to your infrastructure when Postgres has really great
and fast full text search built-in. And since it's part of Postgres it's also available in Super Graph.
```graphql
query {
products(
# Search for all products that contain 'ale' or some version of it
search: "ale"
# Return only matches where the price is less than 10
where: { price: { lt: 10 } }
# Use the search_rank to order from the best match to the worst
order_by: { search_rank: desc }
) {
id
name
search_rank
search_headline_description
}
}
```
This query will use the `tsvector` column in your database table to search for products that contain the query phrase or some version of it. To get the internal relevance ranking for the search results using the `search_rank` field. And to get the highlighted context within any of the table columns you can use the `search_headline_` field prefix. For example `search_headline_name` will return the contents of the products name column which contains the matching query marked with the `<b></b>` html tags.
```json
{
"data": {
"products": [
{
"id": 11,
"name": "Maharaj",
"search_rank": 0.243171,
"search_headline_description": "Blue Moon, Vegetable Beer, Willamette, 1007 - German <b>Ale</b>, 48 IBU, 7.9%, 11.8°Blg"
},
{
"id": 12,
"name": "Schneider Aventinus",
"search_rank": 0.243171,
"search_headline_description": "Dos Equis, Wood-aged Beer, Magnum, 1099 - Whitbread <b>Ale</b>, 15 IBU, 9.5%, 13.0°Blg"
},
...
```

View File

@ -1,335 +0,0 @@
---
id: config
title: Configuration
sidebar_label: Configuration
---
Configuration files can either be in YAML or JSON their names are derived from the `GO_ENV` variable, for example `GO_ENV=prod` will cause the `prod.yaml` config file to be used. or `GO_ENV=dev` will use the `dev.yaml`. A path to look for the config files in can be specified using the `-path <folder>` command line argument.
We're tried to ensure that the config file is self documenting and easy to work with.
```yaml
# Inherit config from this other config file
# so I only need to overwrite some values
inherits: base
app_name: "Super Graph Development"
host_port: 0.0.0.0:8080
web_ui: true
# debug, error, warn, info
log_level: "debug"
# enable or disable http compression (uses gzip)
http_compress: true
# When production mode is 'true' only queries
# from the allow list are permitted.
# When it's 'false' all queries are saved to the
# the allow list in ./config/allow.list
production: false
# Throw a 401 on auth failure for queries that need auth
auth_fail_block: false
# Latency tracing for database queries and remote joins
# the resulting latency information is returned with the
# response
enable_tracing: true
# Watch the config folder and reload Super Graph
# with the new configs when a change is detected
reload_on_config_change: true
# File that points to the database seeding script
# seed_file: seed.js
# Path pointing to where the migrations can be found
migrations_path: ./migrations
# Postgres related environment Variables
# SG_DATABASE_HOST
# SG_DATABASE_PORT
# SG_DATABASE_USER
# SG_DATABASE_PASSWORD
# Auth related environment Variables
# SG_AUTH_RAILS_COOKIE_SECRET_KEY_BASE
# SG_AUTH_RAILS_REDIS_URL
# SG_AUTH_RAILS_REDIS_PASSWORD
# SG_AUTH_JWT_PUBLIC_KEY_FILE
# inflections:
# person: people
# sheep: sheep
auth:
# Can be 'rails' or 'jwt'
type: rails
cookie: _app_session
# Comment this out if you want to disable setting
# the user_id via a header for testing.
# Disable in production
creds_in_header: true
rails:
# Rails version this is used for reading the
# various cookies formats.
version: 5.2
# Found in 'Rails.application.config.secret_key_base'
secret_key_base: 0a248500a64c01184edb4d7ad3a805488f8097ac761b76aaa6c17c01dcb7af03a2f18ba61b2868134b9c7b79a122bc0dadff4367414a2d173297bfea92be5566
# Remote cookie store. (memcache or redis)
# url: redis://redis:6379
# password: ""
# max_idle: 80
# max_active: 12000
# In most cases you don't need these
# salt: "encrypted cookie"
# sign_salt: "signed encrypted cookie"
# auth_salt: "authenticated encrypted cookie"
# jwt:
# provider: auth0
# secret: abc335bfcfdb04e50db5bb0a4d67ab9
# public_key_file: /secrets/public_key.pem
# public_key_type: ecdsa #rsa
# header:
# name: dnt
# exists: true
# value: localhost:8080
# You can add additional named auths to use with actions
# In this example actions using this auth can only be
# called from the Google Appengine Cron service that
# sets a special header to all it's requests
auths:
- name: from_taskqueue
type: header
header:
name: X-Appengine-Cron
exists: true
database:
type: postgres
host: db
port: 5432
dbname: app_development
user: postgres
password: postgres
#schema: "public"
#pool_size: 10
#max_retries: 0
#log_level: "debug"
# Set session variable "user.id" to the user id
# Enable this if you need the user id in triggers, etc
set_user_id: false
# database ping timeout is used for db health checking
ping_timeout: 1m
# Set up an secure tls encrypted db connection
enable_tls: false
# Required for tls. For example with Google Cloud SQL it's
# <gcp-project-id>:<cloud-sql-instance>"
# server_name: blah
# Required for tls. Can be a file path or the contents of the pem file
# server_cert: ./server-ca.pem
# Required for tls. Can be a file path or the contents of the pem file
# client_cert: ./client-cert.pem
# Required for tls. Can be a file path or the contents of the pem file
# client_key: ./client-key.pem
# Define additional variables here to be used with filters
variables:
admin_account_id: "5"
# Field and table names that you wish to block
blocklist:
- ar_internal_metadata
- schema_migrations
- secret
- password
- encrypted
- token
# Create custom actions with their own api endpoints
# For example the below action will be available at /api/v1/actions/refresh_leaderboard_users
# A request to this url will execute the configured SQL query
# which in this case refreshes a materialized view in the database.
# The auth_name is from one of the configured auths
actions:
- name: refresh_leaderboard_users
sql: REFRESH MATERIALIZED VIEW CONCURRENTLY "leaderboard_users"
auth_name: from_taskqueue
tables:
- name: customers
remotes:
- name: payments
id: stripe_id
url: http://rails_app:3000/stripe/$id
path: data
# debug: true
pass_headers:
- cookie
set_headers:
- name: Host
value: 0.0.0.0
# - name: Authorization
# value: Bearer <stripe_api_key>
- # You can create new fields that have a
# real db table backing them
name: me
table: users
roles_query: "SELECT * FROM users WHERE id = $user_id"
roles:
- name: anon
tables:
- name: products
limit: 10
query:
columns: ["id", "name", "description"]
aggregation: false
insert:
allow: false
update:
allow: false
delete:
allow: false
- name: user
tables:
- name: users
query:
filters: ["{ id: { _eq: $user_id } }"]
- name: products
query:
limit: 50
filters: ["{ user_id: { eq: $user_id } }"]
columns: ["id", "name", "description"]
disable_functions: false
insert:
filters: ["{ user_id: { eq: $user_id } }"]
columns: ["id", "name", "description"]
set:
- created_at: "now"
update:
filters: ["{ user_id: { eq: $user_id } }"]
columns:
- id
- name
set:
- updated_at: "now"
delete:
block: true
- name: admin
match: id = 1000
tables:
- name: users
filters: []
```
If deploying into environments like Kubernetes it's useful to be able to configure things like secrets and hosts though environment variables therfore we expose the below environment variables. This is escpecially useful for secrets since they are usually injected in via a secrets management framework ie. Kubernetes Secrets
Keep in mind any value can be overwritten using environment variables for example `auth.jwt.public_key_type` converts to `SG_AUTH_JWT_PUBLIC_KEY_TYPE`. In short prefix `SG_`, upper case and all `.` should changed to `_`.
#### Postgres environment variables
```bash
SG_DATABASE_HOST
SG_DATABASE_PORT
SG_DATABASE_USER
SG_DATABASE_PASSWORD
```
#### Auth environment variables
```bash
SG_AUTH_RAILS_COOKIE_SECRET_KEY_BASE
SG_AUTH_RAILS_REDIS_URL
SG_AUTH_RAILS_REDIS_PASSWORD
SG_AUTH_JWT_PUBLIC_KEY_FILE
```
## YugabyteDB
Yugabyte is an open-source, geo-distrubuted cloud-native relational DB that scales horizontally. Super Graph works with Yugabyte right out of the box. If you think you're data needs will outgrow Postgres and you don't really want to deal with sharding then Yugabyte is the way to go. Just point Super Graph to your Yugabyte DB and everything will just work including running migrations, seeding, querying, mutations, etc.
To use Yugabyte in your local development flow just uncomment the following lines in the `docker-compose.yml` file that is part of your Super Graph app. Also remember to comment out the originl postgres `db` config.
```yaml
# Postgres DB
# db:
# image: postgres:latest
# ports:
# - "5432:5432"
#Standard config to run a single node of Yugabyte
yb-master:
image: yugabytedb/yugabyte:latest
container_name: yb-master-n1
command: [ "/home/yugabyte/bin/yb-master",
"--fs_data_dirs=/mnt/disk0,/mnt/disk1",
"--master_addresses=yb-master-n1:7100",
"--replication_factor=1",
"--enable_ysql=true"]
ports:
- "7000:7000"
environment:
SERVICE_7000_NAME: yb-master
db:
image: yugabytedb/yugabyte:latest
container_name: yb-tserver-n1
command: [ "/home/yugabyte/bin/yb-tserver",
"--fs_data_dirs=/mnt/disk0,/mnt/disk1",
"--start_pgsql_proxy",
"--tserver_master_addrs=yb-master-n1:7100"]
ports:
- "9042:9042"
- "6379:6379"
- "5433:5433"
- "9000:9000"
environment:
SERVICE_5433_NAME: ysql
SERVICE_9042_NAME: ycql
SERVICE_6379_NAME: yedis
SERVICE_9000_NAME: yb-tserver
depends_on:
- yb-master
# Environment variables to point Super Graph to Yugabyte
# This is required since it uses a different user and port number
yourapp_api:
image: dosco/super-graph:latest
environment:
GO_ENV: "development"
Uncomment below for Yugabyte DB
SG_DATABASE_PORT: 5433
SG_DATABASE_USER: yugabyte
SG_DATABASE_PASSWORD: yugabyte
volumes:
- ./config:/config
ports:
- "8080:8080"
depends_on:
- db
```

View File

@ -1,159 +0,0 @@
---
id: deploy
title: How to deploy Super Graph
sidebar_label: Deploy to Prod.
---
Since you're reading this you're probably considering deploying Super Graph. You're in luck it's really easy and there are several ways to choose from. Keep in mind Super Graph can be used as a pre-built docker image or you can easily customize it and build your own docker image.
:::info JWT tokens (Auth0, etc)
When deploying on a subdomain and configure this service to use JWT authentication. You will need the public key file or secret key. Ensure your web app passes the JWT token with every GraphQL request (Cookie recommended). You have to add the web domain to the `cors_allowed_origins` config option so CORS can allow the browser to do cross-domain ajax requests.
:::
## Google Cloud Run (Fully Managed)
Cloud Run is a fully managed compute platform for deploying and scaling containerized applications quickly and securely.
Your Super Graph app comes with a `cloudbuild.yaml` file so it's really easy to use Google Cloud Build to build and deploy your Super Graph app to Google Cloud Run.
:::note
Remember to give Cloud Build permission to deploy to Cloud Run first this can be done in the Cloud Build settings screen. Also the service account you use with Cloud Run must have the IAM permissions to connect to CloudSQL. https://cloud.google.com/sql/docs/postgres/connect-run
:::
Use the command below to tell Cloud Build to build and deploy your app.
```bash
gcloud build submit --substitutions=SERVICE_ACCOUNT=admin@my-project.iam.gserviceaccount.com,REGION=us-central1 .
```
:::info Secrets Management
Your secrets like the database password should be managed by the Mozilla SOPS app. This is a secrets management app that encrypts all your secrets and stores them in a file to be decrypted in production using the Cloud KMS (Google Cloud KMS Or Amazon KMS). Our cloud build file above expects the secrets file to be `config/prod.secrets.yml`. You can find more information on Mozilla SOPS on their site. https://github.com/mozilla/sops
:::
## Build Docker Image Locally
If for whatever reason you decide to build your own Docker images then just use the command below.
```bash
docker build -t your-api-service-name .
```
## With a Rails app
Super Graph can read Rails session cookies, like those created by authentication gems (Devise or Warden). Based on how you've configured your Rails app the cookie can be signed, encrypted, both, include the user ID or just have the ID of the session. If you have choosen to use Redis or Memcache as your session store then Super Graph can read the session cookie and then lookup the user in the session store. In short it works really well with almost all Rails apps.
For any of this to work Super Graph must be deployed in a way that make the browser send the apps cookie to it along with the GraphQL query. That means Super Graph needs to be either on the same domain as your app or on a subdomain.
:::info I need an example
Say your Rails app runs on `myrailsapp.com` then Super Graph should be on the same domain or on a subdomain like `graphql.myrailsapp.com`. If you choose subdomain then remeber read the [Deploy under a subdomain](#deploy-under-a-subdomain) section.
:::
## Deploy under a subdomain
For this to work you have to ensure that the option `:domain => :all` is added to your Rails app config `Application.config.session_store` this will cause your rails app to create session cookies that can be shared with sub-domains. More info here [/sharing-a-devise-user-session-across-subdomains-with-rails](http://excid3.com/blog/sharing-a-devise-user-session-across-subdomains-with-rails-3/)
## With NGINX
If your infrastructure is fronted by NGINX then it should be configured so that all requests to your GraphQL API path are proxyed to Super Graph. In the example NGINX config below all requests to the path `/api/v1/graphql` are routed to wherever you have Super Graph installed within your architecture. This example is derived from the config file example at [/microservices-nginx-gateway/nginx.conf](https://github.com/launchany/microservices-nginx-gateway/blob/master/nginx.conf)
:::info NGINX with sub-domain
Yes, NGINX is very flexible and you can configure it to keep Super Graph a subdomain instead of on the same top level domain. I'm sure a little Googleing will get you some great example configs for that.
:::
```nginx
# Configuration for the server
server {
# Running port
listen 80;
# Proxy the graphql api path to Super Graph
location /api/v1/graphql {
proxy_pass http://super-graph-service:8080;
proxy_redirect off;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Host $server_name;
}
# Proxying all other paths to your Rails app
location / {
proxy_pass http://your-rails-app:3000;
proxy_redirect off;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Host $server_name;
}
}
```
## On Kubernetes
If your Rails app runs on Kubernetes then ensure you have an ingress config deployed that points the path to the service that you have deployed Super Graph under.
### Ingress config
```yaml
apiVersion: extensions/v1beta1
kind: Ingress
metadata:
name: simple-rails-app
annotations:
nginx.ingress.kubernetes.io/rewrite-target: /
spec:
rules:
- host: myrailsapp.com
http:
paths:
- path: /api/v1/graphql
backend:
serviceName: graphql-service
servicePort: 8080
- path: /
backend:
serviceName: rails-app
servicePort: 3000
```
### Service and deployment config
```yaml
apiVersion: v1
kind: Service
metadata:
name: graphql-service
labels:
run: super-graph
spec:
ports:
- port: 8080
protocol: TCP
selector:
run: super-graph
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: super-graph
spec:
selector:
matchLabels:
run: super-graph
replicas: 2
template:
metadata:
labels:
run: super-graph
spec:
containers:
- name: super-graph
image: docker.io/dosco/super-graph:latest
ports:
- containerPort: 8080
```

View File

@ -1,678 +0,0 @@
---
id: graphql
title: How to GraphQL
sidebar_label: GraphQL Syntax
---
GraphQL (GQL) is a simple query syntax that's fast replacing REST APIs. GQL is great since it allows web developers to fetch the exact data that they need without depending on changes to backend code. Also if you squint hard enough it looks a little bit like JSON :smiley:
The below query will fetch an `users` name, email and avatar image (renamed as picture). If you also need the users `id` then just add it to the query.
```graphql
query {
user {
full_name
email
picture: avatar
}
}
```
Multiple tables can also be fetched using a single GraphQL query. This is very fast since the entire query is converted into a single SQL query which the database can efficiently run.
```graphql
query {
user {
full_name
email
}
products {
name
description
}
}
```
### Fetching data
To fetch a specific `product` by it's ID you can use the `id` argument. The real name id field will be resolved automatically so this query will work even if your id column is named something like `product_id`.
```graphql
query {
products(id: 3) {
name
}
}
```
Postgres also supports full text search using a TSV index. Super Graph makes it easy to use this full text search capability using the `search` argument.
```graphql
query {
products(search: "ale") {
name
}
}
```
### Fragments
Fragments make it easy to build large complex queries with small composible and re-usable fragment blocks.
```graphql
query {
users {
...userFields2
...userFields1
picture_url
}
}
fragment userFields1 on user {
id
email
}
fragment userFields2 on user {
first_name
last_name
}
```
### Sorting
To sort or ordering results just use the `order_by` argument. This can be combined with `where`, `search`, etc to build complex queries to fit you needs.
```graphql
query {
products(order_by: { cached_votes_total: desc }) {
id
name
}
}
```
### Filtering
Super Graph support complex queries where you can add filters, ordering,offsets and limits on the query. For example the below query will list all products where the price is greater than 10 and the id is not 5.
```graphql
query {
products(where: { and: { price: { gt: 10 }, not: { id: { eq: 5 } } } }) {
name
price
}
}
```
#### Nested where clause targeting related tables
Sometimes you need to query a table based on a condition that applies to a related table. For example say you need to list all users who belong to an account. This query below will fetch the id and email or all users who belong to the account with id 3.
```graphql
query {
users(where: {
accounts: { id: { eq: 3 } }
}) {
id
email
}
}`
```
#### Logical Operators
| Name | Example | Explained |
| ---- | ------------------------------------------------------------ | ------------------------------- |
| and | price : { and : { gt: 10.5, lt: 20 } | price > 10.5 AND price < 20 |
| or | or : { price : { greater_than : 20 }, quantity: { gt : 0 } } | price >= 20 OR quantity > 0 |
| not | not: { or : { quantity : { eq: 0 }, price : { eq: 0 } } } | NOT (quantity = 0 OR price = 0) |
#### Other conditions
| Name | Example | Explained |
| ---------------------- | -------------------------------------- | -------------------------------------------------------------------------------------------------------- |
| eq, equals | id : { eq: 100 } | id = 100 |
| neq, not_equals | id: { not_equals: 100 } | id != 100 |
| gt, greater_than | id: { gt: 100 } | id > 100 |
| lt, lesser_than | id: { gt: 100 } | id < 100 |
| gte, greater_or_equals | id: { gte: 100 } | id >= 100 |
| lte, lesser_or_equals | id: { lesser_or_equals: 100 } | id <= 100 |
| in | status: { in: [ "A", "B", "C" ] } | status IN ('A', 'B', 'C) |
| nin, not_in | status: { in: [ "A", "B", "C" ] } | status IN ('A', 'B', 'C) |
| like | name: { like "phil%" } | Names starting with 'phil' |
| nlike, not_like | name: { nlike "v%m" } | Not names starting with 'v' and ending with 'm' |
| ilike | name: { ilike "%wOn" } | Names ending with 'won' case-insensitive |
| nilike, not_ilike | name: { nilike "%wOn" } | Not names ending with 'won' case-insensitive |
| similar | name: { similar: "%(b\|d)%" } | [Similar Docs](https://www.postgresql.org/docs/9/functions-matching.html#FUNCTIONS-SIMILARTO-REGEXP) |
| nsimilar, not_similar | name: { nsimilar: "%(b\|d)%" } | [Not Similar Docs](https://www.postgresql.org/docs/9/functions-matching.html#FUNCTIONS-SIMILARTO-REGEXP) |
| has_key | column: { has_key: 'b' } | Does JSON column contain this key |
| has_key_any | column: { has_key_any: [ a, b ] } | Does JSON column contain any of these keys |
| has_key_all | column: [ a, b ] | Does JSON column contain all of this keys |
| contains | column: { contains: [1, 2, 4] } | Is this array/json column a subset of value |
| contained_in | column: { contains: "{'a':1, 'b':2}" } | Is this array/json column a subset of these value |
| is_null | column: { is_null: true } | Is column value null or not |
### Aggregations
You will often find the need to fetch aggregated values from the database such as `count`, `max`, `min`, etc. This is simple to do with GraphQL, just prefix the aggregation name to the field name that you want to aggregrate like `count_id`. The below query will group products by name and find the minimum price for each group. Notice the `min_price` field we're adding `min_` to price.
```graphql
query {
products {
name
min_price
}
}
```
| Name | Explained |
| ----------- | ---------------------------------------------------------------------- |
| avg | Average value |
| count | Count the values |
| max | Maximum value |
| min | Minimum value |
| stddev | [Standard Deviation](https://en.wikipedia.org/wiki/Standard_deviation) |
| stddev_pop | Population Standard Deviation |
| stddev_samp | Sample Standard Deviation |
| variance | [Variance](https://en.wikipedia.org/wiki/Variance) |
| var_pop | Population Standard Variance |
| var_samp | Sample Standard variance |
All kinds of queries are possible with GraphQL. Below is an example that uses a lot of the features available. Comments `# hello` are also valid within queries.
```graphql
query {
products(
# returns only 30 items
limit: 30
# starts from item 10, commented out for now
# offset: 10,
# orders the response items by highest price
order_by: { price: desc }
# no duplicate prices returned
distinct: [price]
# only items with an id >= 30 and < 30 are returned
where: { id: { and: { greater_or_equals: 20, lt: 28 } } }
) {
id
name
price
}
}
```
### Custom Functions
Any function defined in the database like the below `add_five` that adds 5 to any number given to it can be used
within your query. The one limitation is that it should be a function that only accepts a single argument. The function is used within you're GraphQL in similar way to how aggregrations are used above. Example below
```grahql
query {
thread(id: 5) {
id
total_votes
add_five_total_votes
}
}
```
Postgres user-defined function `add_five`
```
CREATE OR REPLACE FUNCTION add_five(a integer) RETURNS integer AS $$
BEGIN
RETURN a + 5;
END;
$$ LANGUAGE plpgsql;
```
In GraphQL mutations is the operation type for when you need to modify data. Super Graph supports the `insert`, `update`, `upsert` and `delete`. You can also do complex nested inserts and updates.
When using mutations the data must be passed as variables since Super Graphs compiles the query into an prepared statement in the database for maximum speed. Prepared statements are are functions in your code when called they accept arguments and your variables are passed in as those arguments.
### Insert
```json
{
"data": {
"name": "Art of Computer Programming",
"description": "The Art of Computer Programming (TAOCP) is a comprehensive monograph written by computer scientist Donald Knuth",
"price": 30.5
}
}
```
```graphql
mutation {
product(insert: $data) {
id
name
}
}
```
#### Bulk insert
```json
{
"data": [
{
"name": "Art of Computer Programming",
"description": "The Art of Computer Programming (TAOCP) is a comprehensive monograph written by computer scientist Donald Knuth",
"price": 30.5
},
{
"name": "Compilers: Principles, Techniques, and Tools",
"description": "Known to professors, students, and developers worldwide as the 'Dragon Book' is available in a new edition",
"price": 93.74
}
]
}
```
```graphql
mutation {
product(insert: $data) {
id
name
}
}
```
### Update
```json
{
"data": {
"price": 200.0
},
"product_id": 5
}
```
```graphql
mutation {
product(update: $data, id: $product_id) {
id
name
}
}
```
#### Bulk update
```json
{
"data": {
"price": 500.0
},
"gt_product_id": 450.0,
"lt_product_id:": 550.0
}
```
```graphql
mutation {
product(
update: $data
where: { price: { gt: $gt_product_id, lt: lt_product_id } }
) {
id
name
}
}
```
### Delete
```json
{
"data": {
"price": 500.0
},
"product_id": 5
}
```
```graphql
mutation {
product(delete: true, id: $product_id) {
id
name
}
}
```
#### Bulk delete
```json
{
"data": {
"price": 500.0
}
}
```
```graphql
mutation {
product(delete: true, where: { price: { eq: { 500.0 } } }) {
id
name
}
}
```
### Upsert
```json
{
"data": {
"id": 5,
"name": "Art of Computer Programming",
"description": "The Art of Computer Programming (TAOCP) is a comprehensive monograph written by computer scientist Donald Knuth",
"price": 30.5
}
}
```
```graphql
mutation {
product(upsert: $data) {
id
name
}
}
```
#### Bulk upsert
```json
{
"data": [
{
"id": 5,
"name": "Art of Computer Programming",
"description": "The Art of Computer Programming (TAOCP) is a comprehensive monograph written by computer scientist Donald Knuth",
"price": 30.5
},
{
"id": 6,
"name": "Compilers: Principles, Techniques, and Tools",
"description": "Known to professors, students, and developers worldwide as the 'Dragon Book' is available in a new edition",
"price": 93.74
}
]
}
```
```graphql
mutation {
product(upsert: $data) {
id
name
}
}
```
Often you will need to create or update multiple related items at the same time. This can be done using nested mutations. For example you might need to create a product and assign it to a user, or create a user and his products at the same time. You just have to use simple json to define you mutation and Super Graph takes care of the rest.
### Nested Insert
Create a product item first and then assign it to a user
```json
{
"data": {
"name": "Apple",
"price": 1.25,
"created_at": "now",
"updated_at": "now",
"user": {
"connect": { "id": 5 }
}
}
}
```
```graphql
mutation {
product(insert: $data) {
id
name
user {
id
full_name
email
}
}
}
```
Or it's reverse, create the user first and then his product
```json
{
"data": {
"email": "thedude@rug.com",
"full_name": "The Dude",
"created_at": "now",
"updated_at": "now",
"product": {
"name": "Apple",
"price": 1.25,
"created_at": "now",
"updated_at": "now"
}
}
}
```
```graphql
mutation {
user(insert: $data) {
id
full_name
email
product {
id
name
price
}
}
}
```
### Nested Update
Update a product item first and then assign it to a user
```json
{
"data": {
"name": "Apple",
"price": 1.25,
"user": {
"connect": { "id": 5 }
}
}
}
```
```graphql
mutation {
product(update: $data, id: 5) {
id
name
user {
id
full_name
email
}
}
}
```
Or it's reverse, update a user first and then his product
```json
{
"data": {
"email": "newemail@me.com",
"full_name": "The Dude",
"product": {
"name": "Banana",
"price": 1.25
}
}
}
```
```graphql
mutation {
user(update: $data, id: 1) {
id
full_name
email
product {
id
name
price
}
}
}
```
### Pagination
This is a must have feature of any API. When you want your users to go through a list page by page or implement some fancy infinite scroll you're going to need pagination. There are two ways to paginate in Super Graph.
Limit-Offset
This is simple enough but also inefficient when working with a large number of total items. Limit, limits the number of items fetched and offset is the point you want to fetch from. The below query will fetch 10 results at a time starting with the 100th item. You will have to keep updating offset (110, 120, 130, etc ) to walk thought the results so make offset a variable.
```graphql
query {
products(limit: 10, offset: 100) {
id
slug
name
}
}
```
#### Cursor
This is a powerful and highly efficient way to paginate a large number of results. Infact it does not matter how many total results there are this will always be lighting fast. You can use a cursor to walk forward or backward through the results. If you plan to implement infinite scroll this is the option you should choose.
When going this route the results will contain a cursor value this is an encrypted string that you don't have to worry about just pass this back in to the next API call and you'll received the next set of results. The cursor value is encrypted since its contents should only matter to Super Graph and not the client. Also since the primary key is used for this feature it's possible you might not want to leak it's value to clients.
You will need to set this config value to ensure the encrypted cursor data is secure. If not set a random value is used which will change with each deployment breaking older cursor values that clients might be using so best to set it.
```yaml
# Secret key for general encryption operations like
# encrypting the cursor data
secret_key: supercalifajalistics
```
Paginating forward through your results
```json
{
"variables": {
"cursor": "MJoTLbQF4l0GuoDsYmCrpjPeaaIlNpfm4uFU4PQ="
}
}
```
```graphql
query {
products(first: 10, after: $cursor) {
slug
name
}
}
```
Paginating backward through your results
```graphql
query {
products(last: 10, before: $cursor) {
slug
name
}
}
```
```graphql
"data": {
"products": [
{
"slug": "eius-nulla-et-8",
"name" "Pale Ale"
},
{
"slug": "sapiente-ut-alias-12",
"name" "Brown Ale"
}
...
],
"products_cursor": "dJwHassm5+d82rGydH2xQnwNxJ1dcj4/cxkh5Cer"
}
```
Nested tables can also have cursors. Requesting multiple cursors are supported on a single request but when paginating using a cursor only one table is currently supported. To explain this better, you can only use a `before` or `after` argument with a cursor value to paginate a single table in a query.
```graphql
query {
products(last: 10) {
slug
name
customers(last: 5) {
email
full_name
}
}
}
```
Multiple order-by arguments are supported. Super Graph is smart enough to allow cursor based pagination when you also need complex sort order like below.
```graphql
query {
products(
last: 10
before: $cursor
order_by: [ price: desc, total_customers: asc ]) {
slug
name
}
}
```
## Using Variables
Variables (`$product_id`) and their values (`"product_id": 5`) can be passed along side the GraphQL query. Using variables makes for better client side code as well as improved server side SQL query caching. The built-in web-ui also supports setting variables. Not having to manipulate your GraphQL query string to insert values into it makes for cleaner
and better client side code.
```javascript
// Define the request object keeping the query and the variables seperate
var req = {
query: "{ product(id: $product_id) { name } }",
variables: { product_id: 5 },
};
// Use the fetch api to make the query
fetch("http://localhost:8080/api/v1/graphql", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify(req),
})
.then((res) => res.json())
.then((res) => console.log(res.data));
```

View File

@ -1,190 +0,0 @@
---
id: home
title: Super Graph
hide_title: true
sidebar_label: Home
---
import useBaseUrl from '@docusaurus/useBaseUrl'; // Add to the top of the file below the front matter.
<div class="hero shadow--lw margin-bottom--lg">
<div class="container">
<div class="row">
<div class="col col--2">
<img
class="avatar__photo avatar__photo--xl"
alt="Super Graph Logo"
src={useBaseUrl('img/super-graph-logo.svg')}
height="70"
/>
</div>
<div class="col col--10"><h1 class="hero__title">Super Graph</h1></div>
</div>
<p class="hero__subtitle">Fetch data without code!</p>
<div class="margin-bottom--lg">
<a class="button button--secondary button--outline button--lg" href="start">
Skip Intro
</a>
</div>
<p>Stop fighting ORM's and complex SQL just to fetch the data you need. Instead try Super Graph it automagically tranforms GraphQL into efficient SQL.</p>
</div>
</div>
:::info cut development time
80% of all web app development is either reading from or writing to a database. 100x your developer productivity and save valuable time by making that super simple.
:::
### Fetching data with GraphQL
Just imagine the code or SQL you'll need to fetch this data, the user, all his posts, all the votes on the posts, the authors information and the related tags. Oh yeah and you also need efficient cursor based pagination. And Remember you also need to maintain this code forever.
Instead just describe the data you need in GraphQL and give that to Super Graph it'll automatically learn your database and generate the most efficient SQL query fetching your data in the JSON structure you expected.
```graphql
query {
user(id: 5) {
id
first_name
last_name
picture_url
posts(first: 20, order_by: { score: desc }) {
slug
title
created_at
votes_total
votes {
created_at
}
author {
id
name
}
tags {
id
name
}
}
posts_cursor
}
}
```
### Instant results
Here's the data Super Graph fetched using the GraphQL above, it's even in the JSON structure you
wanted it in. All this without you writing any code or SQL.
```json
{
"data": {
"user": {
"id": 5,
"threads": [
{
"id": 1,
"title": "This is a sample tite for this thread.",
"topics": [
{
"id": 3,
"name": "CloudRun"
}
]
}]
},
"posts": [
{
"id": 1477,
"body": "These are some example contents for this post.",
"slug": "monitor-negotiate-store-1476",
"votes": [],
"author": {
"id": 5,
"email": "jordanecruickshank@ferry.io"
},
"created_at": "2020-05-13T13:51:21.729501+00:00"
},
...
],
"posts_cursor": "a8d4j2k9d83dy373hd2nskw2sjs8"
}
}
```
## How do I use it?
Super Graph can be used in two ways. You can run it as a standalone service serving as an API backend for your app or as a library within your own app code. Super Graph is built in GO a secure and high-performance language from Google used to build cloud infratructure.
### Using it as a service
```bash
go get github.com/dosco/super-graph
super-graph new <app_name>
```
### Using it in your own code
```bash
go get github.com/dosco/super-graph/core
```
```go
package main
import (
"database/sql"
"fmt"
"time"
"github.com/dosco/super-graph/core"
_ "github.com/jackc/pgx/v4/stdlib"
)
func main() {
db, err := sql.Open("pgx", "postgres://postgrs:@localhost:5432/example_db")
if err != nil {
log.Fatal(err)
}
sg, err := core.NewSuperGraph(nil, db)
if err != nil {
log.Fatal(err)
}
query := `
query {
posts {
id
title
}
}`
ctx = context.WithValue(ctx, core.UserIDKey, 1)
res, err := sg.GraphQL(ctx, query, nil)
if err != nil {
log.Fatal(err)
}
fmt.Println(string(res.Data))
}
```
## Follow us for updates
For when you need help or just want to stay in the loop
[Twitter](https://twitter.com/dosco) or [Discord](https://discord.gg/6pSWCTZ).
## Why I created Super Graph
After working on several products through my career I found that we spend way too much time on building API backends. Most APIs also need constant updating, and this costs time and money.
It's always the same thing, figure out what the UI needs then build an endpoint for it. Most API code involves struggling with an ORM to query a database and mangle the data into a shape that the UI expects to see.
I didn't want to write this code anymore, I wanted the computer to do it. Enter GraphQL, to me it sounded great, but it still required me to write all the same database query code.
Having worked with compilers before I saw this as a compiler problem. Why not build a compiler that converts GraphQL to highly efficient SQL.
This compiler is what sits at the heart of Super Graph, with layers of useful functionality around it like authentication, remote joins, rails integration, database migrations, and everything else needed for you to build production-ready apps with it.
## Apache License 2.0
Apache Public License 2.0 | Copyright © 2018-present Vikram Rangnekar

View File

@ -1,271 +0,0 @@
---
id: internals
title: Codebase Explained
sidebar_label: Code Internals
---
Super Graph code is made up of a number of packages. We have done our best to keep each package small and focused. Let us begin by looking at some of these packages.
1. qcode - GraphQL lexer and parser.
2. psql - SQL generator
3. serv - HTTP Endpoint, Configs, CLI, etc
4. rails - Rails cookie and session store decoders
## QCODE
This package contains the core of the GraphQL compiler it handling the lexing and parsing of the GraphQL query transforming it into an internal representation called
`QCode`.
This is the first step of the compiling process the `func NewCompiler(c Config)` function creates a new instance of this compiler which has it's own config.
Keep in mind QCode has no knowledge of the Database structure it is designed to be a fast GraphQL parser. Care is taken to keep memory allocations to a minimum.
```go
const (
opQuery
opMutate
...
)
type QCode struct {
Type QType
Selects []Select
...
}
type Select struct {
ID int32
ParentID int32
Args map[string]*Node
Name string
FieldName string
Cols []Column
Where *Exp
OrderBy []*OrderBy
DistinctOn []string
Paging Paging
Children []int32
Functions bool
Allowed map[string]struct{}
PresetMap map[string]string
PresetList []string
}
```
But before the incoming GraphQL query can be turned into QCode it must first be tokenzied by the lexer `lex.go`. As the tokenzier walks the bytes of the query it generates tokens `item` structs which are then consumed by the next step the parser `parse.go`.
```go
type item struct {
typ itemType
pos Pos
end Pos
}
```
For exmple a simple query like `query getUser { user { id } }` will be converted into several tokens like below.
```go
item{itemQuery, 0, 4} // query
item{itemName, 6, 12} // getUser
item{itemObjOpen, 16, 20} // {
...
```
These tokens are then fed into the parser `parse.go` the parser does the work of generating an abstract syntax tree (AST) from the tokens. This AST is an internal representation (data structure) and is not exposed outside the package. Since the AST is a tree a stack `stack.go` is used to walk the tree and generate the QCode AST. The QCode data structure is also a tree (represented as an array). This is then returned to the caller of the compile function.
```go
type Operation struct {
Type parserType
Name string
Args []Arg
Fields []Field
}
type Field struct {
ID int32
ParentID int32
Name string
Alias string
Args []Arg
Children []int32
}
```
## PSQL
This package is responsible for generating Postgres SQL from the QCode AST. There are various GraphQL query types (Query, Mutation, etc). And several more sub types like single root or multi-root queries, various types of mutations (insert, update delete, bulk insert, etc). This package is designed to be able to generate SQL for all of those types.
In addition to QCode variable data is also passed to the compile function within this package. Variables are decoded to derive what is being inserted and what kind of insert is it single or bulk. This information is not available in the GraphQL query its passed in seperatly via variables. This package is able to put all this together and generate the right SQL code.
The entry point of this package is in `query.go`. The database schema must be passed in the config object when creating a new compiler instance `NewCompiler`. The functions to extract this schema from the database are also part of this package `tables.go`. The `GetTables` functions fetches all the tables from the database and `GetColumns` fetches columns and relationship information.
```go
func NewCompiler(conf Config) *Compiler {
return &Compiler{conf.Schema, conf.Vars}
}
func (co *Compiler) Compile(qc *qcode.QCode, w io.Writer, vars Variables) (uint32, error) {
switch qc.Type {
case qcode.QTQuery:
return co.compileQuery(qc, w)
case qcode.QTInsert, qcode.QTUpdate, qcode.QTDelete, qcode.QTUpsert:
return co.compileMutation(qc, w, vars)
}
return 0, fmt.Errorf("Unknown operation type %d", qc.Type)
}
```
GraphQL, input is first converted to QCode.
```graphql
query {
user {
id
}
posts {
title
}
}
```
SQL, in reality the generated SQL is far more complex single it has to be very efficient, leverage the power of Postgres, support RBAC (Role based access control) and all of this must be done in a single SQL query.
```sql
SELECT users.id, posts.title FROM users, posts;
```
## SERV
The `serv` package constains most of code that turns the above compiler into an HTTP service. It also includes authentication middleware, remote join resolvers, config parsering, database migrations and seeding commands.
Another big feature that this package handles is the `allow.list` management code. In production mode parsing the allow list file and registering prepared statements to adding GraphQL queries to this file in development mode.
Currently the following global variables are referrenced across the package. In future I'd prefer to move these into a context struct and pass that around instead.
```go
var (
logger zerolog.Logger // logger for everything but errors
errlog zerolog.Logger // logger for errors includes line numbers
conf *config // parsed config
confPath string // path to the config file
db *pgxpool.Pool // database connection pool
schema *psql.DBSchema // database tables, columns and relationships
qcompile *qcode.Compiler // qcode compiler
pcompile *psql.Compiler // postgres sql compiler
)
```
## Testing
There are several unit tests and benchmark tests `parse_test.go`) included. There are also scripts included for memory `pprof_cpu.sh` and cpu `pprof_cpu.sh` profiling.
```go
// Test to ensure synthetic tables gnerate the correct SQL
func syntheticTables(t *testing.T) {
gql := `query {
me {
email
}
}`
sql := `SELECT json_object_agg('me', json_0) FROM (SELECT row_to_json((SELECT "json_row_0" FROM (SELECT ) AS "json_row_0")) AS "json_0" FROM (SELECT "users"."email" FROM "users" WHERE ((("users"."id") = '{{user_id}}' :: bigint)) LIMIT ('1') :: integer) AS "users_0" LIMIT ('1') :: integer) AS "sel_0"`
resSQL, err := compileGQLToPSQL(gql, nil, "user")
if err != nil {
t.Fatal(err)
}
if string(resSQL) != sql {
t.Fatal(errNotExpected)
}
}
```
You can run tests within each package or across the entire app. It is usually the fastest to first write a test and then build the feature to satisfy it.
```
go test -v ./...
```
Memory profiling can help find where allocations are happining within the package code.
```bash
$ cd ./psql
$ ./pprof_mem.sh
goos: darwin
goarch: amd64
pkg: github.com/dosco/super-graph/psql
BenchmarkCompile-8 52567 19401 ns/op 3918 B/op 61 allocs/op
BenchmarkCompileParallel-8 219548 5684 ns/op 3938 B/op 61 allocs/op
PASS
ok github.com/dosco/super-graph/psql 2.582s
Type: alloc_space
Time: Nov 29, 2019 at 11:59pm (EST)
Entering interactive mode (type "help" for commands, "o" for options)
(pprof) top
Showing nodes accounting for 880.59MB, 80.63% of 1092.14MB total
Dropped 33 nodes (cum <= 5.46MB)
Showing top 10 nodes out of 35
flat flat% sum% cum cum%
22MB 2.01% 2.01% 903.57MB 82.73% github.com/dosco/super-graph/qcode.(*Compiler).Compile
0 0% 2.01% 862.98MB 79.02% github.com/dosco/super-graph/psql.BenchmarkCompileParallel.func1
0 0% 2.01% 862.98MB 79.02% testing.(*B).RunParallel.func1
461.95MB 42.30% 44.31% 760.53MB 69.64% github.com/dosco/super-graph/qcode.(*Compiler).compileQuery
396.63MB 36.32% 80.63% 396.63MB 36.32% github.com/dosco/super-graph/util.NewStack
0 0% 80.63% 252.07MB 23.08% github.com/dosco/super-graph/qcode.(*Compiler).compileArgs
0 0% 80.63% 228.15MB 20.89% testing.(*B).runN
0 0% 80.63% 227.63MB 20.84% github.com/dosco/super-graph/psql.BenchmarkCompile
0 0% 80.63% 227.63MB 20.84% testing.(*B).launch
0 0% 80.63% 187.04MB 17.13% github.com/dosco/super-graph/psql.(*Compiler).Compile
```
## Benchmarking
Most packages contain benchmark tests to ensure new features don't introduce a significant regression to performance.
```bash
$ cd ./psql
$ go test -v -run=xx -bench=.
goos: darwin
goarch: amd64
pkg: github.com/dosco/super-graph/psql
BenchmarkCompile-8 60775 19076 ns/op 3919 B/op 61 allocs/op
BenchmarkCompileParallel-8 207847 5172 ns/op 3937 B/op 61 allocs/op
PASS
ok github.com/dosco/super-graph/psql 2.530s
```
## Reach out
If you'd like me to explain other parts of the code please reach out over Twitter or Discord. I'll keep adding to this doc as I get time.
## Developing Super Graph
If you want to build and run Super Graph from code then the below commands will build the web ui and launch Super Graph in developer mode with a watcher to rebuild on code changes. And the demo rails app is also launched to make it easier to test changes.
```bash
# yarn install dependencies and build the web ui
make build
# do this the only the time to setup the database
docker-compose run rails_app rake db:create db:migrate db:seed
# start super graph in development mode with a change watcher
docker-compose up
```
## Developing on the Super Graph UI
```bash
# yarn is needed to build the web ui
brew install yarn
# this is where the react app for the ui lives
cd internals/serv/web
# launch it in development mpde
yarn start
```

View File

@ -1,144 +0,0 @@
---
id: intro
title: Introduction
sidebar_label: Introduction
---
import useBaseUrl from '@docusaurus/useBaseUrl'; // Add to the top of the file below the front matter.
Super Graph is a service that instantly and without code gives you a high performance and secure GraphQL API. Your GraphQL queries are auto translated into a single fast SQL query. No more spending weeks or months writing backend API code. Just make the query you need and Super Graph will do the rest.
Super Graph has a rich feature set like integrating with your existing Ruby on Rails apps, joining your DB with data from remote APIs, Role and Attribute based access control, Support for JWT tokens, DB migrations, seeding and a lot more.
## Features
- Role and Attribute based access control
- Works with existing Ruby-On-Rails apps
- Automatically learns database schemas and relationships
- Full text search and aggregations
- Rails authentication supported (Redis, Memcache, Cookie)
- JWT tokens supported (Auth0, etc)
- Join database with remote REST APIs
- Highly optimized and fast Postgres SQL queries
- GraphQL queries and mutations
- A simple config file
- High performance GO codebase
- Tiny docker image and low memory requirements
- Fuzz tested for security
- Database migrations tool
- Database seeding tool
- Works with Postgres and Yugabyte DB
- OpenCensus Support: Zipkin, Prometheus, X-Ray, Stackdriver
## Try the demo app
```bash
# clone the repository
git clone https://github.com/dosco/super-graph
# run db in background
docker-compose up -d db
# see logs and wait until DB is really UP
docker-compose logs db
# setup the demo rails app & database and run it
docker-compose run rails_app rake db:create db:migrate db:seed
# run the demo
docker-compose up
# signin to the demo app (user1@demo.com / 123456)
open http://localhost:3000
# try the super graph web ui
open http://localhost:8080
```
:::note Docker?
This demo requires `docker` you can either install it using `brew` or from the
docker website [https://docs.docker.com/docker-for-mac/install/](https://docs.docker.com/docker-for-mac/install/)
:::
## GraphQL
We fully support queries and mutations. For example the below GraphQL query would fetch two products that belong to the current user where the price is greater than 10.
### GraphQL Query
```graphql
query {
users {
id
email
picture: avatar
password
full_name
products(limit: 2, where: { price: { gt: 10 } }) {
id
name
description
price
}
}
}
```
#### JSON Result
```json
{
"data": {
"users": [
{
"id": 1,
"email": "odilia@west.info",
"picture": "https://robohash.org/simur.png?size=300x300",
"full_name": "Edwin Orn",
"products": [
{
"id": 16,
"name": "Sierra Nevada Style Ale",
"description": "Belgian Abbey, 92 IBU, 4.7%, 17.4°Blg",
"price": 16.47
},
...
]
}
]
}
}
```
:::note Set User ID
In development mode you can use the `X-User-ID: 4` header to set a user id so you don't have to worries about cookies etc. This can be set using the _HTTP Headers_ tab at the bottom of the web UI.
:::
### GraphQL Mutation
In another example the below GraphQL mutation would insert a product into the database. The first part of the below example is the variable data and the second half is the GraphQL mutation. For mutations data has to always ben passed as a variable.
```json
{
"data": {
"name": "Art of Computer Programming",
"description": "The Art of Computer Programming (TAOCP) is a comprehensive monograph written by computer scientist Donald Knuth",
"price": 30.5
}
}
```
```graphql
mutation {
product(insert: $data) {
id
name
}
}
```
### Built-in GraphQL Editor
Quickly craft and test your queries with a full-featured GraphQL editor. Auto-complete and schema documentation is automatically available.
<img alt="Zipkin Traces" src={useBaseUrl("img/webui.jpg")} />

View File

@ -1,101 +0,0 @@
---
id: react
title: ReactJS Examples
sidebar_label: ReactJS Examples
---
## Apollo Client
We recommend using the new [Apollo Client v3](https://www.apollographql.com/docs/react/v3.0-beta/) `@apollo/client` this is the latest version of the Apollo GraphQL javascript client library. Previous versions of this library were called `apollo-boost`.
This library contains react hooks `useQuery`, `useMutation`, `useLazyQuery`, etc that make it easy to add GraphQL queries to your React app.
```bash
npm install @apollo/client graphql
```
### Creating a client
```jsx title="App.js"
import React from "react";
import { ApolloClient, ApolloProvider } from "@apollo/client";
const client = new ApolloClient({ uri: `/api/v1/graphql` });
```
### Or a client with caching enabled
```jsx title="App.js"
import React from "react";
import { ApolloClient, ApolloProvider, InMemoryCache } from "@apollo/client";
// Enable GraphQL caching really speeds up your app
const cache = new InMemoryCache({
freezeResults: true,
// Set `dataIdFromObject` id is not called `id`
// dataIdFromObject: obj => { return obj.slug }
});
const client = new ApolloClient({ uri: `/api/v1/graphql`, cache: cache });
```
### Use the client in components to query for data
In this example we create a component `UserProfile` that user's name by his `id` and displays that on the page.
```jsx
import React from "react";
import { gql, useQuery } from "@apollo/client";
const UserProfile = { id } => {
const query = gql`
query getUser {
user(id: $id) {
id
name
}
}`;
const variables = { id };
const { error, loading, data } = useQuery(query, { variables });
if (loading) {
return <p>Loading</p>;
}
return <h1>{data.name}</h1>
}
```
### Use the client in components to post data back
In this example we create a component `UpdateUserProfile` displays a text input which when changed causes a mutation query to be trigger to update the users name in the backend database. You can also insert or delete in mutation. More complex mutations like bulk insert, or nested insert and updates are also supported. Nested inserts are create to create an entry and create or update a related entitiy in the same request.
```jsx
import React from "react";
import { gql, useQuery } from "@apollo/client";
const UpdateUserProfile = { id } => {
const query = gql`
mutation setUserName {
user(id: $id, update: $data) {
id
name
}
}`;
const [setUserName] = useMutation(query);
const updateName = (e) => {
let name = e.target.value;
let variables = { data: { name } };
setUserName({ variables })
}
return <input type="text" onChange={updateName}>
}
```

View File

@ -1,230 +0,0 @@
---
id: security
title: API Security
sidebar_label: API Security
---
One of the the most common questions I get asked is what happens if a user out on the internet sends queries
that we don't want run. For example how do we stop him from fetching all users or the emails of users. Our answer to this is that it is not an issue as this cannot happen, let me explain.
Super Graph runs in one of two modes `development` or `production`, this is controlled via the config value `production: false` when it's false it's running in development mode and when true, production. In development mode all the **named** queries (including mutations) are saved to the allow list `./config/allow.list`. While in production mode when Super Graph starts only the queries from this allow list file are registered with the database as [prepared statements](https://stackoverflow.com/questions/8263371/how-can-prepared-statements-protect-from-sql-injection-attacks).
Prepared statements are designed by databases to be fast and secure. They protect against all kinds of sql injection attacks and since they are pre-processed and pre-planned they are much faster to run then raw sql queries. Also there's no GraphQL to SQL compiling happening in production mode which makes your queries lighting fast as they are directly sent to the database with almost no overhead.
In short in production only queries listed in the allow list file `./config/allow.list` can be used, all other queries will be blocked.
:::tip How to think about the allow list?
The allow list file is essentially a list of all your exposed API calls and the data that passes within them. It's very easy to build tooling to do things like parsing this file within your tests to ensure fields like `credit_card_no` are not accidently leaked. It's a great way to build compliance tooling and ensure your user data is always safe.
:::
This is an example of a named query, `getUserWithProducts` is the name you've given to this query it can be anything you like but should be unique across all you're queries. Only named queries are saved in the allow list in development mode.
```graphql
query getUserWithProducts {
users {
id
name
products {
id
name
price
}
}
}
```
## Authentication
You can only have one type of auth enabled either Rails or JWT.
### Ruby on Rails
Almost all Rails apps use Devise or Warden for authentication. Once the user is
authenticated a session is created with the users ID. The session can either be
stored in the users browser as a cookie, memcache or redis. If memcache or redis is used then a cookie is set in the users browser with just the session id.
Super Graph can handle all these variations including the old and new session formats. Just enable the right `auth` config based on how your rails app is configured.
#### Cookie session store
```yaml
auth:
type: rails
cookie: _app_session
rails:
# Rails version this is used for reading the
# various cookies formats.
version: 5.2
# Found in 'Rails.application.config.secret_key_base'
secret_key_base: 0a248500a64c01184edb4d7ad3a805488f8097ac761b76aaa6c17c01dcb7af03a2f18ba61b2868134b9c7b79a122bc0dadff4367414a2d173297bfea92be5566
```
#### Memcache session store
```yaml
auth:
type: rails
cookie: _app_session
rails:
# Memcache remote cookie store.
url: memcache://127.0.0.1
```
#### Redis session store
```yaml
auth:
type: rails
cookie: _app_session
rails:
# Redis remote cookie store
url: redis://127.0.0.1:6379
password: ""
max_idle: 80
max_active: 12000
```
### JWT Tokens
```yaml
auth:
type: jwt
jwt:
# valid providers are auth0, firebase and none
provider: auth0
secret: abc335bfcfdb04e50db5bb0a4d67ab9
public_key_file: /secrets/public_key.pem
public_key_type: ecdsa #rsa
```
For JWT tokens we currently support tokens from a provider like Auth0 or if you have a custom solution then we look for the `user_id` in the `subject` claim of of the `id token`. If you pick Auth0 then we derive two variables from the token `user_id` and `user_id_provider` for to use in your filters.
We can get the JWT token either from the `authorization` header where we expect it to be a `bearer` token or if `cookie` is specified then we look there.
For validation a `secret` or a public key (ecdsa or rsa) is required. When using public keys they have to be in a PEM format file.
### Firebase Auth
```yaml
auth:
type: jwt
jwt:
provider: firebase
audience: <firebase-project-id>
```
Firebase auth also uses JWT the keys are auto-fetched from Google and used according to their documentation mechanism. The `audience` config value needs to be set to your project id and everything else is taken care for you.
### HTTP Headers
```yaml
header:
name: X-AppEngine-QueueName
exists: true
#value: default
```
Header auth is usually the best option to authenticate requests to the action endpoints. For example you
might want to use an action to refresh a materalized view every hour and only want a cron service like the Google AppEngine Cron service to make that request in this case a config similar to the one above will do.
The `exists: true` parameter ensures that only the existance of the header is checked not its value. The `value` parameter lets you confirm that the value matches the one assgined to the parameter. This helps in the case you are using a shared secret to protect the endpoint.
### Named Auth
```yaml
# You can add additional named auths to use with actions
# In this example actions using this auth can only be
# called from the Google Appengine Cron service that
# sets a special header to all it's requests
auths:
- name: from_taskqueue
type: header
header:
name: X-Appengine-Cron
exists: true
```
In addition to the default auth configuration you can create additional named auth configurations to be used
with features like `actions`. For example while your main GraphQL endpoint uses JWT for authentication you may want to use a header value to ensure your actions can only be called by clients having access to a shared secret
or security header.
## Actions
Actions is a very useful feature that is currently work in progress. For now the best use case for actions is to
refresh database tables like materialized views or call a database procedure to refresh a cache table, etc. An action creates an http endpoint that anyone can call to have the SQL query executed. The below example will create an endpoint `/api/v1/actions/refresh_leaderboard_users` any request send to that endpoint will cause the sql query to be executed. the `auth_name` points to a named auth that should be used to secure this endpoint. In future we have big plans to allow your own custom code to run using actions.
```yaml
actions:
- name: refresh_leaderboard_users
sql: REFRESH MATERIALIZED VIEW CONCURRENTLY "leaderboard_users"
auth_name: from_taskqueue
```
#### Using CURL to test a query
```bash
# fetch the response json directly from the endpoint using user id 5
curl 'http://localhost:8080/api/v1/graphql' \
-H 'content-type: application/json' \
-H 'X-User-ID: 5' \
--data-binary '{"query":"{ products { name price users { email }}}"}'
```
## Access Control
It's common for APIs to control what information they return or insert based on the role of the user. In Super Graph we have two primary roles `user` and `anon` the first for users where a `user_id` is available the latter for users where it's not.
:::info Define authenticated request?
An authenticated request is one where Super Graph can extract an `user_id` based on the configured authentication method (jwt, rails cookies, etc).
:::
The `user` role can be divided up into further roles based on attributes in the database. For example when fetching a list of users, a normal user can only fetch his own entry while an admin can fetch all the users within a company and an admin user can fetch everyone. In some places this is called Attribute based access control. So in way we support both. Role based access control and Attribute based access control.
Super Graph allows you to create roles dynamically using a `roles_query` and `match` config values.
### Configure RBAC
```yaml
roles_query: "SELECT * FROM users WHERE users.id = $user_id"
roles:
- name: user
tables:
- name: users
query:
filters: ["{ id: { _eq: $user_id } }"]
insert:
filters: ["{ user_id: { eq: $user_id } }"]
columns: ["id", "name", "description"]
presets:
- created_at: "now"
update:
filters: ["{ user_id: { eq: $user_id } }"]
columns:
- id
- name
presets:
- updated_at: "now"
delete:
block: true
- name: admin
match: users.id = 1
tables:
- name: users
query:
filters: []
```
This configuration is relatively simple to follow the `roles_query` parameter is the query that must be run to help figure out a users role. This query can be as complex as you like and include joins with other tables.
The individual roles are defined under the `roles` parameter and this includes each table the role has a custom setting for. The role is dynamically matched using the `match` parameter for example in the above case `users.id = 1` means that when the `roles_query` is executed a user with the id `1` will be assigned the admin role and those that don't match get the `user` role if authenticated successfully or the `anon` role.

View File

@ -1,249 +0,0 @@
---
id: seed
title: Database Seeding
sidebar_label: Seed Scripts
---
While developing it's often useful to be able to have fake data available in the database. Fake data can help with building the UI and save you time when trying to get the GraphQL query correct. Super Graph has the ability do this for you. All you have to do is write a seed script `config/seed.js` (In Javascript) and use the `db:seed` command line option. Below is an example of kind of things you can do in a seed script.
## Creating fake users
Since all mutations and queries are in standard GraphQL you can use all the features available in Super Graph GraphQL.
```javascript
var users = [];
for (i = 0; i < 20; i++) {
var data = {
slug: util.make_slug(fake.first_name() + "-" + fake.last_name()),
first_name: fake.first_name(),
last_name: fake.last_name(),
picture_url: fake.avatar_url(),
email: fake.email(),
bio: fake.sentence(10),
};
var res = graphql(" \
mutation { \
user(insert: $data) { \
id \
} \
}", { data: data });
users.push(res.user);
}
```
## Inserting the users fake blog posts
Another example highlighting how the `connect` syntax of Super Graph GraphQL can be used to connect inserted posts
to random users that were previously created. For futher details checkout the [seed script](/seed) documentation.
```javascript
var posts = [];
for (i = 0; i < 1500; i++) {
var user.id = users[Math.floor(Math.random() * 10)];
var data = {
slug: util.make_slug(fake.sentence(3) + i),
body: fake.sentence(100),
published: true,
thread: {
connect: { user: user.id }
}
}
var res = graphql(" \
mutation { \
post(insert: $data) { \
id \
} \
}",
{ data: data },
{ user_id: u.id })
posts.push(res.post.slug)
}
```
## Insert a large number of rows efficiently
This feature uses the `COPY` functionality available in Postgres this is the best way to
insert a large number of rows into a table. The `import_csv` function reads in a CSV file using the first
line of the file as column names.
```javascript
import_csv("post_tags", "./tags.csv");
```
## A list of fake data functions available to you.
```
person
name
name_prefix
name_suffix
first_name
last_name
gender
ssn
contact
email
phone
phone_formatted
username
password
// Address
address
city
country
country_abr
state
state_abr
street
street_name
street_number
street_prefix
street_suffix
zip
latitude
latitude_in_range
longitude
longitude_in_range
// Beer
beer_alcohol
beer_hop
beer_ibu
beer_blg
beer_malt
beer_name
beer_style
beer_yeast
// Cars
car
car_type
car_maker
car_model
// Text
word
sentence
paragraph
question
quote
// Misc
generate
boolean
uuid
// Colors
color
hex_color
rgb_color
safe_color
// Internet
url
image_url
avatar_url
domain_name
domain_suffix
ipv4_address
ipv6_address
http_method
user_agent
user_agent_firefox
user_agent_chrome
user_agent_opera
user_agent_safari
// Date / Time
date
date_range
nano_second
second
minute
hour
month
day
weekday
year
timezone
timezone_abv
timezone_full
timezone_offset
// Payment
price
credit_card
credit_card_cvv
credit_card_number
credit_card_type
currency
currency_long
currency_short
// Company
bs
buzzword
company
company_suffix
job
job_description
job_level
job_title
// Hacker
hacker_abbreviation
hacker_adjective
hacker_noun
hacker_phrase
hacker_verb
//Hipster
hipster_word
hipster_paragraph
hipster_sentence
// File
file_extension
file_mine_type
// Numbers
number
numerify
int8
int16
int32
int64
uint8
uint16
uint32
uint64
float32
float32_range
float64
float64_range
shuffle_ints
mac_address
// String
digit
letter
lexify
rand_string
numerify
```
## Some more utility functions
```
shuffle_strings(string_array)
make_slug(text)
make_slug_lang(text, lang)
```

View File

@ -1,170 +0,0 @@
---
id: start
title: Getting Started
sidebar_label: Getting Started
---
Super Graph can generate your initial app for you. The generated app will have config files, database migrations and seed files among other things like docker related files.
You can then add your database schema to the migrations, maybe create some seed data using the seed script and launch Super Graph. You're now good to go and can start working on your UI frontend in React, Vue or whatever.
```bash
# Download and install Super Graph. You will need Go 1.14 or above
go get github.com/dosco/super-graph
```
And then create and launch your new app
```bash
# Create a new app and change to it's directory
super-graph new blog
cd blog
# Setup the app database and seed it with fake data.
# Docker compose will start a Postgres database for your app
docker-compose run blog_api ./super-graph db:setup
# Finally launch Super Graph configured for your app
docker-compose up
```
Lets take a look at the files generated by Super Graph when you create a new app
```bash
super-graph new blog
> created 'blog'
> created 'blog/Dockerfile'
> created 'blog/docker-compose.yml'
> created 'blog/config'
> created 'blog/config/dev.yml'
> created 'blog/config/prod.yml'
> created 'blog/config/seed.js'
> created 'blog/config/migrations'
> created 'blog/config/migrations/100_init.sql'
> app 'blog' initialized
```
:::note Docker
Docker Compose is a great way to run multiple services while developing on your desktop or laptop. In our case we need Postgres and Super Graph to both be running and the `docker-compose.yml` is configured to do just that. The Super Graph service is named after your app postfixed with `_api`. The Dockerfile can be used build a containr of your app for production deployment.
:::
Run Super Graph with Docker compose
```bash
docker-compose run blog_api ./super-graph help
```
### Config files
All the config files needs to configure Super Graph for your app are contained in this folder for starters you have two `dev.yaml` and `prod.yaml`. When the `GO_ENV` environment variable is set to `development` then `dev.yaml` is used and the prod one when it's set to `production`. Stage and Test are the other two environment options, but you can set the `GO_ENV` to whatever you like (eg. `alpha-test`) and Super Graph will look for a yaml file with that name to load config from.
### Seed.js
Having data flowing through your API makes building your frontend UI so much easier. When creafting say a user profile wouldn't it be nice for the API to return a fake user with name, picture and all. This is why having the ability to seed your database is important. Seeding cn also be used in production to setup some initial users like the admins or to add an initial set of products to a ecommerce store.
Super Graph makes this easy by allowing you to write your seeding script in plan old Javascript. The below file that auto-generated for new apps uses our built-in functions `fake` and `graphql` to generate fake data and use GraphQL mutations to insert it into the database.
```javascript
// Example script to seed database
var users = [];
for (i = 0; i < 10; i++) {
var data = {
full_name: fake.name(),
email: fake.email(),
};
var res = graphql(" \
mutation { \
user(insert: $data) { \
id \
} \
}", { data: data });
users.push(res.user);
}
```
If you want to import a lot of data using a CSV file is the best and fastest option. The `import_csv` command uses the `COPY FROM` Postgres method to load massive amounts of data into tables. The first line of the CSV file must be the header with column names.
```javascript
var post_count = import_csv("posts", "posts.csv");
```
You can generate the following fake data for your seeding purposes. Below is the list of fake data functions supported by the built-in fake data library. For example `fake.image_url()` will generate a fake image url or `fake.shuffle_strings(['hello', 'world', 'cool'])` will generate a randomly shuffled version of that array of strings or `fake.rand_string(['hello', 'world', 'cool'])` will return a random string from the array provided.
### Migrations
Easy database migrations is the most important thing when building products backend by a relational database. We make it super easy to manage and migrate your database.
```bash
super-graph db:new create_users
> created migration 'config/migrations/101_create_users.sql'
```
Migrations in Super Graph are plain old Postgres SQL. Here's an example for the above migration.
```sql
-- Write your migrate up statements here
CREATE TABLE public.users (
id bigint GENERATED ALWAYS AS IDENTITY PRIMARY KEY,
full_name text,
email text UNIQUE NOT NULL CHECK (length(email) < 255),
created_at timestamptz NOT NULL NOT NULL DEFAULT NOW(),
updated_at timestamptz NOT NULL NOT NULL DEFAULT NOW()
);
---- create above / drop below ----
-- Write your down migrate statements here. If this migration is irreversible
-- then delete the separator line above.
DROP TABLE public.users
```
We would encourage you to leverage triggers to maintain consistancy of your data for example here are a couple triggers that you can add to you init migration and across your tables.
```sql
-- This trigger script will set the updated_at column everytime a row is updated
CREATE OR REPLACE FUNCTION trigger_set_updated_at()
RETURNS TRIGGER SET SCHEMA 'public' LANGUAGE 'plpgsql' AS $$
BEGIN
new.updated_at = now();
RETURN new;
END;
$$;
...
-- An exmple of adding this trigger to the users table
CREATE TRIGGER set_updated_at BEFORE UPDATE ON public.users
FOR EACH ROW EXECUTE PROCEDURE trigger_set_updated_at();
```
```sql
-- This trigger script will set the user_id column to the current
-- Super Graph user.id value everytime a row is created or updated
CREATE OR REPLACE FUNCTION trigger_set_user_id()
RETURNS TRIGGER SET SCHEMA 'public' LANGUAGE 'plpgsql' AS $$
BEGIN
IF TG_OP = 'UPDATE' THEN
new.user_id = old.user_id;
ELSE
new.user_id = current_setting('user.id')::int;
END IF;
RETURN new;
END;
$$;
...
-- An exmple of adding this trigger to the blog_posts table
CREATE TRIGGER set_user_id BEFORE INSERT OR UPDATE ON public.blog_posts
FOR EACH ROW EXECUTE PROCEDURE trigger_set_user_id();
```

View File

@ -1,82 +0,0 @@
---
id: telemetry
title: Tracing and Metrics
sidebar_label: Telemetry
---
import useBaseUrl from '@docusaurus/useBaseUrl'; // Add to the top of the file below the front matter.
Having observability and telemetry is at the core of any production ready service. Super Graph has built-in support for OpenCensus for tracing requests all the way from HTTP to the database and providing all kinds of metrics.
OpenCensus has a concept called exporters these are external services that can consume this data and make to give you graphs, charts, alerting etc. Super Graph again has built-in support for Zipkin, Prometheus, Google Stackdriver and the AWS X-Ray exporters.
## Telemetry config
The `telemetry` section of the standard config files is where you set values to configure this feature to your needs.
```yaml
telemetry:
debug: true
interval: 5s
metrics:
exporter: "prometheus"
endpoint: ""
namespace: "web api"
key: "1234xyz"
tracing:
exporter: "zipkin"
endpoint: "http://zipkin:9411/api/v2/spans"
sample: 0.2
include_query: false
include_params: false
```
**debug**: Enabling debug enables an embedded web ui to test and debug tracing and metrics. This UI called `zPages` is provided by OpenCensus and will be made available on the `/telemetry` path. For more information on using `zPages` https://opencensus.io/zpages/. Remeber to disable this in production.
**interval**: This controls the interval setting for OpenCensus metrics collection. This deafaults to `5 seconds` if not set.
**metric.exporters** Setting this enables metrics collection. The supported values for this field are `prometheus` and `stackdriver`. The Prometheus exporter requires `metric.namespace` to be set. The Sackdriver exporter requires the `metric.key` to be set to the Google Cloud Project ID.
**metric.endpoint** Is not currently used by any of the exporters.
**tracing.exporter** Setting this enables request tracing. The supported values for this field are `zipkin`, `aws` and `xray`. Zipkin requires `tracing.endpoint` to be set. AWS and Xray are the same and do not require any addiitonal settings.
**tracing.sample** This controls what percentage of the requests should be traced. By default `0.5` or 50% of the requests are traced, `always` is also a valid value for this field and it means all requests will be traced.
**include_query** Include the Super Graph SQL query to the trace. Be careful with this setting in production it will add the entire SQL query to the trace. This can be veru useful to debug slow requests.
**include_params** Include the Super Graph SQL query parameters to the trace. Be careful with this setting in production it will it can potentially leak sensitive user information into tracing logs.
## Using Zipkin
Zipkin is a really great open source request tracing project. It's easy to add to your current Super Graph app as a way to test tracing in development. Add the following to the Super Graph generated `docker-compose.yml` file. Also add `zipkin` in your current apps `depends_on` list. Once setup the Zipkin UI is available at http://localhost:9411
```yaml
your_api:
...
depends_on:
- db
- zipkin
zipkin:
image: openzipkin/zipkin-slim
container_name: zipkin
# Environment settings are defined here https://github.com/openzipkin/zipkin/blob/master/zipkin-server/README.md#environment-variables
environment:
- STORAGE_TYPE=mem
# Uncomment to enable self-tracing
# - SELF_TRACING_ENABLED=true
# Uncomment to enable debug logging
# - JAVA_OPTS=-Dorg.slf4j.simpleLogger.log.zipkin2=debug
ports:
# Port used for the Zipkin UI and HTTP Api
- 9411:9411
```
### Zipkin HTTP to DB traces
<img alt="Zipkin Traces" src={useBaseUrl("img/zipkin1.png")} />
### Zipkin trace details
<img alt="Zipkin Traces" src={useBaseUrl('img/zipkin2.png')} />

View File

@ -1,13 +0,0 @@
---
id: webui
title: Web UI / GraphQL Editor
sidebar_label: Web UI
---
import useBaseUrl from '@docusaurus/useBaseUrl'; // Add to the top of the file below the front matter.
<img alt="Zipkin Traces" src={useBaseUrl("img/webui.jpg")} />
Super Graph comes with a build-in GraphQL editor that only runs in development. Use it to craft your queries and copy-paste them into you're app once you're ready. The editor supports auto-completation and schema documentation. This makes it super easy to craft and test your query all in one go without knowing anything about the underlying database structure.
You can even set query variables or http headers as required. To simulate an authenticated user set the http header `"X-USER-ID": 5` to the user id of the user you want to test with.

Some files were not shown because too many files have changed in this diff Show More