development approach b benchmark - #672
Conversation
vharkins1
commented
Aug 15, 2026
vharkins1
commented
- Remove Pipeline C-only contract (extraction, incidents, reporting) from approach-b
- Reshape templates & forms contract to Pipeline B
- Wire fill flow to accept input_id and use stored transcript ([Pipeline B] Wire the fill flow to accept input_id and use the stored transcript #638)
- Updated and pinned ruff latest version. Created ruff.toml
- ci: run lint and tests on development- PR bases*
- Link FormSubmission to Input by foreign key ([Pipeline B] Link FormSubmission to Input by foreign key #639)
- Apply routes->services->repositories layering to forms ([Pipeline B] Apply routes → services → repositories layering to forms & templates #640, part 1)
- Apply routes->services->repositories layering to templates ([Pipeline B] Apply routes → services → repositories layering to forms & templates #640, part 2)
- Route Controller through FormService in the async fill task ([Pipeline B] Route Controller/FileManipulator through the form service #641)
- Match PDF fields by name instead of position in Filler ([Pipeline B] Fix positional PDF field-matching in Filler #642)
- feat: ✨ first approach of pipeline B implementation
- feat: ✨ template creation adjusted based on pdf_path
- refactor: ♻️ linter
- fix: 🐛 removing field confidence in pipelineextractionoutput structure
- fix: 🐛 sqlalchemy errors
- fix: 🐛 celery.py was importing celery but never calling it
- fix: 🐛 Added ConfigDict to the pydantic import. Replaced the nested class Config with model_config = ConfigDict(from_attributes=True) at the class body level — the Pydantic v2 way.
- style: 💄 runner logging
- The validator for Big Little model testing
…database to accuracy validators
659 dataset creations
…using-pytest-benchmark-to-compare-pipeline-performance [FEAT]: Benchmark Module Using pytest-benchmark to Compare Pipeline Performance
667 add pdfs to benchmark
…lass Config with model_config = ConfigDict(from_attributes=True) at the class body level — the Pydantic v2 way.
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: bbcfa76054
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| template_id: int = Field(foreign_key="template.id") | ||
| input_id: UUID | None = Field(default=None, foreign_key="inputs.input_id") | ||
| input_text: str | ||
| extracted_fields: dict | None = Field(default=None, sa_column=Column(JSON)) |
There was a problem hiding this comment.
The new ORM column is not created by migration 003, which only adds input_id. On any database initialized or upgraded through Alembic, ORM reads and writes of FormSubmission will therefore reference a nonexistent extracted_fields column, breaking form filling, submission listing, analytics, and deletion; the migration and downgrade need to add and remove this JSON column as well.
Useful? React with 👍 / 👎.
| from celery.schedules import crontab | ||
|
|
||
| from app.core.config import CELERY_BROKER_URL, CELERY_RESULT_BACKEND | ||
| celery_app = Celery("fireform") |
There was a problem hiding this comment.
The checked Docker worker starts from app.core.celery:celery_app, but this module no longer imports or includes app.tasks.fill, app.tasks.purge, or app.tasks.transcribe. Consequently the separate worker process does not register those tasks and will reject queued form-fill and transcription messages as unregistered; restore the task includes/imports (and the existing app configuration) when constructing the Celery application.
Useful? React with 👍 / 👎.
2 participants
