from __future__ import annotations import uuid from pathlib import Path from typing import Literal, Optional import structlog from fastapi import APIRouter, BackgroundTasks, Depends, File, Query, UploadFile from fastapi.responses import FileResponse, JSONResponse from app.core.auth import require_api_key from app.core.config import settings from app.core.exceptions import InvalidParameterError, JobNotFoundError from app.models.schemas import ( AsyncJobResponse, ConversionParams, ConversionResult, HealthResponse, ImageFormat, JobStatus, JobStatusResponse, ) from app.services.conversion import conversion_service from app.services.file_service import upload_file_to_service from app.services.upload_orchestrator import convert_and_upload from app.utils.validators import fetch_pdf_from_url, read_and_validate_pdf logger = structlog.get_logger(__name__) router = APIRouter() _jobs: dict[str, JobStatusResponse] = {} _MEDIA_TYPES: dict[str, str] = { "png": "image/png", "jpg": "image/jpeg", "jpeg": "image/jpeg", } def build_conversion_params( format: ImageFormat = Query(ImageFormat.PNG, description="Output image format"), dpi: int = Query(settings.DEFAULT_DPI, ge=72, le=settings.MAX_DPI, description="Render DPI (72–600)"), quality: int = Query(settings.DEFAULT_JPEG_QUALITY, ge=1, le=100, description="JPEG/WEBP quality (1–100)"), pages: Optional[str] = Query(None, description="Page spec: '1', '1-3', '1,3,5-7'. Empty = all pages."), grayscale: bool = Query(False, description="Convert output to grayscale"), transparent_bg: bool = Query(False, description="Preserve transparency (PNG only)"), split_page: bool = Query(False, description="True: each page separate image. False: stitch into one tall image."), ) -> ConversionParams: try: return ConversionParams( format=format, dpi=dpi, quality=quality, pages=pages, grayscale=grayscale, transparent_bg=transparent_bg, split_page=split_page, ) except Exception as exc: raise InvalidParameterError(str(exc)) from exc def _upload_mode_param( upload_mode: Optional[Literal["per_page", "merged"]] = Query( None, description=( "Upload behaviour after conversion. " "'per_page' uploads each page image immediately as it is rendered (lowest latency). " "'merged' waits for all pages to render, then uploads them. " "Omit to use the server default (PDF_UPLOAD_MODE env var)." ), ), ) -> Optional[str]: return upload_mode @router.post( "/convert", response_model=ConversionResult, summary="Convert PDF to images — file upload (sync)", tags=["PDF Conversion"], ) async def convert_sync_upload( file: UploadFile = File(...), params: ConversionParams = Depends(build_conversion_params), upload_mode: Optional[str] = Depends(_upload_mode_param), _: str = Depends(require_api_key), ): """ Upload a PDF and convert all (or selected) pages to JPEG or PNG images synchronously. - Each page is rendered in **parallel** (split into individual page blobs when split_page=true). - Converted images are uploaded to the File Upload Service automatically. - Returns file URLs and metadata for every converted page. - Supported output formats: JPEG, PNG. """ pdf_bytes = await read_and_validate_pdf(file) job_id = str(uuid.uuid4()) return await convert_and_upload(pdf_bytes, params, job_id, upload_mode) @router.post( "/convert/url", response_model=ConversionResult, summary="Convert PDF to images — URL (sync)", tags=["PDF Conversion"], ) async def convert_sync_url( pdf_url: str = Query(..., description="Publicly accessible HTTPS URL to a PDF"), params: ConversionParams = Depends(build_conversion_params), upload_mode: Optional[str] = Depends(_upload_mode_param), _: str = Depends(require_api_key), ): """Fetch a PDF from a public URL and convert it. Private/loopback IPs are blocked.""" pdf_bytes = await fetch_pdf_from_url(pdf_url) job_id = str(uuid.uuid4()) return await convert_and_upload(pdf_bytes, params, job_id, upload_mode) @router.post( "/convert/async", response_model=AsyncJobResponse, status_code=202, summary="Submit async conversion job — file upload", tags=["PDF Conversion"], ) async def convert_async_upload( background_tasks: BackgroundTasks, file: UploadFile = File(...), params: ConversionParams = Depends(build_conversion_params), upload_mode: Optional[str] = Depends(_upload_mode_param), _: str = Depends(require_api_key), ): """Submit a PDF for async conversion. Returns 202 immediately; poll /jobs/{job_id}.""" pdf_bytes = await read_and_validate_pdf(file) job_id = str(uuid.uuid4()) _jobs[job_id] = JobStatusResponse(job_id=job_id, status=JobStatus.PENDING, progress=0.0) background_tasks.add_task(_run_conversion_job, job_id, pdf_bytes, params, upload_mode) return AsyncJobResponse( job_id=job_id, status=JobStatus.PENDING, message="Job accepted. Poll the status URL for progress.", status_url=f"/api/v1/jobs/{job_id}", ) @router.post( "/convert/async/url", response_model=AsyncJobResponse, status_code=202, summary="Submit async conversion job — URL", tags=["PDF Conversion"], ) async def convert_async_url( background_tasks: BackgroundTasks, pdf_url: str = Query(..., description="Publicly accessible HTTPS URL to a PDF"), params: ConversionParams = Depends(build_conversion_params), upload_mode: Optional[str] = Depends(_upload_mode_param), _: str = Depends(require_api_key), ): """Submit a URL-based PDF conversion job asynchronously.""" pdf_bytes = await fetch_pdf_from_url(pdf_url) job_id = str(uuid.uuid4()) _jobs[job_id] = JobStatusResponse(job_id=job_id, status=JobStatus.PENDING, progress=0.0) background_tasks.add_task(_run_conversion_job, job_id, pdf_bytes, params, upload_mode) return AsyncJobResponse( job_id=job_id, status=JobStatus.PENDING, message="Job accepted. Poll the status URL for progress.", status_url=f"/api/v1/jobs/{job_id}", ) @router.get( "/jobs/{job_id}", response_model=JobStatusResponse, summary="Get async job status", tags=["Jobs"], ) async def get_job_status( job_id: str, _: str = Depends(require_api_key), ): job = _jobs.get(job_id) if not job: raise JobNotFoundError(job_id) return job @router.get( "/files/{job_id}/{filename}", summary="Download a converted image", tags=["Files"], ) async def download_file( job_id: str, filename: str, _: str = Depends(require_api_key), ): output_root = Path(settings.OUTPUT_DIR).resolve() file_path = (output_root / job_id / filename).resolve() try: file_path.relative_to(output_root) except ValueError: raise JobNotFoundError(f"{job_id}/{filename}") if not file_path.exists() or not file_path.is_file(): raise JobNotFoundError(f"{job_id}/{filename}") ext = file_path.suffix.lstrip(".").lower() media_type = _MEDIA_TYPES.get(ext, "application/octet-stream") return FileResponse(path=str(file_path), media_type=media_type, filename=filename) @router.get("/health", response_model=HealthResponse, summary="Liveness probe", tags=["Observability"]) async def health(): return HealthResponse(status="ok", version=settings.VERSION, environment=settings.ENV) @router.get("/ready", response_model=HealthResponse, summary="Readiness probe", tags=["Observability"]) async def ready(): try: probe = Path(settings.OUTPUT_DIR) / ".probe" probe.touch() probe.unlink() except OSError as exc: return JSONResponse(status_code=503, content={"status": "unavailable", "detail": str(exc)}) return HealthResponse(status="ready", version=settings.VERSION, environment=settings.ENV) async def _run_conversion_job( job_id: str, pdf_bytes: bytes, params: ConversionParams, upload_mode: Optional[str], ) -> None: _jobs[job_id] = JobStatusResponse(job_id=job_id, status=JobStatus.PROCESSING, progress=10.0) try: result = await convert_and_upload(pdf_bytes, params, job_id, upload_mode) _jobs[job_id] = JobStatusResponse( job_id=job_id, status=JobStatus.COMPLETED, progress=100.0, result=result, ) logger.info("async_job_complete", job_id=job_id) except Exception as exc: logger.exception("async_job_failed", job_id=job_id) _jobs[job_id] = JobStatusResponse( job_id=job_id, status=JobStatus.FAILED, progress=0.0, error=str(exc), )