Files
pdf/gateway/app/routers/render.py
T
2026-06-29 10:51:04 +05:30

217 lines
7.4 KiB
Python

from typing import Annotated
from fastapi import APIRouter, HTTPException, Path, Query, Response, status
from app.schemas.font import FontInfoResponse
from app.services import engine
from app.services.font import font_info_to_response
from app.services.store import document_store
router = APIRouter(prefix="/documents/{document_id}/pages", tags=["render"])
compat_router = APIRouter(tags=["render"])
@router.get("/{page_index}/render")
def render_page(
document_id: str, page_index: Annotated[int, Path(ge=0)], dpi: int = 96
) -> Response:
if not engine.is_available():
raise HTTPException(
status_code=status.HTTP_501_NOT_IMPLEMENTED,
detail="Engine bridge (bindings/python) not yet available.",
)
doc_info = document_store.get_document(document_id)
if not doc_info:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Document not found")
try:
doc = doc_info["doc_instance"]
page = doc.get_page(page_index)
img = page.render(dpi)
return Response(content=img.data, media_type="image/png")
except IndexError:
raise HTTPException(
status_code=status.HTTP_404_NOT_FOUND, detail="Page index out of bounds"
)
except Exception as e:
raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail=str(e))
@router.get("/{page_index}/text")
def extract_page_text(document_id: str, page_index: Annotated[int, Path(ge=0)]):
if not engine.is_available():
raise HTTPException(
status_code=status.HTTP_501_NOT_IMPLEMENTED,
detail="Engine bridge (bindings/python) not yet available.",
)
doc_info = document_store.get_document(document_id)
if not doc_info:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Document not found")
try:
doc = doc_info["doc_instance"]
page = doc.get_page(page_index)
text = page.extract_text()
annots = page.extract_annotations_text()
if annots:
text += "\n" + "\n".join(annots)
page_height = page.height
glyphs = page.extract_text_with_bounds()
for g in glyphs:
g["y"] = page_height - (g["y"] + g["h"])
return {"text": text, "glyphs": glyphs}
except IndexError:
raise HTTPException(
status_code=status.HTTP_404_NOT_FOUND, detail="Page index out of bounds"
)
except Exception as e:
raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail=str(e))
@compat_router.get("/render/{document_id}")
def render_page_compat(
document_id: str, page: int = 0, zoom: float = 1.0, rotation: int = 0
) -> Response:
dpi = int(96 * zoom)
return render_page(document_id, page, dpi)
@router.get("/{page_index}")
def get_page_info(document_id: str, page_index: Annotated[int, Path(ge=0)]):
if not engine.is_available():
raise HTTPException(
status_code=status.HTTP_501_NOT_IMPLEMENTED,
detail="Engine bridge (bindings/python) not yet available.",
)
doc_info = document_store.get_document(document_id)
if not doc_info:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Document not found")
try:
doc = doc_info["doc_instance"]
page = doc.get_page(page_index)
return {"width": page.width, "height": page.height}
except IndexError:
raise HTTPException(
status_code=status.HTTP_404_NOT_FOUND, detail="Page index out of bounds"
)
except Exception as e:
raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail=str(e))
@router.get("/{page_index}/transform/page-to-device")
def transform_page_to_device(
document_id: str,
page_index: int,
x: float,
y: float,
device_width: int,
device_height: int,
rotate: int = 0,
):
if not engine.is_available():
raise HTTPException(
status_code=status.HTTP_501_NOT_IMPLEMENTED, detail="Engine unavailable"
)
doc_info = document_store.get_document(document_id)
if not doc_info:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Document not found")
try:
pdfengine = engine.require()
doc = doc_info["doc_instance"]
page = doc.get_page(page_index)
pt = pdfengine.Point2D(x=x, y=y)
res = page.page_to_device(pt, device_width, device_height, rotate)
return {"x": res.x, "y": res.y}
except Exception as e:
raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail=str(e))
@router.get("/{page_index}/transform/device-to-page")
def transform_device_to_page(
document_id: str,
page_index: int,
x: int,
y: int,
device_width: int,
device_height: int,
rotate: int = 0,
):
if not engine.is_available():
raise HTTPException(
status_code=status.HTTP_501_NOT_IMPLEMENTED, detail="Engine unavailable"
)
doc_info = document_store.get_document(document_id)
if not doc_info:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Document not found")
try:
pdfengine = engine.require()
doc = doc_info["doc_instance"]
page = doc.get_page(page_index)
pt = pdfengine.DevicePoint(x=x, y=y)
res = page.device_to_page(pt, device_width, device_height, rotate)
return {"x": res.x, "y": res.y}
except Exception as e:
raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail=str(e))
@router.get("/{page_index}/fonts", response_model=list[FontInfoResponse])
def get_page_fonts(document_id: str, page_index: int) -> list[FontInfoResponse]:
if not engine.is_available():
raise HTTPException(
status_code=status.HTTP_501_NOT_IMPLEMENTED,
detail="Engine bridge (bindings/python) not yet available.",
)
doc_info = document_store.get_document(document_id)
if not doc_info:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Document not found")
try:
doc = doc_info["doc_instance"]
page = doc.get_page(page_index)
fonts = page.get_fonts()
return [font_info_to_response(f) for f in fonts]
except IndexError:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Page index out of range")
except Exception as e:
raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail=str(e))
@router.get("/{page_index}/fonts/glyph-width")
def get_page_glyph_width(
document_id: str,
page_index: Annotated[int, Path(ge=0)],
font_name: str,
charcode: int,
font_size: Annotated[float, Query(gt=0)] = 12.0,
):
if not engine.is_available():
raise HTTPException(
status_code=status.HTTP_501_NOT_IMPLEMENTED,
detail="Engine bridge (bindings/python) not yet available.",
)
doc_info = document_store.get_document(document_id)
if not doc_info:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Document not found")
try:
doc = doc_info["doc_instance"]
page = doc.get_page(page_index)
width = page.get_glyph_width(font_name, charcode, font_size)
return {"width": width}
except IndexError:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Page index out of range")
except Exception as e:
raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail=str(e))