mirror of
https://github.com/snapotter-hq/SnapOtter.git
synced 2026-08-03 07:46:42 +02:00
SnapOtterandGitHub
7d37f6e6f5
When a PDF has no text layer (scanned or image-only), pdf-to-text now returns a 422 that points at the OCR tool instead of a silent empty file, and text downloads carry charset=utf-8 so UTF-8 Arabic renders correctly when the .txt is viewed inline. Fixes #589
37 lines
1.3 KiB
Python
37 lines
1.3 KiB
Python
"""Extract plain text. Args: {"path": in, "out": out-txt-path}. Prints {"chars": N, "hasText": bool}."""
|
|
import json
|
|
import sys
|
|
|
|
|
|
def main():
|
|
args = json.loads(sys.argv[1]) if len(sys.argv) > 1 else {}
|
|
path, out = args.get("path"), args.get("out")
|
|
if not path or not out:
|
|
print(json.dumps({"error": "missing path/out"}))
|
|
sys.exit(1)
|
|
try:
|
|
import fitz
|
|
except ImportError:
|
|
print(json.dumps({"error": "PyMuPDF not installed"}))
|
|
sys.exit(1)
|
|
try:
|
|
doc = fitz.open(path)
|
|
parts = [page.get_text() for page in doc]
|
|
doc.close()
|
|
text = "\n".join(parts)
|
|
# hasText distinguishes a PDF with a real text layer from a scanned or
|
|
# image-only PDF, where every page returns "" and only the join newlines
|
|
# remain. len(text) alone can't tell them apart, so the caller uses this
|
|
# to offer OCR instead of handing back an empty file (#589).
|
|
has_text = any(part.strip() for part in parts)
|
|
with open(out, "w", encoding="utf-8") as fh:
|
|
fh.write(text)
|
|
print(json.dumps({"chars": len(text), "hasText": has_text}))
|
|
except Exception as exc: # noqa: BLE001
|
|
print(json.dumps({"error": str(exc)}))
|
|
sys.exit(1)
|
|
|
|
|
|
if __name__ == "__main__":
|
|
main()
|