paradizelost/paperless-ngx
1
0
Fork 0
mirror of https://github.com/paperless-ngx/paperless-ngx.git synced 2025-08-14 00:26:21 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fixhancement: improve text thumbnail generation for large files (#10483)

Browse Source
This commit is contained in:
shamoon
2025-08-01 13:26:35 -04:00
committed by GitHub
parent f09965464a
commit 23501b9060
3 changed files with 35 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
src/paperless_text/parsers.py
View File

@@ -16,7 +16,15 @@ class TextDocumentParser(DocumentParser):
logging_name = "paperless.parsing.text"
def get_thumbnail(self, document_path: Path, mime_type, file_name=None) -> Path:
text = self.read_file_handle_unicode_errors(document_path)
# Avoid reading entire file into memory
max_chars = 100_000
file_size_limit = 50 * 1024 * 1024
if document_path.stat().st_size > file_size_limit:
text = "[File too large to preview]"
else:
with Path(document_path).open("r", encoding="utf-8", errors="replace") as f:
text = f.read(max_chars)
img = Image.new("RGB", (500, 700), color="white")
draw = ImageDraw.Draw(img)
@@ -25,7 +33,7 @@ class TextDocumentParser(DocumentParser):
size=20,
layout_engine=ImageFont.Layout.BASIC,
)
draw.text((5, 5), text, font=font, fill="black")
draw.multiline_text((5, 5), text, font=font, fill="black", spacing=4)
out_path = self.tempdir / "thumb.webp"
img.save(out_path, format="WEBP")