AI/ragflow
1
0
Fork 0
mirror of https://git.mirrors.martin98.com/https://github.com/infiniflow/ragflow.git synced 2025-06-04 11:24:00 +08:00
Code Issues Packages Projects Releases Wiki Activity

Ensure thumbnail be smaller than 64K (#3722)

Browse Source 
### What problem does this PR solve?

Ensure thumbnail be smaller than 64K. Close #1443
### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

---------

Co-authored-by: Kevin Hu <kevinhu.sh@gmail.com>
This commit is contained in:
Zhichang Yu 2024-11-28 19:15:31 +08:00 committed by GitHub
parent 91f1814a87
commit 112ef42a19
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 30 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

35
api/utils/file_utils.py
View File

@ -170,29 +170,52 @@ def filename_type(filename):
return FileType.OTHER.value return FileType.OTHER.value
def thumbnail_img(filename, blob): def thumbnail_img(filename, blob):
"""
MySQL LongText max length is 65535
"""
filename = filename.lower() filename = filename.lower()
if re.match(r".*\.pdf$", filename): if re.match(r".*\.pdf$", filename):
pdf = pdfplumber.open(BytesIO(blob)) pdf = pdfplumber.open(BytesIO(blob))
buffered = BytesIO() buffered = BytesIO()
pdf.pages[0].to_image(resolution=32).annotated.save(buffered, format="png") resolution = 32
return buffered.getvalue() img = None
for _ in range(10):
# https://github.com/jsvine/pdfplumber?tab=readme-ov-file#creating-a-pageimage-with-to_image
pdf.pages[0].to_image(resolution=resolution).annotated.save(buffered, format="png")
img = buffered.getvalue()
if len(img) >= 64000 and resolution >= 2:
resolution = resolution / 2
buffered = BytesIO()
else:
break
return img
if re.match(r".*\.(jpg|jpeg|png|tif|gif|icon|ico|webp)$", filename): elif re.match(r".*\.(jpg|jpeg|png|tif|gif|icon|ico|webp)$", filename):
image = Image.open(BytesIO(blob)) image = Image.open(BytesIO(blob))
image.thumbnail((30, 30)) image.thumbnail((30, 30))
buffered = BytesIO() buffered = BytesIO()
image.save(buffered, format="png") image.save(buffered, format="png")
return buffered.getvalue() return buffered.getvalue()
if re.match(r".*\.(ppt|pptx)$", filename): elif re.match(r".*\.(ppt|pptx)$", filename):
import aspose.slides as slides import aspose.slides as slides
import aspose.pydrawing as drawing import aspose.pydrawing as drawing
try: try:
with slides.Presentation(BytesIO(blob)) as presentation: with slides.Presentation(BytesIO(blob)) as presentation:
buffered = BytesIO() buffered = BytesIO()
presentation.slides[0].get_thumbnail(0.03, 0.03).save( scale = 0.03
img = None
for _ in range(10):
# https://reference.aspose.com/slides/python-net/aspose.slides/slide/get_thumbnail/#float-float
presentation.slides[0].get_thumbnail(scale, scale).save(
buffered, drawing.imaging.ImageFormat.png) buffered, drawing.imaging.ImageFormat.png)
return buffered.getvalue() img = buffered.getvalue()
if len(img) >= 64000:
scale = scale / 2.0
buffered = BytesIO()
else:
break
return img
except Exception: except Exception:
pass pass
return None return None