almet/dangerzone
1
0
Fork 0
mirror of https://github.com/freedomofpress/dangerzone.git synced 2025-04-28 18:02:38 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 11

Replace 'convert' with PyMuPDF for images

Browse source 
PyMuPDF can also convert images of the types we already support so we
don't need ImageMagick's 'convert'.
This commit is contained in:
deeplow 2023-11-16 11:23:55 +00:00
parent a3a64882a3
commit e5dbe25abb
No known key found for this signature in database
GPG key ID: 577982871529A52A
1 changed files with 7 additions and 25 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

32
dangerzone/conversion/doc_to_pixels.py
View file

@ -146,14 +146,14 @@ class DocumentToPixels(DangerzoneConverter):
"type": "libreoffice", "type": "libreoffice",
}, },
# .jpg # .jpg
"image/jpeg": {"type": "convert"}, "image/jpeg": {"type": None},
# .gif # .gif
"image/gif": {"type": "convert"}, "image/gif": {"type": None},
# .png # .png
"image/png": {"type": "convert"}, "image/png": {"type": None},
# .tif # .tif
"image/tiff": {"type": "convert"}, "image/tiff": {"type": None},
"image/x-tiff": {"type": "convert"}, "image/x-tiff": {"type": None},
} }
# Detect MIME type # Detect MIME type
@ -182,7 +182,7 @@ class DocumentToPixels(DangerzoneConverter):
# Convert input document to PDF # Convert input document to PDF
conversion = conversions[mime_type] conversion = conversions[mime_type]
if conversion["type"] is None: if conversion["type"] is None:
pdf_filename = "/tmp/input_file" doc = fitz.open("/tmp/input_file", filetype=mime_type)
elif conversion["type"] == "libreoffice": elif conversion["type"] == "libreoffice":
libreoffice_ext = conversion.get("libreoffice_ext", None) libreoffice_ext = conversion.get("libreoffice_ext", None)
# Disable conversion for HWP/HWPX on specific platforms. See: # Disable conversion for HWP/HWPX on specific platforms. See:
@ -220,24 +220,7 @@ class DocumentToPixels(DangerzoneConverter):
# https://github.com/freedomofpress/dangerzone/issues/494 # https://github.com/freedomofpress/dangerzone/issues/494
if not os.path.exists(pdf_filename): if not os.path.exists(pdf_filename):
raise errors.LibreofficeFailure() raise errors.LibreofficeFailure()
elif conversion["type"] == "convert": doc = fitz.open(pdf_filename)
self.update_progress("Converting to PDF using GraphicsMagick")
args = [
"gm",
"convert",
"/tmp/input_file",
"/tmp/input_file.pdf",
]
await self.run_command(
args,
error_message="Conversion to PDF with GraphicsMagick failed",
timeout_message=(
"Error converting document to PDF, GraphicsMagick timed out after"
f" {timeout} seconds"
),
timeout=timeout,
)
pdf_filename = "/tmp/input_file.pdf"
else: else:
raise errors.InvalidGMConversion( raise errors.InvalidGMConversion(
f"Invalid conversion type {conversion['type']} for MIME type {mime_type}" f"Invalid conversion type {conversion['type']} for MIME type {mime_type}"
@ -245,7 +228,6 @@ class DocumentToPixels(DangerzoneConverter):
self.percentage += 3 self.percentage += 3
# Obtain number of pages # Obtain number of pages
doc = fitz.open(pdf_filename)
if doc.page_count > errors.MAX_PAGES: if doc.page_count > errors.MAX_PAGES:
raise errors.MaxPagesException() raise errors.MaxPagesException()
await self.write_page_count(doc.page_count) await self.write_page_count(doc.page_count)