Fix DPI mismatch between doc2pixels and pixels2pdf

The original document was larger in dimensions than the original one due
to a mismatch in DPI settings. When converting documents to pixels we
were setting the DPI to 150 pixels per inch. Then when converting back
into a PDF we were using 70 DPI. This difference would result in an
overall larger document in dimensions (though not necessarily in file
size).

Fixes #626

dangerzone/conversion/common.py

@@ -15,6 +15,7 @@
 TIMEOUT_PER_PAGE: float = 30  # (seconds)
 TIMEOUT_PER_MB: float = 30  # (seconds)
 TIMEOUT_MIN: float = 60  # (seconds)
+DEFAULT_DPI = 150  # Pixels per inch
 
 
 def running_on_qubes() -> bool:

dangerzone/conversion/doc_to_pixels.py

@@ -20,7 +20,7 @@
 import magic
 
 from . import errors
-from .common import DangerzoneConverter, running_on_qubes
+from .common import DEFAULT_DPI, DangerzoneConverter, running_on_qubes
 
 
 class DocumentToPixels(DangerzoneConverter):
@@ -253,7 +253,7 @@ async def convert(self) -> None:
             self.update_progress(
                 f"Converting page {page_num}/{doc.page_count} to pixels"
             )
-            pix = page.get_pixmap(dpi=150)
+            pix = page.get_pixmap(dpi=DEFAULT_DPI)
             rgb_buf = pix.samples
             await self.write_page_width(pix.width, width_filename)
             await self.write_page_height(pix.height, height_filename)

dangerzone/conversion/pixels_to_pdf.py

@@ -15,7 +15,7 @@
 
 import fitz
 
-from .common import DangerzoneConverter, get_tessdata_dir, running_on_qubes
+from .common import DEFAULT_DPI, DangerzoneConverter, get_tessdata_dir, running_on_qubes
 
 
 class PixelsToPDF(DangerzoneConverter):
@@ -56,6 +56,7 @@ async def convert(
             pixmap = fitz.Pixmap(
                 fitz.Colorspace(fitz.CS_RGB), width, height, untrusted_rgb_data, False
             )
+            pixmap.set_dpi(DEFAULT_DPI, DEFAULT_DPI)
             if ocr_lang:  # OCR the document
                 try:
                     ocr_pdf_bytes = pixmap.pdfocr_tobytes(