paradizelost/paperless-ngx
1
0
Fork 0
mirror of https://github.com/paperless-ngx/paperless-ngx.git synced 2025-04-02 13:45:10 -05:00
Code Issues Packages Projects Releases Wiki Activity

code review changes

Browse Source 
Signed-off-by: Florian Brandes <florian.brandes@posteo.de>
This commit is contained in:
Florian Brandes 2022-04-20 00:55:45 +02:00
parent bf57b6e4a2
commit a7b1658ee1
No known key found for this signature in database
GPG Key ID: 074048E893713170
2 changed files with 30 additions and 36 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

63
src/documents/tasks.py
View File

@ -108,23 +108,22 @@ def convert_from_tiff_to_pdf(filepath: str) -> str:
newpath = os.path.join(tempdir, file_name + ".pdf") newpath = os.path.join(tempdir, file_name + ".pdf")
else: else:
logger.warning(f"Cannot convert from {str(file_extension)} to pdf.") logger.warning(f"Cannot convert from {str(file_extension)} to pdf.")
return "" return None
image = Image.open(filepath) with Image.open(filepath) as image:
images = [] images = []
for i, page in enumerate(ImageSequence.Iterator(image)): for i, page in enumerate(ImageSequence.Iterator(image)):
page = page.convert("RGB") page = page.convert("RGB")
images.append(page) images.append(page)
try: try:
if len(images) == 1: if len(images) == 1:
images[0].save(newpath) images[0].save(newpath)
else: else:
images[0].save(newpath, save_all=True, append_images=images[1:]) images[0].save(newpath, save_all=True, append_images=images[1:])
except OSError as e: except OSError as e:
logger.warning( logger.warning(
f"Could not save the file as pdf. Error: {str(e)}", f"Could not save the file as pdf. Error: {str(e)}",
) )
return "" return None
image.close()
return newpath return newpath
@ -242,23 +241,19 @@ def consume_file(
f"Unsupported file format for barcode reader: {str(file_extension)}", f"Unsupported file format for barcode reader: {str(file_extension)}",
) )
else: else:
if file_extension == ".tif" or file_extension == ".tiff": if file_extension in {".tif", ".tiff"}:
converted_tiff = convert_from_tiff_to_pdf(path) file_to_process = convert_from_tiff_to_pdf(path)
if converted_tiff:
separators = scan_file_for_separating_barcodes(converted_tiff)
else: else:
separators = scan_file_for_separating_barcodes(path) file_to_process = path
separators = scan_file_for_separating_barcodes(file_to_process)
if separators: if separators:
if converted_tiff: logger.debug(
logger.debug( f"Pages with separators found in: {str(path)}",
f"Pages with separators found in: {str(converted_tiff)}", )
) document_list = separate_pages(file_to_process, separators)
document_list = separate_pages(converted_tiff, separators)
else:
logger.debug(
f"Pages with separators found in: {str(path)}",
)
document_list = separate_pages(path, separators)
if document_list: if document_list:
for n, document in enumerate(document_list): for n, document in enumerate(document_list):
# save to consumption dir # save to consumption dir
@ -271,8 +266,8 @@ def consume_file(
# if we got here, the document was successfully split # if we got here, the document was successfully split
# and can safely be deleted # and can safely be deleted
if converted_tiff: if converted_tiff:
logger.debug("Deleting file {}".format(converted_tiff)) logger.debug("Deleting file {}".format(file_to_process))
os.unlink(converted_tiff) os.unlink(file_to_process)
logger.debug("Deleting file {}".format(path)) logger.debug("Deleting file {}".format(path))
os.unlink(path) os.unlink(path)
# notify the sender, otherwise the progress bar # notify the sender, otherwise the progress bar

3
src/documents/tests/test_tasks.py
View File

@ -225,8 +225,7 @@ class TestTasks(DirectoriesMixin, TestCase):
) )
dst = os.path.join(settings.SCRATCH_DIR, "simple.pdf") dst = os.path.join(settings.SCRATCH_DIR, "simple.pdf")
shutil.copy(test_file, dst) shutil.copy(test_file, dst)
target_file = tasks.convert_from_tiff_to_pdf(dst) self.assertIsNone(tasks.convert_from_tiff_to_pdf(dst))
self.assertFalse(os.path.isfile(target_file))
def test_scan_file_for_separating_barcodes(self): def test_scan_file_for_separating_barcodes(self):
test_file = os.path.join( test_file = os.path.join(