0x2620/pandora
2
1
Fork 1
Code Issues 1.2k Pull requests Releases Wiki Activity

support more image formats, server avif is supported for larger images

Browse source
This commit is contained in:
j 2025-01-29 19:44:28 +05:30
commit 406237b837
9 changed files with 31 additions and 11 deletions
Download patch file Download diff file Expand all files Collapse all files

3
pandora/archive/extract.py
View file

@ -19,10 +19,13 @@ import ox.image
from ox.utils import json
from django.conf import settings
from PIL import Image, ImageOps
import pillow_avif
from pillow_heif import register_heif_opener
from .chop import Chop, make_keyframe_index
register_heif_opener()
logger = logging.getLogger('pandora.' + __name__)
img_extension = 'jpg'

5
pandora/document/fulltext.py
View file

@ -8,6 +8,7 @@ from django.conf import settings
logger = logging.getLogger('pandora.' + __name__)
IMAGE_EXTENSIONS = ('png', 'jpg', 'webp', 'heic', 'heif')
def extract_text(pdf, page=None):
if page is not None:
@ -53,7 +54,7 @@ class FulltextMixin:
if self.file:
if self.extension == 'pdf':
return extract_text(self.file.path)
elif self.extension in ('png', 'jpg'):
elif self.extension in IMAGE_EXTENSIONS:
return ocr_image(self.file.path)
elif self.extension == 'html':
return self.data.get('text', '')
@ -180,7 +181,7 @@ class FulltextPageMixin(FulltextMixin):
if self.document.file:
if self.document.extension == 'pdf':
return extract_text(self.document.file.path, self.page)
elif self.extension in ('png', 'jpg'):
elif self.extension in IMAGE_EXTENSIONS:
return ocr_image(self.document.file.path)
elif self.extension == 'html':
# FIXME: is there a nice way to split that into pages

11
pandora/document/models.py
View file

@ -521,14 +521,17 @@ class Document(models.Model, FulltextMixin):
return save_chunk(self, self.file, chunk, offset, name, done_cb)
return False, 0
def thumbnail(self, size=None, page=None):
def thumbnail(self, size=None, page=None, accept=None):
if not self.file:
return os.path.join(settings.STATIC_ROOT, 'png/document.png')
src = self.file.path
folder = os.path.dirname(src)
if size:
size = int(size)
path = os.path.join(folder, '%d.jpg' % size)
ext = 'jpg'
if accept and 'image/avif' in accept and size > 512:
ext = 'avif'
path = os.path.join(folder, '%d.%s' % (size, ext))
else:
path = src
if self.extension == 'pdf':
@ -561,7 +564,7 @@ class Document(models.Model, FulltextMixin):
path = os.path.join(folder, '%dp%d,%s.jpg' % (size, page, ','.join(map(str, crop))))
if not os.path.exists(path):
resize_image(src, path, size=size)
elif self.extension in ('jpg', 'png', 'gif'):
elif self.extension in ('jpg', 'png', 'gif', 'webp', 'heic', 'heif'):
if os.path.exists(src):
if size and page:
crop = list(map(int, page.split(',')))
@ -574,7 +577,7 @@ class Document(models.Model, FulltextMixin):
img = open_image_rgb(path)
src = path
if size < max(img.size):
path = os.path.join(folder, '%sp%s.jpg' % (size, ','.join(map(str, crop))))
path = os.path.join(folder, '%sp%s.%s' % (size, ','.join(map(str, crop)), ext))
if not os.path.exists(path):
resize_image(src, path, size=size)
if os.path.exists(src) and not os.path.exists(path):

9
pandora/document/views.py
View file

@ -378,15 +378,22 @@ actions.register(sortDocuments, cache=False)
def file(request, id, name=None):
document = get_document_or_404_json(request, id)
accept = request.headers.get("Accept")
if accept and 'image/' in accept and document.extension in (
'webp', 'heif', 'heic', 'avif'
) and document.extension not in accept:
image_size = max(document.width, document.height)
return HttpFileResponse(document.thumbnail(image_size))
return HttpFileResponse(document.file.path)
def thumbnail(request, id, size=256, page=None):
size = int(size)
document = get_document_or_404_json(request, id)
accept = request.headers.get("Accept")
if "q" in request.GET and page:
img = document.highlight_page(page, request.GET["q"], size)
return HttpResponse(img, content_type="image/jpeg")
return HttpFileResponse(document.thumbnail(size, page=page))
return HttpFileResponse(document.thumbnail(size, page=page, accept=accept))
@login_required_json

2
pandora/urls.py
View file

@ -53,7 +53,7 @@ urlpatterns += [
re_path(r'^resetUI$', user.views.reset_ui),
re_path(r'^collection/(?P<id>.*?)/icon(?P<size>\d*).jpg$', documentcollection.views.icon),
re_path(r'^documents/(?P<id>[A-Z0-9]+)/(?P<size>\d*)p(?P<page>[\d,]*).jpg$', document.views.thumbnail),
re_path(r'^documents/(?P<id>[A-Z0-9]+)/(?P<name>.*?\.[^\d]{3})$', document.views.file),
re_path(r'^documents/(?P<id>[A-Z0-9]+)/(?P<name>.*?\.[^\d]{3,4})$', document.views.file),
re_path(r'^documents/(?P<fragment>.*?)$', document.views.document),
re_path(r'^edit/(?P<id>.*?)/icon(?P<size>\d*).jpg$', edit.views.icon),
re_path(r'^list/(?P<id>.*?)/icon(?P<size>\d*).jpg$', itemlist.views.icon),