Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-09 03:04:10 -04:00
Code Issues Packages Projects Releases Wiki Activity

News download: Add support for <img> tags that link to SVG images. Fixes #1031553 (web2disk doesn't support SVG images)

Browse Source
This commit is contained in:
Kovid Goyal 2012-08-01 13:28:49 +05:30
parent 0d843a89ea
commit 2848314bf9
1 changed files with 16 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
src/calibre/web/fetch/simple.py
View File

@ -7,7 +7,7 @@ __copyright__ = '2008, Kovid Goyal <kovid at kovidgoyal.net>'
Fetch a webpage and its links recursively. The webpages are saved to disk in Fetch a webpage and its links recursively. The webpages are saved to disk in
UTF-8 encoding with any charset declarations removed. UTF-8 encoding with any charset declarations removed.
''' '''
import sys, socket, os, urlparse, re, time, copy, urllib2, threading, traceback import sys, socket, os, urlparse, re, time, copy, urllib2, threading, traceback, imghdr
from urllib import url2pathname, quote from urllib import url2pathname, quote
from httplib import responses from httplib import responses
from PIL import Image from PIL import Image
@ -375,16 +375,25 @@ class RecursiveFetcher(object):
if isinstance(fname, unicode): if isinstance(fname, unicode):
fname = fname.encode('ascii', 'replace') fname = fname.encode('ascii', 'replace')
imgpath = os.path.join(diskpath, fname+'.jpg') imgpath = os.path.join(diskpath, fname+'.jpg')
try: if (imghdr.what(None, data) is None and b'<svg' in data[:1024]):
im = Image.open(StringIO(data)).convert('RGBA') # SVG image
imgpath = os.path.join(diskpath, fname+'.svg')
with self.imagemap_lock: with self.imagemap_lock:
self.imagemap[iurl] = imgpath self.imagemap[iurl] = imgpath
with open(imgpath, 'wb') as x: with open(imgpath, 'wb') as x:
im.save(x, 'JPEG') x.write(data)
tag['src'] = imgpath tag['src'] = imgpath
except: else:
traceback.print_exc() try:
continue im = Image.open(StringIO(data)).convert('RGBA')
with self.imagemap_lock:
self.imagemap[iurl] = imgpath
with open(imgpath, 'wb') as x:
im.save(x, 'JPEG')
tag['src'] = imgpath
except:
traceback.print_exc()
continue
def absurl(self, baseurl, tag, key, filter=True): def absurl(self, baseurl, tag, key, filter=True):
iurl = tag[key] iurl = tag[key]