f2009fd991
py3: make rb output work
2019-04-29 02:03:50 -04:00
336d4054d1
DOCX Input: Fix cross-references using the "fldSimple" markup not being recognized by calibre.
...
Apparently some software in some circumstances generate field
using fldSimple rather than instrText. fldSimple is in the spec, so add
support for it to calibre's DOCX engine.
2019-04-29 10:15:21 +05:30
97b20f38f5
Fix #1822018 [Does not auto-convert when adding existing book with overwrite selected]( https://bugs.launchpad.net/calibre/+bug/1822018 )
2019-04-28 18:25:52 +05:30
5b3376c3f5
py3: Ensure cssText is of well-known type
2019-04-28 15:10:18 +05:30
e99c7fd296
Conversion: Transform styles: When using regular expressions fix substitution groups not working
...
Fixes #1825764 [Problem with regex Pattern Matching in Transform Style](https://bugs.launchpad.net/calibre/+bug/1825764 )
2019-04-28 14:13:43 +05:30
fda49f12d5
py3: Fix transform css test dialog
2019-04-28 14:07:46 +05:30
8ea5ddbbaa
Fix a regression in the previous release that caused the conversion of HTML to text in comments when output to catalogs or converting downloaded metadata to behave slightly differently. Fixes #1826654 [Comments in CSV-Catalog]( https://bugs.launchpad.net/calibre/+bug/1826654 )
2019-04-28 12:51:32 +05:30
ddbda87137
py3: more hashlib fixes
2019-04-28 12:03:47 +05:30
74c3bfa9fe
py3 fix OPDS feed generation
2019-04-28 11:48:32 +05:30
3265d1ce52
Magyar Nemzet by pofa
2019-04-28 07:35:08 +05:30
578b5f2a8e
remove no longer working recipe
2019-04-28 07:26:53 +05:30
59271d0bc2
Edit book: Fix beautify not handling <colgroup> and <th> tags correctly. Fixes #1826631 [Beautify does not work with colgroup]( https://bugs.launchpad.net/calibre/+bug/1826631 )
2019-04-27 14:15:50 +05:30
9d268ed2c9
EPUB/MOBI Catalogs: Fix a regression in the previous release that caused ampersands and middle-dots to appear as HTML entities in the output. Fixes #1826127 [special signs are ignored in catalog]( https://bugs.launchpad.net/calibre/+bug/1826127 )
...
BS 4 now auto-escape unsafe characters in HTML when creating markup. So
get rid of the explicit escaping code and use unicode symbols instead of
entities everywhere.
2019-04-27 13:17:10 +05:30
ea5b6da55a
Get rid of workaround for people running from source with obsolete binaries
2019-04-27 13:15:24 +05:30
18fd3ae063
Amazon metadata: Make title processing more robust
...
Also make it easier to run tests selectively from the command line
2019-04-27 12:05:24 +05:30
c485690a23
Update poppler
2019-04-27 09:11:33 +05:30
2983cb20e9
Fix #1826619 [GMX email setup instructions enhancement]( https://bugs.launchpad.net/calibre/+bug/1826619 )
2019-04-27 08:12:38 +05:30
330a93f001
Update more deps
2019-04-26 14:56:00 +05:30
0cc769a0b9
Update more deps
2019-04-25 17:01:51 +05:30
99b92ae390
Update more deps
2019-04-25 15:38:37 +05:30
3657252039
Merge branch 'py3' of https://github.com/eli-schwartz/calibre
2019-04-25 13:30:35 +05:30
2663085c55
py3: don't unpack dict before passing it to new defaultdict constructor
...
This was never needed, but it breaks on python3 since e.g. None cannot
be a keyword. defaultdict can just accept the dict itself, though.
2019-04-25 03:32:21 -04:00
99cb0c4417
py3: make conversion in the GUI work
2019-04-25 03:01:15 -04:00
5a3678d8c6
Prevent a malformed tweaks.json file from stopping calibre startup
2019-04-25 09:42:23 +05:30
7d6338caa8
Merge branch 'py3' of https://github.com/eli-schwartz/calibre
2019-04-25 09:37:07 +05:30
5331522226
Update versions for some deps
2019-04-25 09:33:19 +05:30
9699bcbb1d
Update Global Times
2019-04-25 09:06:26 +05:30
5ba0e29dd0
fix CLI completion for ebook-convert
...
msgpack.loads will produce mutable, unhashable lists by default when
unpacking values that were originally saved as dicts with tuple
keys. Which ebook-convert-complete does. This results in errors when
trying to reassemble the dict.
Fix by allowing our msgpack wrapper to forward arguments, and by having
ebook-convert-complete unpack lists as tuples.
2019-04-24 19:52:17 -04:00
b42963af61
Merge branch 'py3' of https://github.com/eli-schwartz/calibre
2019-04-24 12:11:36 +05:30
3972c79375
py3: port one last use of im_func
2019-04-24 02:37:29 -04:00
54ddf5ec5f
Log underlying exception when converting RTF fails
2019-04-24 12:03:12 +05:30
8f829e6de6
Update Global Times
...
Fixes #1825937 [can't fetch news of globaltimes](https://bugs.launchpad.net/calibre/+bug/1825937 )
2019-04-24 11:21:47 +05:30
4b8d853999
http -> https
2019-04-24 05:29:51 +05:30
be27f464b2
Update Ars Technica
2019-04-23 20:46:38 +05:30
c12c80e174
py3: More fixes for news downloads
2019-04-23 16:04:20 +05:30
bfbc31fa9f
Merge branch 'py3' of https://github.com/eli-schwartz/calibre
2019-04-23 15:55:08 +05:30
b9224f17c4
py3: basic news fetching works
...
Just some more juggling around of bytes types. Specifically note that
urls are encoded and quoted, then decoded before being given to
mechanize as mechanize expects to see unicode. Furthermore,
ascii_filename is already there to sanitize filenames.
2019-04-23 03:31:46 -04:00
84d939fb06
py3: Avoid chained exception
2019-04-23 12:17:34 +05:30
bc661a812d
News fetching: Fix a regression that broke removal of attributes/javascript/srcset because of bugs in the new BeautifulSoup
2019-04-23 12:15:47 +05:30
7270eaa91f
Apparently python 3.4+ removes the __func__ attribute, sigh
2019-04-23 11:00:49 +05:30
ac2a83d82a
oops
2019-04-23 10:49:21 +05:30
a6eb1d79ea
py3: Replace useage of im_func
2019-04-23 10:46:58 +05:30
f4cf168a72
plugins_mirror should not rely on getcwdu
2019-04-23 10:34:37 +05:30
22b1a4843b
Merge branch 'py3' of https://github.com/eli-schwartz/calibre
2019-04-23 10:30:19 +05:30
bca47af433
py3: make Polish Book work
...
str() template cannot be formatted with an encoded bytestring because
b''.__str__() is unuseful. So just defer encoding to right before
writing.
2019-04-23 00:49:42 -04:00
adbe76079e
py3: make Check Book work
...
when working with lxml.etree strings, everything is explicitly bytes
2019-04-23 00:49:42 -04:00
570b02c59b
Update Corriere della Serra
2019-04-23 10:14:17 +05:30
76bcf5ca41
Merge branch 'py3' of https://github.com/eli-schwartz/calibre
2019-04-23 09:56:15 +05:30
4fc124adb0
py3: also make plugin installer work
...
We check if versions compare favorably to the installed version, and if
it None, then we report that it can be installed. But this check always
returns None < available_version on python2, and on python3 it is a
comparison error -- so lift out the None handling to return as early as
possible.
2019-04-22 21:36:57 -04:00
eab2c99b60
py3: make plugin updater work
...
calibre's internal version representation is a tuple. On python3, it no
longer suffices to compare this to json.loads() list-based
representations.
Also update one instance of map where we need something subscriptable.
2019-04-22 13:34:01 -04:00
Eli Schwartz
Kovid Goyal