Paul Tremberth
14f5817d6b
Modify ItemLoader to support XPath and CSS selectors
...
Deprecate XPathItemLoader (now an alias to the new ItemLoader)
2013-11-21 18:05:24 +01:00
Pablo Hoffman
f87be371a2
better names for HANDLE_* settings, and added doc
2013-11-21 14:33:17 -02:00
Daniel Graña
ab01e9e9e4
Merge pull request #466 from kalessin/httperror
...
allow to use settings for defining http error handling defaults
2013-11-21 03:55:44 -08:00
Martin Olveyra
55bee912a2
allow to use settings for defining http error handling defaults
2013-11-20 20:12:49 -02:00
Mikhail Korobov
8416cc7515
Merge pull request #465 from bjlange/master
...
Add note to item-pipeline documentation explaining order
2013-11-20 09:40:19 -08:00
Brian Lange
e4c1d8d37d
Elaborate on use of order numbers
2013-11-19 17:51:50 -06:00
Daniel Graña
2564c21d4c
add a tox env for Python 3.3
2013-11-19 20:15:15 -02:00
Daniel Graña
526a944eda
lxml is required, no need to skip tests.
2013-11-19 20:14:48 -02:00
Daniel Graña
3f156ad845
Do not call body_as_unicode on non text responses. closes #462
2013-11-19 20:13:34 -02:00
Brian Lange
b878f60b5a
Add note to item-pipeline documentation explaining order in the ITEM_PIPELINES setting.
2013-11-19 16:12:54 -06:00
Daniel Graña
ec7833a910
Deprecate body_or_str helper function only used by xml iterators
2013-11-19 19:21:54 -02:00
Pablo Hoffman
2d91c7136d
Merge pull request #464 from kalessin/telnet
...
telnet client: fix unexisting reference to engine.slots
2013-11-19 05:22:58 -08:00
olveyra
755b9ba5a4
telnet client: fix unexisting reference to engine.slots
2013-11-19 04:52:24 +01:00
Pablo Hoffman
afe6eaa2fe
Merge pull request #460 from tntC4stl3/master
...
duplicate 'use' in line 87
2013-11-15 04:10:49 -08:00
tntC4stl3
b51d5d81e4
duplicate 'use' in line 87
2013-11-15 13:56:44 +08:00
Daniel Graña
c74903f9da
process_parallel was leaking the failures on its internal deferreds. closes #458
...
DeferredList implemented cancellation in Twisted 13.2.0 by holding a
reference to the affected deferreds objects, if a deferred errored the
result was propagated to the DeferredList but still referenced by the
original deferred and nobody was consuming it.
The tests started to fail because the reference from DeferredList
prevented the underlining deferred from been collected before the test
finish invalidating the effect of self.flushedLoggedErrors() call.
2013-11-09 02:12:52 -02:00
Daniel Graña
04ff7ecebf
improve 0.20 release notes
...
Conflicts:
docs/news.rst
2013-11-08 17:45:03 -02:00
Daniel Graña
3d18a3c49e
bumped version to 0.21.0
0.21.0
2013-11-08 17:09:00 -02:00
Daniel Graña
d0980e5c9b
Merge 0.20 release notes
2013-11-08 17:06:10 -02:00
Daniel Graña
60516123a5
Merge branch 'travix-toxed'
...
Conflicts:
requirements.txt
2013-11-07 11:29:58 -02:00
Daniel Graña
d29791d7ab
building Pillow with pypy requries dev headers
2013-11-07 10:58:08 -02:00
Daniel Graña
971f60d796
TOXENV is tox supported env
2013-11-07 10:58:08 -02:00
Daniel Graña
ecfa743105
install updated pypy from ppa
2013-11-07 10:58:08 -02:00
Daniel Graña
fabb351097
map travis-ci matrix to tox environments
2013-11-07 10:58:08 -02:00
Daniel Graña
bc7fa61136
Django 1.6 form validation errors now include ValidationError exception instances instead of just strings
2013-11-07 10:36:39 -02:00
Rolando Espinoza La fuente
6f5423aebd
Replaced remaning __import__(module) calls.
...
This commit replaces the statements __import__(module) as the previous
replaced the statements __import__(module, {}, {}, ['']).
At first I thought leaving the single-argument calls, but perhaps it's
better to be strict rather than having exceptions to the rule in this
case.
2013-11-07 10:36:39 -02:00
Rolando Espinoza La fuente
6b1760d7a1
replaced __import__ by importlib.import_module.
...
Since python 2.7, importlib.import_module is the recommended way to
import modules programmatically.
From __import__'s doc:
Import a module. Because this function is meant for use by the
Python
interpreter and not for general use it is better to use
importlib.import_module() to programmatically import a module.
2013-11-07 10:36:39 -02:00
Daniel Graña
b6bed44c2b
Django 1.6 form validation errors now include ValidationError exception instances instead of just strings
2013-11-07 02:32:10 -02:00
Daniel Graña
b78e76108f
Merge pull request #445 from darkrho/import-module
...
Use `importlib.import_module` instead of `__import__`
2013-11-06 20:14:27 -08:00
Daniel Graña
2318c56f14
shutdown the active crawler on SIGINT. fixes #450
2013-11-05 00:02:17 -02:00
Mikhail Korobov
f80f10ae7e
Merge pull request #452 from alexanderlukanin13/python3
...
PY3: scrapy.__version__, NoneType, urlparse_monkeypatches
2013-11-04 00:54:02 -08:00
alexanderlukanin13
6c7292a08e
python3: scrapy.__version__, NoneType, urlparse_monkeypatches
2013-11-03 22:20:33 +06:00
Rolando Espinoza La fuente
d1b912890b
Merge branch 'master' into import-module
2013-10-30 21:07:53 -04:00
Pablo Hoffman
fa245af6d2
Merge pull request #448 from dangra/drop-py26
...
Drop Python 2.6 support
2013-10-30 05:53:06 -07:00
Daniel Graña
2df8156431
Drop Python 2.6 support
2013-10-29 13:44:00 -02:00
Rolando Espinoza La fuente
10e22aa5fb
Replaced remaning __import__(module) calls.
...
This commit replaces the statements __import__(module) as the previous
replaced the statements __import__(module, {}, {}, ['']).
At first I thought leaving the single-argument calls, but perhaps it's
better to be strict rather than having exceptions to the rule in this
case.
2013-10-27 19:10:25 -04:00
Rolando Espinoza La fuente
343f997ed6
replaced __import__ by importlib.import_module.
...
Since python 2.7, importlib.import_module is the recommended way to
import modules programmatically.
From __import__'s doc:
Import a module. Because this function is meant for use by the
Python
interpreter and not for general use it is better to use
importlib.import_module() to programmatically import a module.
2013-10-27 18:33:51 -04:00
Daniel Graña
bd79b6e1d3
debian package requires python-cssselect
2013-10-24 16:47:46 -02:00
Mikhail Korobov
247b2ad596
Merge pull request #431 from alexanderlukanin13/syntax
...
Python 3 compatible syntax: print, except, raise, octal numbers; removed...
2013-10-22 10:38:45 -07:00
alexanderlukanin13
6b598476eb
print_function in xlib
2013-10-22 22:49:18 +06:00
Mikhail Korobov
3d4904be16
Merge pull request #436 from alexanderlukanin13/syntax-next
...
Using next() for Python 3 compatibility; ordereddict.py converted from DOS to Unix
2013-10-22 08:35:09 -07:00
Mikhail Korobov
bf6ef29182
Merge pull request #435 from alexanderlukanin13/syntax-map
...
Converted map() to list comprehension/generator for Python3 compatibility
2013-10-22 07:37:57 -07:00
alexanderlukanin13
a958cb2c5d
Using next() for Python 3 compatibility
2013-10-22 13:54:22 +06:00
alexanderlukanin13
6b35166de6
Converted map() to list comprehension or generator for Python 3 compatibility
2013-10-22 13:18:26 +06:00
Pablo Hoffman
911c8082b0
simplified description of crawl command
2013-10-21 14:42:51 -02:00
Pablo Hoffman
e8ee449a2a
Merge pull request #432 from darkrho/crawl-url
...
Removed URL reference in crawl command and .tld suffix in docs for spider names
2013-10-21 09:40:58 -07:00
alexanderlukanin13
ae4a61a2d5
Added "from __future__ import print_function" in all relevant places except xlib
2013-10-20 11:55:07 +06:00
Rolando Espinoza La fuente
34543c2b2e
DOCS removed .tld suffix for spider names for the sake of consistency.
2013-10-19 23:03:20 -04:00
Rolando Espinoza La fuente
c2ced9a234
Removed reference to URL in crawl command as it's no longer supported.
2013-10-19 22:57:37 -04:00
alexanderlukanin13
d381a35732
Python 3 compatible syntax: print, except, raise, octal numbers; removed Python 2.2 boolean compatibility code in xlib/pydispatch/dispatcher.py
2013-10-20 01:59:35 +06:00