John Vandenberg has uploaded a new change for review.
https://gerrit.wikimedia.org/r/179071
Change subject: Use cmtype to restrict page type for Cat.articles
......................................................................
Use cmtype to restrict page type for Cat.articles
Bug: T67981
Change-Id: I46a53497a86c3ab3761f9166b6c268c7a3c5f6be
---
M pywikibot/page.py
M pywikibot/site.py
2 files changed, 50 insertions(+), 9 deletions(-)
git pull ssh://gerrit.wikimedia.org:29418/pywikibot/core
refs/changes/71/179071/1
diff --git a/pywikibot/page.py b/pywikibot/page.py
index 192d3e1..8c42050 100644
--- a/pywikibot/page.py
+++ b/pywikibot/page.py
@@ -49,7 +49,8 @@
SiteDefinitionError
)
from pywikibot.tools import (
- ComparableMixin, deprecated, deprecate_arg, deprecated_args,
+ ComparableMixin, MediaWikiVersion as LV,
+ deprecated, deprecate_arg, deprecated_args,
remove_last_args
)
from pywikibot import textlib
@@ -2062,7 +2063,7 @@
@deprecate_arg("startFrom", "startsort")
def articles(self, recurse=False, step=None, total=None,
- content=False, namespaces=None, sortby="",
+ content=False, namespaces=None, sortby=None,
starttime=None, endtime=None, startsort=None,
endsort=None):
"""
@@ -2102,9 +2103,6 @@
@type endsort: str
"""
- if namespaces is None:
- namespaces = [x for x in self.site.namespaces()
- if x >= 0 and x != 14]
for member in self.site.categorymembers(self,
namespaces=namespaces,
step=step, total=total,
@@ -2112,7 +2110,8 @@
starttime=starttime,
endtime=endtime,
startsort=startsort,
- endsort=endsort
+ endsort=endsort,
+ member_type=['page', 'file']
):
yield member
if total is not None:
diff --git a/pywikibot/site.py b/pywikibot/site.py
index 3c2c782..82427be 100644
--- a/pywikibot/site.py
+++ b/pywikibot/site.py
@@ -2788,10 +2788,10 @@
step=step, total=total, g_content=content)
return tlgen
- def categorymembers(self, category, namespaces=None, sortby="",
+ def categorymembers(self, category, namespaces=None, sortby=None,
reverse=False, starttime=None, endtime=None,
startsort=None, endsort=None, step=None, total=None,
- content=False):
+ content=False, member_type=None):
"""Iterate members of specified category.
@param category: The Category to iterate.
@@ -2822,7 +2822,9 @@
@type endsort: str
@param content: if True, load the current content of each iterated page
(default False)
-
+ @type content: bool
+ @param member_type: member type
+ @type member_type: str or iterable of str; values: page, subcat, file
"""
if category.namespace() != 14:
raise Error(
@@ -2844,6 +2846,46 @@
if startsort and endsort and startsort > endsort:
raise ValueError(
"categorymembers: startsort must be less than endsort")
+
+ assert(not member_type or sortby != 'timestamp')
+ if isinstance(member_type, basestring):
+ member_type = set([member_type])
+ if member_type and LV(self.site.version()) < LV("1.12"):
+ # Retrofit cmtype/member_type, available on WM API 1.12+,
+ # to use namespaces available on earlier versions.
+ if 'page' in member_type:
+ excluded_namespaces = set()
+ if 'file' not in member_type:
+ excluded_namespaces.add(6)
+ if 'subcat' not in member_type:
+ excluded_namespaces.add(14)
+
+ if namespaces:
+ for exclude_ns in excluded_namespaces:
+ if exclude_ns in namespaces:
+ raise ValueError(
+ 'incompatible namespaces %r and member_type %r'
+ % (namespaces, member_type))
+ else:
+ # if the number of namespaces is greater than permitted by
+ # the API, it will issue a warning and use the namespaces
+ # up until the limit, which will usually be sufficient.
+ # TODO: QueryGenerator should detect when the number of
+ # namespaces requested is higher than available, and split
+ # the request into several batches.
+ namespaces = [x for x in self.namespaces()
+ if x >= 0 and x not in excluded_namespaces]
+ else:
+ namespaces = []
+ if 'file' in member_type:
+ namespaces.append(6)
+ if 'subcat' in member_type:
+ namespaces.append(14)
+
+ member_type = None
+ if member_type:
+ cmargs['gcmtype'] = member_type
+
if reverse:
cmargs["gcmdir"] = "desc"
# API wants start/end params in opposite order if using descending
--
To view, visit https://gerrit.wikimedia.org/r/179071
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: newchange
Gerrit-Change-Id: I46a53497a86c3ab3761f9166b6c268c7a3c5f6be
Gerrit-PatchSet: 1
Gerrit-Project: pywikibot/core
Gerrit-Branch: master
Gerrit-Owner: John Vandenberg <jay...@gmail.com>
_______________________________________________
MediaWiki-commits mailing list
MediaWiki-commits@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
