Revision: 5923
Author: cydeweys
Date: 2008-09-29 02:58:20 +0000 (Mon, 29 Sep 2008)
Log Message:
-----------
Getting rid of the obsolete catmove.pl Perl script (probably not suitable
for a Python bot anyway) and replacing it with a more capable bot written
purely in Python. This is the same code running Cydebot on the English
Wikipedia and should be easily adaptable for other languages' wikis if
they have the same concept of a CFD working page.
Added Paths:
-----------
trunk/pywikipedia/cfd.py
Removed Paths:
-------------
trunk/pywikipedia/catmove.pl
Deleted: trunk/pywikipedia/catmove.pl
===================================================================
--- trunk/pywikipedia/catmove.pl 2008-09-28 17:59:40 UTC (rev 5922)
+++ trunk/pywikipedia/catmove.pl 2008-09-29 02:58:20 UTC (rev 5923)
@@ -1,86 +0,0 @@
-#!/usr/bin/perl
-# $Id$
-
-# This Perl script takes a list of category moves or removes to make and uses
category.py.
-# The input format is as follows:
-# * Category:US to Category:United States
-# OR
-# * Category:US
-# Just copy-paste the list of categories to move or remove from the Wiki, put them in a
-# text-file, and redirect to stdin, i.e., perl catmove.pl < catmoves.txt
-# If you want to use an edit summary, then pass it in as a parameter, i.e.,
-# perl catmove.pl "Emptying dead category" < catmoves.txt
-# Note that if your summary has multiple words in it then enclose it in quotes.
-#
-# To set edit summaries, you can also preface a line with a semicolon ;
-# That will be the edit summary for all subsequent executions ... unless you modify the
-# edit summary again with another semicolon. In this way you can take care of many days
-# worth of WP:CFD backlog with a single execution.
-
-#Necessary for thread handling.
-use POSIX ":sys_wait_h";
-
-#Preliminary setup.
-my $editSummary = '';
-my $customSummary = 0;
-my $throttle = 6;
-my $sysCall = '';
-my $errorLog = "";
-if ($#ARGV >= 0) {
- $customSummary = 1;
- $editSummary = shift;
-}
-
-#my $args = "-batch -putthrottle:$throttle -inplace";
-my $args = "-batch -putthrottle:$throttle";
-
-#Loop through all standard input.
-while (<STDIN>) {
- #Matches category moves and category removals.
- if ($_ =~ m/^\s*[\#\*]?\s*[Cc]ategory:(.*?)\s*to\s*[Cc]ategory:(.*?)\s*$/ || $_ =~
m/^\s*[\#\*]?\s*[Cc]ategory:(.*?)\s*$/) {
- #Matches category moves.
- if ($_ =~ m/^\s*[\#\*]?\s*[Cc]ategory:(.*?)\s*to\s*[Cc]ategory:(.*?)\s*$/) {
- my $from = $1;
- my $to = $2;
- if ($customSummary == 0) {
- $sysCall = "python category.py move -from:\"$from\"
-to:\"$to\"";
- } else {
- $sysCall = "python category.py move -from:\"$from\"
-to:\"$to\" -summary:\"$editSummary\"";
- }
- }
- #Matches category removals.
- elsif ($_ =~ m/^\s*[\#\*]?\s*[Cc]ategory:(.*?)\s*$/) {
- my $from = $1;
- if ($customSummary == 0) {
- $sysCall = "python category.py remove -from:\"$from\"";
- } else {
- $sysCall = "python category.py remove -from:\"$from\"
-summary:\"$editSummary\"";
- }
- }
-
- #Fork off the execution of the Python bot as its own thread. Errors will be skipped and
execution continues
- #through the whole list. Ctrl-C will quit everything immediately, though. (This is why
we can't use the
- #much simpler system() function).
- print "Executing: $sysCall $args\n";
- defined (my $pid = fork) or die "Cannot fork: $!";
- unless ($pid) {
- exec("$sysCall $args");
- }
- waitpid($pid, 0);
- #If the Python script terminates abnormally print something to that effect.
- if ($? == 256) {
- $errorLog .= "* $sysCall $args\n";
- }
- }
- #Setting the edit summary.
- elsif ($_ =~ m/^\s*;\s*(.*)\s*$/) {
- $customSummary = 1;
- $editSummary = $1;
- } else {
- print "Invalid line: $_\n";
- }
-}
-
-if ($errorLog ne "") {
- print "Errors detected with the following commands:\n$errorLog";
-}
Added: trunk/pywikipedia/cfd.py
===================================================================
--- trunk/pywikipedia/cfd.py (rev 0)
+++ trunk/pywikipedia/cfd.py 2008-09-29 02:58:20 UTC (rev 5923)
@@ -0,0 +1,138 @@
+# -*- coding: utf-8 -*-
+"""
+This script processes the Categories for discussion working page. It parses
+out the actions that need to be taken as a result of CFD discussions (as posted
+to the working page by an administrator) and performs them.
+
+Syntax: python cfd.py
+
+"""
+
+# (C) Ben McIlwain, 2008
+#
+# Distributed under the terms of the MIT license.
+
+import wikipedia
+import re
+import category
+
+# Regular expression declarations
+# See the en-wiki CFD working page at [[Wikipedia:Categories for discussion/Working]]
+# to see how these work in context. To get this bot working on other wikis you will
+# need to adjust these regular expressions at the very least.
+nobots = re.compile(r"NO\s*BOTS", re.IGNORECASE);
+example = re.compile(r"\[\[\:Category\:(.)\1\1\1\1\]\]", re.IGNORECASE);
+speedymode = re.compile(r"^===*\s*Speedy Moves\s*===*\s*$", re.IGNORECASE);
+movemode = re.compile(r"^===*\s*Move\/Merge then delete\s*===*\s*$",
re.IGNORECASE);
+emptymode = re.compile(r"^===*\s*Empty then delete\s*===*\s*$", re.IGNORECASE)
+deletemode = re.compile(r"^===*\s*Ready for deletion\s*===*\s*$",
re.IGNORECASE)
+maintenance = re.compile(r"^===*\s*Old by month categories with
entries\s*===*\s*$", re.IGNORECASE)
+dateheader = re.compile(r"(\[\[Wikipedia\:Categories[_ ]for[_
](?:discussion|deletion)\/Log\/([^\]]*?)\]\])", re.IGNORECASE)
+movecat =
re.compile(r"\[\[\:Category\:([^\]]*?)\]\][^\]]*?\[\[\:Category\:([^\]]*?)\]\]",
re.IGNORECASE)
+deletecat = re.compile(r"\[\[\:Category\:([^\]]*?)\]\]", re.IGNORECASE)
+findday = re.compile(r"\[\[(Wikipedia\:Categories for
(?:discussion|deletion)\/Log\/\d{4} \w+ \d+)#", re.IGNORECASE)
+
+class ReCheck:
+ def __init__(self):
+ self.result = None
+ def check(self, pattern, text):
+ self.result = pattern.search(text)
+ return self.result
+
+def main():
+ wikipedia.handleArgs();
+
+ page = wikipedia.Page(wikipedia.getSite(), 'Wikipedia:Categories for
discussion/Working')
+
+ # Variable declarations
+ day = "None"
+ mode = "None"
+ src = "None"
+ dest = "None"
+ line = ""
+ summary = ""
+ robot = None
+
+ m = ReCheck()
+ for line in page.get().split("\n"):
+ if (nobots.search(line)):
+ # NO BOTS!!!
+ pass
+ elif (example.search(line)):
+ # Example line
+ pass
+ elif (speedymode.search(line)):
+ mode = "Speedy"
+ day = "None"
+ elif (movemode.search(line)):
+ mode = "Move"
+ day = "None"
+ elif (emptymode.search(line)):
+ mode = "Empty"
+ day = "None"
+ elif (deletemode.search(line)):
+ mode = "Delete"
+ day = "None"
+ elif (maintenance.search(line)):
+ # It's probably best not to try to handle these in an automated fashion.
+ mode = "None"
+ day = "None"
+ elif (m.check(dateheader, line)):
+ day = m.result.group(1)
+ elif (m.check(movecat, line)):
+ src = m.result.group(1)
+ dest = m.result.group(2)
+ if (mode == "Move" and day != "None"):
+ summary = "Robot - Moving category " + src + " to " +
dest + " per [[WP:CFD|CFD]] at " + findDay(src, day) + "."
+ elif (mode == "Speedy"):
+ summary = "Robot - Speedily moving category " + src + " to
" + dest + " per [[WP:CFD|CFD]]."
+ else:
+ continue
+ robot = category.CategoryMoveRobot(oldCatTitle=src, newCatTitle=dest,
batchMode=True,
+ editSummary=summary, inPlace=True,
moveCatPage=True,
+ deleteEmptySourceCat=True)
+ elif (m.check(deletecat, line)):
+ src = m.result.group(1)
+ # I currently don't see any reason to handle these two cases separately,
though
+ # if are guaranteed that the category in the "Delete" case is
empty, it might be
+ # easier to call delete.py on it.
+ if ((mode == "Empty" or mode == "Delete") and day !=
"None"):
+ summary = "Robot - Removing category " + src + " per
[[WP:CFD|CFD]] at " + findDay(src, day) + "."
+ else:
+ continue
+ robot = category.CategoryRemoveRobot(catTitle=src, batchMode=True,
editSummary=summary,
+ useSummaryForDeletion=True,
inPlace=True)
+ else:
+ # This line does not fit any of our regular expressions, so ignore it.
+ pass
+ if (summary != "" and robot != None):
+ wikipedia.output(summary, toStdout=True)
+ # Run, robot, run!
+ robot.run()
+ summary = ""
+ robot = None
+
+# This function grabs the wiki source of a category page and attempts to
+# extract a link to the CFD per-day discussion page from the CFD template.
+# If the CFD template is not there, it will return the value of the second
+# parameter, which is essentially a fallback that is extracted from the
+# per-day subheadings on the working page.
+def findDay(pageTitle, oldDay):
+ page = wikipedia.Page(wikipedia.getSite(), "Category:" + pageTitle)
+ pageSrc = page.get()
+ m = findday.search(pageSrc)
+ if (m != None):
+ return "[[" + m.group(1) + "]]"
+ else:
+ wikipedia.output("Could not find CFD day link on Category:" + pageTitle
+ "\n", toStdout=True)
+ return oldDay
+
+if __name__ == "__main__":
+ try:
+ main()
+ except:
+ wikipedia.stopme()
+ raise
+ else:
+ wikipedia.stopme()
+