<syntaxhighlight lang="python">
import sys
import catlib
import wikipedia
import codecs
the maximum number of articles per page
MAX = 6000
should we write to file or directly to wikipedia?
DBG = False
class Watchlist:
# the name of the template used to tag articles, e.g., "Numismaticnotice"
template = ""
# the name of the project, e.g., "Numismatics"
project = ""
# the location of the article list (output) -- without prefix, so for
# "Wikipedia:WikiProject Numismatics/Articles", use "Articles"
articleOut = ""
# a list for all articles
articles = []
# a list for all article talk pages
articlesTalk = []
# a list for all Wikipedia pages
wikis = []
# a list for all Wikipedia talk pages
wikisTalk = []
# a list for all templates
templates = []
# a list for all template talk pages
templatesTalk = []
# a list for all categories
categories = []
# a list for all category talk pages
categoriesTalk = []
# a list for all images
images = []
# a list for all image talk pages
imagesTalk = []
# a list for all portals
portals = []
# a list for all portal talk pages
portalsTalk = []
# certain pages need to be included explicitly (for example, if they share
# a talk page)
includePages = []
def getTaggedPages (self):
"""
Get the pages that include templateName
Add the articles to the appropriate lists
"""
page = wikipedia.Page(wikipedia.getSite(), "Template:" + self.template)
refs = page.getReferences(onlyTemplateInclusion=True)
for page in refs:
self.processPageName(page.title())
# include the explicitly named pages
for page in self.includePages:
self.processPageName(page)
# remove duplicates and sort the lists
self.removeDuplicatesAndSort()
# organize the categories hierarchically (actually, no -- this takes too
# much time)
#self.catText = organizeCategories()
# include the explicitly named pages
articles = []
for page in refs:
result = page.title().split(":")
if result[0] == "Category talk": # we expect this
findArticlesInCategory("Category:" + result[1], articles)
# add the category to the list as well
articles.append(page.title())
articles = sorted(set(articles))
for page in articles:
self.processPageName(page)
# remove duplicates and sort the lists
self.removeDuplicatesAndSort()
# organize the categories hierarchically (actually, no -- this takes too
# much time)
#self.catText = organizeCategories()
def writeList (self, taggedPagesFlag):
"""
write the output to the specified page on Wikipedia
taggedPagesFlag tells whether we're looking for tagged pages (true)
or tagged categories (false)
"""
tagText = ""
if not taggedPagesFlag:
tagText = "in categories "
# the output page, without spaces
wikipedia.output(u"Preparing output")
"recreated from time to time. Accordingly, any changes you " + \
"make here will be overwitten. See below for details.
\n\n"
# double the number of articles because of talk pages
splitting = totalArticles*2 > MAX
if splitting:
mainText += "There are too many articles in this project to list " + \
"them all on one page. This article contains the first " + \
str(MAX) + " articles and links to other articles which " + \
"contain "
else:
mainText += "This article contains "
mainText += "links to all articles, categories, images, portal pages " + \
"templates, and project pages " + tagText + "with {{[[Template:" + \
self.template + "|" + \
self.template + "]]}} on their talk page. It was " + \
"generated by " + \
"Ganeshbot. Its purpose is to be able to track " + \
"the project history using [[Special:Recentchangeslinked/" + \
"Wikipedia:WikiProject " + output + \
"|related changes]] or [http://tools.wikimedia.de/~interiot/" + \
"cgi-bin/offtoolserver/RC_firstonly?url=http%3A%2F%2Fen.wikipedia.org" + \
"%2Fw%2Findex.php%3Ftitle%3DSpecial%3ARecentchangeslinked%26target" + \
"%3DWikipedia:WikiProject_" + output + \
"%26hideminor%3D0%26days%3D7%26limit%3D500 related watchlist] which " + \
"only shows the last change for each article.\n" + \
"\n"
mainText += "==Regular content (count: " + str(totalArticles) + ")==\n"
# the number of articles listed on this page
count = 0
# the page number
pageNo = 1
# the text for this subpage (if no subpages, will just be on the main
# page)
mainText += "===Articles (count: " + str(len(self.articles)) + ")===\n"
prevChar = firstChar = "Z" #initialize to anything but A
subText = ""
# make sure the first batch of articles goes to the main page
firstBatch = True
for s in self.articles:
if s[0] != prevChar:
subText += "\n\n"
subText += "====" + s[0] + "====\n"
prevChar = s[0]
if count == 0:
firstChar = prevChar
subText += "" + s + " - "
count += 1
if count > MAX:
count = 0
if firstBatch:
firstBatch = False
mainText += subText
else:
mainText += "\n"
mainText += "====" + \
firstChar + " through " + prevChar + "====\n"
subText = subText.replace("<range>", firstChar + " through " + \
prevChar)
self.writeProjPage(self.articleOut + "/Page" + str(pageNo),
subText)
pageNo += 1
firstChar = prevChar
subText = "===Articles <range>===\n" + \
"====" + prevChar + "====\n"
if splitting and not firstBatch:
mainText += "====" + \
firstChar + " through " + prevChar + "====\n"
subText = subText.replace("<range>", firstChar + " through " + prevChar)
self.writeProjPage(self.articleOut + "/Page" + str(pageNo), subText)
pageNo += 1
else:
mainText += subText
mainText += "\n\n"
mainText += "==Talk pages==\n"
prevChar = firstChar = "Z" #initialize to anything but A
if splitting:
subText = "This article contains links to some talk pages " + tagText + \
"with {{" + self.template + "}} " + \
"on their talk page. It was generated by " + \
"Ganeshbot. Its purpose is to be able to track " + \
"the project history using [[Special:Recentchangeslinked/" + \
"Wikipedia:WikiProject " + output + \
"/Page" + str(pageNo) + "|related changes]] or [http://tools.wikimedia.de/~interiot/" + \
"cgi-bin/offtoolserver/RC_firstonly?url=http%3A%2F%2Fen.wikipedia.org" + \
"%2Fw%2Findex.php%3Ftitle%3DSpecial%3ARecentchangeslinked%26target" + \
"%3DWikipedia:WikiProject_" + output + \
"/Page" + str(pageNo) + "%26hideminor%3D0%26days%3D7%26limit%3D500 related watchlist] which " + \
"only shows the last change for each article.\n" + \
"\n" + \
"===Articles <range>===\n"
else:
subText = ""
project = "Numismatics", pageName="Categories"):
"""
organize the categories hierarchically
write the results to "Wikipedia:WikiProject <project>/<page>"
"""
# get the list of tagged categories
dummy = []
taggedCatList = []
getTagged(tag, taggedCatList, dummy)
print len(taggedCatList)
text = "This is the category structure for [[Wikipedia:WikiProject " + \
project + "|" + project + "]] \n"
cat = catlib.Category(wikipedia.getSite(), "Category:" + topLevelCat)
text += "Category:"+topLevelCat+" \n"
text = organizeCatsNextLevel(text, cat, "|—", taggedCatList)
"""
recursively organize the category text
text is the text so far, add to that
cat is the catlib.Category of the previous level
substring is the text to put before each category
taggedCatList is the list of tagged categories
returns the text so far
"""
subcats = cat.subcategories()
for subcat in subcats:
# if this subcategory is included in our project
if subcat.title() in taggedCatList:
# if it has not already been listed (to prevent duplication,
# but more importantly, to prevent infinite loops)
if subcat.title() not in text:
text += substring + "" + subcat.title() + " \n"
text = organizeCatsNextLevel(text, subcat,
"| "+substring,
taggedCatList)
else: # it's already been listed
text += substring + "" + subcat.title() + " (already included, see above) \n"
# don't recurse in this case, to prevent infinite loops
return text
def getExcluded (project):
"""
get the list of pages which should not be tagged even though they're in
tagged categories
this can also be used to get excluded categories, if they're listed on
the project exclusion page
"""
page = wikipedia.Page(wikipedia.getSite(), "User:WatchlistBot/" + project)
if page.exists():
text = page.get()
# find the "----" the list of articles is below the line
start = text.find("----\n")
result = text[start+4:].split("[[")
pages = []
for page in result:
end = page.find("]]")
if end != -1:
pages.append(getTalkVersion(page[:end]))
return pages
return []
def getTalkVersion (name):
"""
given a page name, convert it to the associated talk page
"""
result = name.split(":")
if len(result) == 1: #article
return "Talk:"+name
if len(result) == 3: #category
return "Category talk:"+result[2]
if "Talk" in result[0]:
return name
return result[0] + " talk:" + result[1]
"""
remove the tag from the given talk page, if it is there
"""
page = wikipedia.Page(wikipedia.getSite(), pageName)
if page.exists():
if not page.isRedirectPage():
text = page.get()
tagStart = text.find("{{"+tag)
if tagStart == -1:
wikipedia.output("Page " + page.title() + " not tagged")
else:
# find the end of the tag (add 3 for the }}\n)
tagEnd = text[tagStart:].find("}}") + tagStart + 3
text = text[:tagStart] + text[tagEnd:]
return writePage(page, text, "Removing " + tag)
return True
def tagPage (pageName, tag, params = ""):
"""
tag the given talk page with the tag
params is an optional list of parameters for the tag (like class=Stub)
"""
# get the talk page
page = wikipedia.Page(wikipedia.getSite(), pageName)
if page.exists():
if not page.isRedirectPage():
text = page.get()
return tagIt(page, text, tag+params)
else:
wikipedia.output("Page " + page.title() + " is a redirect")
else:
# we don't mind if the page doesn't exist yet, just create it
return tagIt(page, "", tag+params)
return True
def tagIt (page, text, tag):
text = "Template:" + tag + "\n\n" + text
return writePage(page, text, "Adding " + tag)
includeCats = False):
"""
find all the articles in the given category, and return a list
If confirm is true, check each article with the user
articles is the list so far
includeCats indicates whether category talk pages should be included
"""
# get the category (don't include it, since tagging articles and categories
# is handled separately)
cat = catlib.Category(wikipedia.getSite(), catName)
# get all pages in this category
pages = cat.articles()
for page in pages:
# if confirming, check
if confirm:
response = wikipedia.input(u"Do you want to tag " + page.title() + u"? (y for yes)")
if not confirm or response == "y":
# add the appropriate prefix
if page.namespace() == 10: # template
articles.append("Template talk:" + page.titleWithoutNamespace())
elif page.namespace() == 0: # article
articles.append("Talk:" + page.title())
elif page.namespace() == 6: # image
articles.append("Image talk:" + page.titleWithoutNamespace())
elif page.namespace() == 100: # portal
articles.append("Portal talk:" + page.titleWithoutNamespace())
elif page.namespace() == 4: # wikipedia
articles.append("Wikipedia talk:" + page.titleWithoutNamespace())
elif page.namespace() == 1 or \ # article talk
page.namespace() == 5 or \ # wikipedia talk
page.namespace() == 7 or \ # image talk
page.namespace() == 11 or \ # template talk
page.namespace() == 101: # portal talk
articles.append(page.title())
elif page.namespace() == 15: # category talk
if includeCats:
articles.append(page.title())
elif page.namespace() == 2 or \ # user
page.namespace() == 3: # user talk
# ignore these (dummy command)
x = 1
else:
print "Unexpected namespace on " + page.title() + ": " + str(page.namespace())
#remove duplicates
articles = list(set(articles))
keywords, excluded = [],
questionText = u"Do you want to tag ", confirm = True):
"""
if catList starts with "", it means we're trying to quit, so just return
starting at catName, make a list, catList, of all subcategories
ask the user first, and allow the user the choice to recurse
through subcategories
taggedCats is the list of categories that are already tagged and can thus
be skipped
otherTaggedCats is the list (possibly empty) of categories that are
tagged with a related tag -- these should be skipped, with no recursion
keywords are words that if they're in the category, it will be tagged
without confirmation
excluded are categories to skip (treat as if user said 'n')
if confirm is false, no confirmation question will be asked (all will be
included)
"""
# check if we're quitting
if len(catList) > 1 and catList[0] == "":
return catList
cat = catlib.Category(wikipedia.getSite(), "Category:" + catName)
response = "z"
# if we have not already decided to tag this cat
if catName not in catList:
# if the categories is already in the taggedCats, treat that like a
# "y" from the user
if "Category:"+catName in taggedCats:
response = "y"
# if the category is in otherTaggedCats, treat it like a "n"
if "Category:"+catName in otherTaggedCats:
response = "n"
elif "Category talk:"+catName in excluded:
response = "n"
else:
# if the name has a keyword in it, treat that like a "y" from the user
for keyword in keywords:
if keyword in catName:
response = "y"
# if confirm is False, treat it as if the user already said yes
if confirm == False:
response = "y"
# if response is still "z", ask the user
if response == "z":
response = wikipedia.input(questionText + cat.title() + u"? (y for yes, yn for yes but no recursion, s for stop recursion)")
if response == "s":
# put "" into the catlist at the beginning as a marker
catList.insert(0, "")
return catList
# add the category to the list
if response == "y" or response == "yn":
catList.append(cat.titleWithoutNamespace())
# recurse through subcategories
if response == "y":
subcats = cat.subcategories()
for subcat in subcats:
updateCategoryList(catList, subcat.titleWithoutNamespace(),
taggedCats, otherTaggedCats, keywords,
excluded, questionText, confirm)
return catList
def tagCategories (catName = "Electronics", tag = "Electron",
otherTag = "", project = "Electronics",
params = "|class=cat", keywords = []):
"""
tag all categories in the specified category and subcategories with the
specified tag (at the top of the page)
if otherTag is not "", skip categories which are tagged with othertag
check with the user for each category
keywords are words that if they're in the category, it will be tagged
without confirmation
"""
wikipedia.put_throttle.setDelay(10, absolute = True)
# get the list of categories which are already tagged
taggedCatList = []
taggedArticleList = []
getTagged(tag, taggedCatList, taggedArticleList)
otherTaggedCatList = []
if otherTag != "":
getTagged(otherTag, otherTaggedCatList, taggedArticleList)
# get the list of categories and articles that are to be excluded (articles
# will be ignored)
excluded = getExcluded(project)
# get the category list
catList = []
catList = updateCategoryList(catList, catName, taggedCatList, otherTaggedCatList,
keywords, excluded)
# if the first element of catList is "", remove it, it was just a marker
if catList[0] == "":
catList.remove("")
# remove duplicates and sort
catList = sorted(set(catList))
# remove categories which are already tagged
for cat in catList:
if "Category:"+cat not in taggedCatList:
tagPage("Category talk:" + cat, tag, params)
def untagCategories (catList = [],
tag = "Electron", project = "Electronics"):
"""
untag all specified categories
"""
wikipedia.put_throttle.setDelay(10, absolute = True)
for cat in catList:
untagPage("Category talk:" + cat, tag)
def getTagged (tag, catList, articles):
"""
get a list of categories and articles which contain the specified tag
"""
page = wikipedia.Page(wikipedia.getSite(), "Template:" + tag)
refs = page.getReferences(onlyTemplateInclusion=True)
for page in refs:
name = page.title()
result = name.split(":")
if result[0] == "Category talk":
catList.append("Category:"+result[1])
else:
articles.append(name)
def untag (catList = [],
tag = "Numismaticnotice",
returnList = False):
"""
remove the tag from all articles in the specified categories
this is useful when the bot makes a mistake
if returnList is true, just return a list, don't actually untag
"""
articles = []
for catName in catList:
findArticlesInCategory("Category:"+catName, articles, False)
articles = sorted(set(articles))
if returnList:
return articles
else:
for article in articles:
untagPage(article, tag)
wikipedia.stopme()
comment="Numismatics assessment, class="):
"""
go through all articles in the specified category and classify them as
image, template, category, portal, or NA. Articles are left as is (as are
lists and disambig pages)
"""
articles = []
findArticlesInCategory("Category:"+catName, articles, False, True)
for article in articles:
# if this is a template
if "Template talk:" in article:
templatesToTag.append(article)
# if this is a category page
if "Category talk:" in article:
categoriesToTag.append(article)
# if this is an image
if "Image talk:" in article:
imagesToTag.append(article)
# if this is a portal
if "Portal talk:" in article:
portalsToTag.append(article)
# if this is a regular talk page, assume it's disambig
secondCat = "Louisville stubs",
recurse = True,
paramName = "class",
paramValue = "Stub",
tag = "WikiProject Louisville",
comment = "Louisville assessment, adding class=Stub"):
"""
find the articles in the intersection of firstCat and secondCat
if recurse is true, include all subcats of secondCat (but not firstCat)
paramName is the parameter to add (e.g., "class")
paramValue is the value to assign (e.g., "NA")
tag is the name of the template tag
comment is the text to use for the comment when saving
"""
# get the list of articles in the first category
firstArticles = []
findArticlesInCategory("Category:"+firstCat, firstArticles, False)
# get the list of articles in the second category
secondCatList = []
secondCatList = updateCategoryList(secondCatList, secondCat, [], [],
"Do you want to include ", False)
secondArticles = []
for cat in secondCatList:
findArticlesInCategory("Category:"+cat, secondArticles, False)
# get the list of articles that is in both
articles = []
for article in firstArticles:
if article in secondArticles:
articles.append(article)
"""
articles is the list of articles to change
paramName is the parameter to add (e.g., "class")
paramValue is the value to assign (e.g., "NA")
tag is the name of the template tag
comment is the text to use for the comment when saving
"""
for article in articles:
page = wikipedia.Page(wikipedia.getSite(), article)
text = page.get()
# skip the first character so we don't have to worry about upper/lower
tagStart = text.find(tag[1:])
tagEnd = text[tagStart:].find("}}")
tagEnd = tagStart + tagEnd
paramStart = text[tagStart:tagEnd].find(paramName)
if paramStart != -1:
paramStart = tagStart + paramStart - 1
paramEnd = text[paramStart+1:tagEnd].find("|")
if paramEnd != -1:
paramEnd = paramStart + paramEnd + 1
else:
paramEnd = tagEnd
else:
paramStart = tagEnd
paramEnd = tagEnd
text = text[:paramStart] + "|" + paramName + "=" + paramValue + \
text[paramEnd:]
if not writePage(page, text, comment):
break
"""
replace the oldTag with the newTag (can be used to replace a tag with
a tag plus parameters)
"""
articles = []
getTagged(oldTag, [], articles)
for article in articles:
page = wikipedia.Page(wikipedia.getSite(), article)
text = page.get()
text = wikipedia.replaceExceptMathNowikiAndComments(
text, oldTag, newTag)
if not writePage(page, text, "replacing " + oldTag + " with " + newTag):
break
project = "Numismatics", confirm=False, catList = [],
returnList = False, assessmentTag = "numismatic articles"):
"""
tag articles in tagged categories
if a page is already tagged with otherTag, skip it (use otherTag = "" for none)
catList is a list of categories to check in. If empty, use tagged categories
if params is given, include it after the tag, when tagging an article
if returnList is true, don't actually tag anything, just return the list
in this case, also don't skip a page just because it's already tagged
assessmentTag is a text string contained in the assessment categories, use
"" to ignore
"""
# get the list of all tagged articles in taggedArticles
# if catList was given, leave it as is. Otherwise, populate catList with
# all tagged categories
taggedArticles = []
if len(catList) == 0:
getTagged(tag, catList, taggedArticles)
# skip the assessment categories (otherwise, we won't skip articles
# which are currently tagged but shouldn't be)
newCatList = []
for cat in catList:
if assessmentTag != "" and \
assessmentTag not in cat:
newCatList.append(cat)
catList = newCatList
else:
dummy = []
getTagged(tag, dummy, taggedArticles)
# put "Category:" in front of the category names
newCatList = []
for cat in catList:
newCatList.append("Category:"+cat)
catList = newCatList
# add the articles tagged with otherTag to the list of taggedArticles
if otherTag != "":
getTagged(otherTag, [], taggedArticles)
# get the list of untagged articles in the categories in catList (which
# was either supplied as a parameter, or was populated with tagged categories)
untaggedArticles = []
for cat in catList:
findArticlesInCategory(cat, untaggedArticles, confirm)
# remove duplicates and sort
untaggedArticles = dict.fromkeys(untaggedArticles).keys()
untaggedArticles.sort()
# if we're returning a list, stop here
if returnList:
return untaggedArticles
# make a list of articles that need to be tagged (by removing articles
# that are already tagged from list of all articles)
for article in taggedArticles:
if article in untaggedArticles:
untaggedArticles.remove(article)
# remove excluded articles
excluded = getExcluded(project)
for page in excluded:
if page in untaggedArticles:
untaggedArticles.remove(page)
if len(untaggedArticles) == 0:
wikipedia.output(u"No untagged articles")
print "Tagging " + str(len(untaggedArticles)) + " articles"
# tag the articles
for article in untaggedArticles:
tagPage(article, tag, params)
"Digital currency exchangers",
"Digital gold currencies",
"Money", "Money stubs",
"Foreign exchange market", "Ancient mints",
"Challenge coin"]):
"""
untag the articles in the specified categories, but only if they are
not in other categories that require them to be tagged
"""
# find articles that should be tagged
needTagList = tag("Numismaticnotice", "", "Exonumianotice", "Numismatics",
False, [], True)
# now get the list of articles to untag (returns all articles in the
# specified categories, without checking if they're tagged)
untagList = untag(catList, "Numismaticnotice", True)
# if an article is in the untagList and not in the needTagList, untag it
for article in untagList:
if article not in needTagList:
untagPage(article, "Numismaticnotice")
def findDoubleTags (catList = []):
"""
find articles that are in numismatics as well as exonumia categories
"""
# find articles that think they should be tagged Exonumia and Numismaticnotice
numArticles = tag("Numismaticnotice", "", "", "Numismatics", False, [], True)
getTagged("Numismaticnotice", [], numArticles)
exoArticles = tag("Exonumianotice", "", "", "Numismatics", False, [], True)
getTagged("Exonumianotice", [], exoArticles)
bothArticles = []
for article in numArticles:
if article in exoArticles:
bothArticles.append(article)
text = ""
for article in bothArticles:
text += "*"+article+" \n"
print text
wikipedia.stopme()
def listProjects ():
"""
print out a list of active projects, with numbers to use for an individual update
"""
for proj,x in enumerate(projects):
print str(proj) + ": " + x
def main():
"""
update the project watchlists. If projectNum is given, only update the
given project number (see projects for list, remember to start at 0)
"""
