# -*- coding: utf-8 -*- """ Scripts to manage categories. Syntax: python category.py action [-option] where action can be one of these: * add - mass-add a category to a list of pages * remove - remove category tag from all pages in a category * move - move all pages in a category to another category * tidy - tidy up a category by moving its articles into subcategories * tree - show a tree of subcategories of a given category and option can be one of these: * person - sort persons by their last name (for action 'add') * rebuild - reset the database For the actions tidy and tree, the bot will store the category structure locally in category.dump. This saves time and server load, but if it uses these data later, they may be outdated; use the -rebuild parameter in this case. For example, to create a new category from a list of persons, type: python category.py add -person and follow the on-screen instructions. """ # # (C) Rob W.W. Hooft, 2004 # (C) Daniel Herding, 2004 # __version__ = '$Id: category.py,v 1.103 2006/03/12 17:34:04 wikipedian Exp $' # # Distributed under the terms of the MIT license. # import re, sys, string, pickle, bz2 import wikipedia, catlib, config, pagegenerators # Summary messages msg_add={ 'da':u'Robot: Tilføjer [[%s]]', 'de':u'Bot: Ergänze [[%s]]', 'en':u'Robot: Adding [[%s]]', 'es':u'Bot: Añadida [[%s]]', 'fi':u'Botti lisäsi luokkaan [[%s]]', 'fr':u'Robot : ajoute [[%s]]', 'ia':u'Robot: Addition de [[%s]]', 'is':u'Vélmenni: Bæti við [[%s]]', 'no':u'Robot: Legger til [[%s]]', 'pt':u'Bot: Adicionando [[%s]]', 'sr':u'Бот: Додаје [[Категорија:%s]]', } msg_change={ 'da':u'Robot: Ændrer %s', 'de':u'Bot: Ändere %s', 'en':u'Robot: Changing %s', 'es':u'Bot: Cambiada %s', 'fi':u'Botti vaihtoi luokan %s', 'fr':u'Robot : modifie %s', 'ia':u'Robot: Modification de %s', 'is':u'Vélmenni: Breyti flokknum [[%s]]', 'nl':u'Bot: Wijziging %s', 'no':u'Robot: Endrer %s', 'pt':u'Bot: Modificando [[%s]]', 'sr':u'Бот: Измена категорије %s', } deletion_reason_move = { 'de':u'Bot: Kategorie wurde nach %s verschoben', 'en':u'Robot: Category was moved to %s', 'fr':u'Robot : catégorie déplacée sur %s', 'ia':u'Robot: Categoria transferite a %s', 'no':u'Robot: Kategorien ble flyttet til %s', 'pt':u'Bot: Categoria %s foi movida', 'sr':u'Бот: Категорија премештена у %s', } class CategoryDatabase: ''' This is a temporary knowledge base saving for each category the contained subcategories and articles, so that category pages don't need to be loaded over and over again ''' def __init__(self, rebuild = False, filename = 'category.dump.bz2'): if rebuild: self.rebuild() else: try: f = bz2.BZ2File(filename, 'r') wikipedia.output(u'Reading dump from %s' % filename) databases = pickle.load(f) f.close() # keys are categories, values are 2-tuples with lists as entries. self.catContentDB = databases['catContentDB'] # like the above, but for supercategories self.superclassDB = databases['superclassDB'] del databases except: # If something goes wrong, just rebuild the database self.rebuild() def rebuild(self): self.catContentDB={} self.superclassDB={} def getSubcats(self, supercat): ''' For a given supercategory, return a list of Categorys for all its subcategories. Saves this list in a temporary database so that it won't be loaded from the server next time it's required. ''' # if we already know which subcategories exist here if self.catContentDB.has_key(supercat): return self.catContentDB[supercat][0] else: subcatlist = supercat.subcategories() articlelist = supercat.articles() # add to dictionary self.catContentDB[supercat] = (subcatlist, articlelist) return subcatlist def getArticles(self, cat): ''' For a given category, return a list of Pages for all its articles. Saves this list in a temporary database so that it won't be loaded from the server next time it's required. ''' # if we already know which articles exist here if self.catContentDB.has_key(cat): return self.catContentDB[cat][1] else: subcatlist = cat.subcategories() articlelist = cat.articles() # add to dictionary self.catContentDB[cat] = (subcatlist, articlelist) return articlelist def getSupercats(self, subcat): # if we already know which subcategories exist here if self.superclassDB.has_key(subcat): return self.superclassDB[subcat] else: supercatlist = subcat.supercategories() # add to dictionary self.superclassDB[subcat] = supercatlist return supercatlist def dump(self, filename = 'category.dump.bz2'): ''' Saves the contents of the dictionaries superclassDB and catContentDB to disk. ''' wikipedia.output(u'Dumping to %s, please wait...' % filename) f = bz2.BZ2File(filename, 'w') databases = { 'catContentDB': self.catContentDB, 'superclassDB': self.superclassDB } # store dump to disk in binary format pickle.dump(databases, f, protocol=pickle.HIGHEST_PROTOCOL) f.close() def sorted_by_last_name(catlink, pagelink): ''' given a Category, returns a Category which has an explicit sort key which sorts persons by their last names. Trailing words in brackets will be removed. Example: If category_name is 'Author' and pl is a Page to [[Alexandre Dumas (senior)]], this function will return this Category: [[Category:Author|Dumas, Alexandre]] ''' page_name = pagelink.title() site = pagelink.site() # regular expression that matches a name followed by a space and # disambiguation brackets. Group 1 is the name without the rest. bracketsR = re.compile('(.*) \(.+?\)') match_object = bracketsR.match(page_name) if match_object: page_name = match_object.group(1) split_string = page_name.split(' ') if len(split_string) > 1: # pull last part of the name to the beginning, and append the rest after a comma # e.g. "John von Neumann" becomes "Neumann, John von" sorted_key = split_string[-1] + ', ' + string.join(split_string[:-1], ' ') # give explicit sort key return wikipedia.Page(site, catlink.title() + '|' + sorted_key) else: return wikipedia.Page(site, catlink.title()) def add_category(sort_by_last_name = False): ''' A robot to mass-add a category to a list of pages. ''' print "This bot has two modes: you can add a category link to all" print "pages mentioned in a List that is now in another wikipedia page" print "or you can add a category link to all pages that link to a" print "specific page. If you want the second, please give an empty" print "answer to the first question." listpageTitle = wikipedia.input(u'Wiki page with list of pages to change:') site = wikipedia.getSite() pages = [] if listpageTitle: try: listpage = wikipedia.Page(site, listpageTitle) pages = listpage.linkedPages() except wikipedia.NoPage: wikipedia.output(u'%s could not be loaded from the server.' % listpage.aslink()) except wikipedia.IsRedirectPage: wikipedia.output(u'%s is a redirect to %s.' % (listpage.aslink(), listpage.getRedirectTarget())) else: referredPage = wikipedia.input(u'Wikipedia page that is now linked to:') page = wikipedia.Page(wikipedia.getSite(), referredPage) pages = page.getReferences() wikipedia.output(u' ==> %i pages to process\n' % len(pages)) if len(pages) > 0: newcatTitle = wikipedia.input(u'Category to add (do not give namespace):') newcatTitle = newcatTitle[:1].capitalize() + newcatTitle[1:] # set edit summary message wikipedia.setAction(wikipedia.translate(wikipedia.getSite(), msg_add) % newcatTitle) cat_namespace = wikipedia.getSite().category_namespaces()[0] answer = '' for page in pages: if answer != 'a': answer = '' while answer not in ('y','n','a'): answer = wikipedia.input(u'%s [y/n/a(ll)]:' % (page.aslink())) if answer == 'a': confirm = '' while confirm not in ('y','n'): confirm = wikipedia.input(u'This should be used if and only if you are sure that your links are correct! Are you sure? [y/n]:') if confirm == 'n': answer = '' if answer == 'y' or answer == 'a': try: cats = page.categories() except wikipedia.NoPage: wikipedia.output(u"%s doesn't exist yet. Ignoring." % (page.title())) pass except wikipedia.IsRedirectPage,arg: redirTarget = wikipedia.Page(site,arg.args[0]) wikipedia.output(u"WARNING: %s is redirect to %s. Ignoring." % (page.title(), redirTarget.title())) else: wikipedia.output(u"Current categories:") for cat in cats: wikipedia.output(u"* %s" % cat.title()) catpl = wikipedia.Page(site, cat_namespace + ':' + newcatTitle) if sort_by_last_name: catpl = sorted_by_last_name(catpl, page) if catpl in cats: wikipedia.output(u"%s is already in %s." % (page.title(), catpl.title())) else: wikipedia.output(u'Adding %s' % catpl.aslink()) cats.append(catpl) text = page.get() text = wikipedia.replaceCategoryLinks(text, cats) page.put(text) class CategoryMoveRobot: def __init__(self, oldCatTitle, newCatTitle): self.oldCat = catlib.Category(wikipedia.getSite(), 'Category:' + oldCatTitle) self.newCatTitle = newCatTitle # set edit summary message wikipedia.setAction(wikipedia.translate(wikipedia.getSite(),msg_change) % self.oldCat.title()) def run(self): newCat = catlib.Category(wikipedia.getSite(), 'Category:' + self.newCatTitle) gen = pagegenerators.CategorizedPageGenerator(self.oldCat, recurse = False) preloadingGen = pagegenerators.PreloadingGenerator(gen) for article in preloadingGen: catlib.change_category(article, self.oldCat, newCat) # TODO: create subcategory generator subcategories = self.oldCat.subcategories(recurse = 0) if len(subcategories) == 0: wikipedia.output(u'There are no subcategories in category ' + self.oldCat.title()) else: for subcategory in subcategories: catlib.change_category(subcategory, self.oldCat, newCat) if self.oldCat.exists(): # try to copy page contents to new cat page if self.oldCat.copyTo(newCatTitle): if self.oldCat.isEmpty(): reason = wikipedia.translate(wikipedia.getSite(), deletion_reason_move) % newCatTitle self.oldCat.delete(reason) else: wikipedia.output('Couldn\'t copy contents of %s because %s already exists.' % (self.oldCatTitle, self.newCatTitle)) class CategoryRemoveRobot: ''' Removes the category tag from all pages in a given category and from the category pages of all subcategories, without prompting. Doesn't remove category tags pointing at subcategories. ''' deletion_reason_remove = { 'de':u'Bot: Kategorie wurde aufgelöst', 'en':u'Robot: Category was disbanded', 'ia':u'Robot: Categoria esseva dissolvite', } msg_remove={ 'da':u'Robot: Fjerner fra %s', 'de':u'Bot: Entferne aus %s', 'en':u'Robot: Removing from %s', 'es':u'Bot: Eliminada de la %s', 'ia':u'Robot: Eliminate de %s', 'is':u'Vélmenni: Fjarlægi [[%s]]', 'nl':u'Bot: Verwijderd uit %s', 'pt':u'Bot: Removendo [[%s]]', 'sr':u'Бот: Уклањање из категорије [[Категорија:%s|%s]]', } def __init__(self, catTitle): self.cat = catlib.Category(wikipedia.getSite(), 'Category:' + catTitle) # get edit summary message wikipedia.setAction(wikipedia.translate(wikipedia.getSite(), self.msg_remove) % self.cat.title()) def run(self): articles = self.cat.articles(recurse = 0) if len(articles) == 0: wikipedia.output(u'There are no articles in category %s' % self.cat.title()) else: for article in articles: catlib.change_category(article, self.cat, None) # Also removes the category tag from subcategories' pages subcategories = self.cat.subcategories(recurse = 0) if len(subcategories) == 0: wikipedia.output(u'There are no subcategories in category %s' % self.cat.title()) else: for subcategory in subcategories: catlib.change_category(subcategory, self.cat.title(), None) if self.cat.exists() and self.cat.isEmpty(): reason = wikipedia.translate(wikipedia.getSite(), self.deletion_reason_remove) self.cat.delete(reason) class CategoryTidyRobot: """ Script to help a human to tidy up a category by moving its articles into subcategories Specify the category name on the command line. The program will pick up the page, and look for all subcategories and supercategories, and show them with a number adjacent to them. It will then automatically loop over all pages in the category. It will ask you to type the number of the appropriate replacement, and perform the change robotically. If you don't want to move the article to a subcategory or supercategory, but to another category, you can use the 'j' (jump) command. Typing 's' will leave the complete page unchanged. Typing '?' will show you the first few bytes of the current page, helping you to find out what the article is about and in which other categories it currently is. Important: * this bot is written to work with the MonoBook skin, so make sure your bot account uses this skin """ def __init__(self, catTitle, catDB): self.catTitle = catTitle self.catDB = catDB # This is a purely interactive robot. We set the delays lower. wikipedia.get_throttle.setDelay(1) wikipedia.put_throttle.setDelay(10) def move_to_category(self, article, original_cat, current_cat): ''' Given an article which is in category original_cat, ask the user if it should be moved to one of original_cat's subcategories. Recursively run through subcategories' subcategories. NOTE: current_cat is only used for internal recursion. You should always use current_cat = original_cat. ''' print wikipedia.output(u'Treating page %s, currently in category %s' % (article.title(), current_cat.title())) subcatlist = self.catDB.getSubcats(current_cat) supercatlist = self.catDB.getSupercats(current_cat) print if len(subcatlist) == 0: print 'This category has no subcategories.' print if len(supercatlist) == 0: print 'This category has no supercategories.' print # show subcategories as possible choices (with numbers) for i in range(len(supercatlist)): # layout: we don't expect a cat to have more than 10 supercats wikipedia.output(u'u%d - Move up to %s' % (i, supercatlist[i].title())) for i in range(len(subcatlist)): # layout: we don't expect a cat to have more than 100 subcats wikipedia.output(u'%2d - Move down to %s' % (i, subcatlist[i].title())) print ' j - Jump to another category' print ' n - Skip this article' print ' r - Remove this category tag' print ' ? - Read the page' wikipedia.output(u'Enter - Save category as %s' % current_cat.title()) flag = False length = 1000 while not flag: print '' choice=wikipedia.input(u'Choice:') if choice == 'n': flag = True elif choice == '': wikipedia.output(u'Saving category as %s' % current_cat.title()) if current_cat == original_cat: print 'No changes necessary.' else: catlib.change_category(article, original_cat, current_cat) flag = True elif choice == 'j': newCatTitle = wikipedia.input(u'Please enter the category the article should be moved to:') newCat = catlib.Category(wikipedia.getSite(), 'Category:' + newCatTitle) # recurse into chosen category self.move_to_category(article, original_cat, newCat) flag = True elif choice == 'r': # remove the category tag catlib.change_category(article, original_cat, None) flag = True elif choice == '?': print '' full_text = article.get() print '' wikipedia.output(full_text[0:length]) # if categories possibly weren't visible, show them additionally # (maybe this should always be shown?) if len(full_text) > length: print '' print 'Original categories: ' for cat in article.categories(): wikipedia.output(u'* %s' % cat.title()) # show more text if the user uses this function again length = length+500 elif choice[0] == 'u': try: choice=int(choice[1:]) except ValueError: # user pressed an unknown command. Prompt him again. continue self.move_to_category(article, original_cat, supercatlist[choice]) flag = True else: try: choice=int(choice) except ValueError: # user pressed an unknown command. Prompt him again. continue # recurse into subcategory self.move_to_category(article, original_cat, subcatlist[choice]) flag = True def run(self): cat = catlib.Category(wikipedia.getSite(), 'Category:' + self.catTitle) # get edit summary message wikipedia.setAction(wikipedia.translate(wikipedia.getSite(), msg_change) % cat.title()) articles = cat.articles(recurse = 0) if len(articles) == 0: wikipedia.output(u'There are no articles in category ' + catTitle) else: for article in articles: print print '===================================================================' self.move_to_category(article, cat, cat) class CategoryTreeRobot: ''' Robot to create tree overviews of the category structure. Parameters: * catTitle - The category which will be the tree's root. * catDB - A CategoryDatabase object * maxDepth - The limit beyond which no subcategories will be listed. This also guarantees that loops in the category structure won't be a problem. * filename - The textfile where the tree should be saved; None to print the tree to stdout. ''' def __init__(self, catTitle, catDB, filename = None, maxDepth = 10): self.catTitle = catTitle self.catDB = catDB self.filename = filename # TODO: make maxDepth changeable with a parameter or config file entry self.maxDepth = maxDepth def treeview(self, cat, currentDepth = 0, parent = None): ''' Returns a multi-line string which contains a tree view of all subcategories of cat, up to level maxDepth. Recursively calls itself. Parameters: * cat - the Category of the node we're currently opening * currentDepth - the current level in the tree (for recursion) * parent - the Category of the category we're coming from ''' # Translations to say that the current category is in more categories than # the one we're coming from also_in_cats = { 'da': u'(også i %s)', 'de': u'(auch in %s)', 'en': u'(also in %s)', 'fr': u'(également dans %s)', 'ia': u'(equalmente in %s)', 'is': u'(einnig í %s)', 'pt': u'(também em %s)', 'ср': u'(такође у %s)', } result = u'#' * currentDepth result += '[[:%s|%s]]' % (cat.title(), cat.title().split(':', 1)[1]) result += ' (%d)' % len(self.catDB.getArticles(cat)) # We will remove an element of this array, but will need the original array # later, so we create a shallow copy with [:] supercats = self.catDB.getSupercats(cat)[:] # Find out which other cats are supercats of the current cat try: supercats.remove(parent) except: pass if supercats != []: supercat_names = [] for i in range(len(supercats)): # create a list of wiki links to the supercategories supercat_names.append('[[:%s|%s]]' % (supercats[i].title(), supercats[i].title().split(':', 1)[1])) # print this list, separated with commas, using translations given in also_in_cats result += ' ' + wikipedia.translate(wikipedia.getSite(), also_in_cats) % ', '.join(supercat_names) result += '\n' if currentDepth < self.maxDepth: for subcat in self.catDB.getSubcats(cat): # recurse into subdirectories result += self.treeview(subcat, currentDepth + 1, parent = cat) else: if self.catDB.getSubcats(cat) != []: # show that there are more categories beyond the depth limit result += '#' * (currentDepth + 1) + '[...]\n' return result def run(self): """ Prints the multi-line string generated by treeview or saves it to a file. Parameters: * catTitle - the title of the category which will be the tree's root * maxDepth - the limit beyond which no subcategories will be listed """ cat = catlib.Category(wikipedia.getSite(), 'Category:' + self.catTitle) tree = self.treeview(cat) if filename: wikipedia.output(u'Saving results in %s' % filename) import codecs f = codecs.open(filename, 'a', 'utf-8') f.write(tree) f.close() else: wikipedia.output(tree) if __name__ == "__main__": try: catDB = CategoryDatabase() action = None sort_by_last_name = False restore = False for arg in sys.argv[1:]: arg = wikipedia.argHandler(arg, 'category') if arg: if arg == 'add': action = 'add' elif arg == 'remove': action = 'remove' elif arg == 'move': action = 'move' elif arg == 'tidy': action = 'tidy' elif arg == 'tree': action = 'tree' elif arg == '-person': sort_by_last_name = True elif arg == '-rebuild': catDB.rebuild() if action == 'add': add_category(sort_by_last_name) elif action == 'remove': catTitle = wikipedia.input(u'Please enter the name of the category that should be removed:') bot = CategoryRemoveRobot(catTitle) bot.run() elif action == 'move': oldCatTitle = wikipedia.input(u'Please enter the old name of the category:') newCatTitle = wikipedia.input(u'Please enter the new name of the category:') bot = CategoryMoveRobot(oldCatTitle, newCatTitle) bot.run() elif action == 'tidy': catTitle = wikipedia.input(u'Which category do you want to tidy up?') bot = CategoryTidyRobot(catTitle, catDB) bot.run() elif action == 'tree': catTitle = wikipedia.input(u'For which category do you want to create a tree view?') filename = wikipedia.input(u'Please enter the name of the file where the tree should be saved, or press enter to simply show the tree:') bot = CategoryTreeRobot(catTitle, catDB, filename) bot.run() else: # show help wikipedia.output(__doc__, 'utf-8') finally: catDB.dump() wikipedia.stopme()