Coverage for erfgoedbot/categorize_images.py : 38%

#!/usr/bin/python 

# -*- coding: utf-8 -*- 

''' 

Bot to move images from commonsCategoryBase to subcategories based on the monument 

template and Commonscat links at the Wikipedia. 

First checks if monument article or it's categories have commonscat template, 

if not then checks if monuments list or it's categories have commonscat template. 

Usage: 

* To process all: 

python categorize_images.py 

* Just process one source: 

python categorize_images.py -countrycode:ee -langcode:et 

''' 

import json 

import os 

from collections import OrderedDict 

import yaml 

import pywikibot 

from pywikibot import pagegenerators, textlib 

import erfgoedbot.common as common 

import erfgoedbot.monuments_config as mconfig 

from erfgoedbot.database_connection import ( 

close_database_connection, 

connect_to_monuments_database 

) 

from erfgoedbot.statistics_table import StatisticsTable 

_logger = "categorize_images" 

SKIP_LIST = [ 

('au', 'en'), # Australia in English 

('cn', 'en'), 

('gb-sct', 'en'), # Scotland in English 

('dz', 'ar'), # Algeria in Arabic 

('ir', 'fa'), 

('it', 'it'), 

('jo', 'ar'), 

('ge', 'ka'), 

('gh', 'en'), # Ghana in English 

('mt', 'de'), # Malta in German 

('np', 'en'), 

('pe', 'es'), # Peru in Spanish 

('sv', 'es'), # El Salvador in Spanish 

('ve', 'es'), # Venezuela in Spanish 

] 

class NoMonumentIdentifierFoundException(pywikibot.exceptions.PageRelatedError): 

message = "No Monument Identifier could be found for %s" 

pass 

class NoCommonsCatFromWikidataItemException(pywikibot.exceptions.PageRelatedError): 

message = "No CommonsCat could be retrieved through Wikidata for %s" 

pass 

class NoCategoryToAddException(Exception): 

pass 

def _load_wikipedia_commonscat_templates(): 

data_dir = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'data') 

json_file = os.path.join(data_dir, 'wikipedia_commonscat_templates.json') 

return json.load(open(json_file, 'r')) 

def _load_ignored_categories(): 

data_dir = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'data') 

yaml_file = os.path.join(data_dir, 'ignore_commons_categories.yml') 

raw_list = yaml.load(open(yaml_file, 'r')) 

commons_site = pywikibot.Site('commons', 'commons') 

return [pywikibot.Category(commons_site, cat) for cat in raw_list] 

def _get_commons_template(template_name): 

site = pywikibot.Site('commons', 'commons') 

return pywikibot.Page(site, 'Template:%s' % template_name) 

def categorizeImage( 

countrycode, lang, commonsTemplateName, commonsCategoryBase, 

commonsCatTemplates, page, conn, cursor, harvest_type): 

pywikibot.log('Working on: %s' % page.title()) 

commonsTemplate = _get_commons_template(commonsTemplateName) 

currentcats = list(page.categories()) 

if commonsCategoryBase not in currentcats: 

pywikibot.log('%s category not found at: %s. Someone probably already categorized it.' % ( 

commonsCategoryBase, page.title())) 

return False 

if 'Wikipedia image placeholders for cultural heritage monuments' in currentcats: 

pywikibot.log('%s in %s is a placeholder, skipping it.' % ( 

page.title(), commonsCategoryBase)) 

return False 

templates = page.templates() 

if commonsTemplate not in templates: 

pywikibot.log('%s template not found at: %s' % ( 

commonsTemplate, page.title())) 

return False 

try: 

monumentId = get_monument_id(page, commonsTemplate) 

except NoMonumentIdentifierFoundException: 

pywikibot.warning('Didn\'t find a valid monument identifier at: %s' % ( 

page.title(),)) 

return False 

monData = getMonData(countrycode, lang, monumentId, conn, cursor) 

if not monData: 

try: 

monumentId = int(monumentId) 

monData = getMonData(countrycode, lang, monumentId, conn, cursor) 

except ValueError: 

pywikibot.debug( 

'Can\'t convert %s to an integer' % (monumentId,), _logger) 

if not monData: 

# Triage as log since there are plenty of valid reasons for this 

pywikibot.log( 

'Monument with id %s not in monuments database' % (monumentId, )) 

return False 

(newcats, categorisation_method) = get_new_categories(monumentId, monData, lang, commonsCatTemplates, harvest_type) 

# See if one of the three options worked 

if newcats: 

comment = 'Adding categories based on [[Template:%s]] with identifier %s (method %s)' % ( 

commonsTemplateName, monumentId, categorisation_method) 

return replace_default_cat_with_new_categories_in_image( 

page, commonsCategoryBase, newcats, comment, verbose=True) 

else: 

pywikibot.log('Categories not found for %s' % page.title()) 

def get_monument_id(page, commonsTemplate): 

monumentId = None 

for (template, params) in page.templatesWithParams(): 

if template == commonsTemplate: 

if len(params) >= 1: 

try: 

monumentId = params[0] 

except ValueError: 

pywikibot.warning( 

'Unable to extract a valid id for %s on %s' % ( 

template, page.title())) 

break 

if not monumentId: 

raise NoMonumentIdentifierFoundException(page) 

return monumentId 

def get_new_categories(monumentId, monData, lang, commonsCatTemplates, harvest_type): 

(monumentName, monumentCommonscat, 

monumentArticleTitle, monumentSource, project) = monData 

commons_site = pywikibot.Site('commons', 'commons') 

newcats = [] 

categorisation_method = '' 

# First try to add a category based on the commonscat field in the list 

if monumentCommonscat: 

# Might want to include some error checking here 

try: 

cat = pywikibot.Category(commons_site, monumentCommonscat) 

newcats.append(cat) 

categorisation_method = 'A: CommonsCat in the monument list' 

except ValueError: 

pywikibot.warning( 

'The Commonscat field for %s contains an invalid category %s' % ( 

monumentId, monumentCommonscat)) 

except pywikibot.exceptions.InvalidTitle: 

pywikibot.warning( 

'Incorrect category title %s' % (monumentCommonscat,)) 

# Option two is to use the article about the monument and see if it has 

# Commonscat links 

if not newcats: 

monumentArticle = None 

if monumentArticleTitle: 

project_site = pywikibot.Site(lang, project) 

monumentArticle = pywikibot.Page(project_site, monumentArticleTitle) 

if monumentArticle: 

try: 

if monumentArticle.isRedirectPage(): 

monumentArticle = monumentArticle.getRedirectTarget() 

try: 

for commonsCatTemplateName in commonsCatTemplates: 

commonsCatTemplate = pywikibot.Page(project_site, 'Template:%s' % commonsCatTemplateName) 

if is_template_present_in_page(commonsCatTemplate, monumentArticle): 

(new_cat, method) = getCategoryFromCommonscat(monumentArticle, commonsCatTemplates) 

newcats.append(new_cat) 

categorisation_method = 'B%s: CommonsCat on the monument article' % method 

except pywikibot.SectionError: 

pywikibot.warning('Incorrect redirect at %s' % ( 

monumentArticle.title(),)) 

except pywikibot.exceptions.InvalidTitle: 

pywikibot.warning('Incorrect article title %s' % ( 

monumentArticleTitle,)) 

except pywikibot.exceptions.Error as e: 

pywikibot.error('Error occured with monument %s: %s' % ( 

monumentId, str(e))) 

# Option three is to see if the list contains Commonscat links (whole list) 

if not newcats and harvest_type != 'sparql': 

monumentList = getList(lang, project, monumentSource) 

# print monumentList 

if not monumentList: 

return ([], '') 

if monumentList.isRedirectPage(): 

monumentList = monumentList.getRedirectTarget() 

try: 

(newcats, categorisation_method) = get_categories_from_source_page(monumentList, commonsCatTemplates) 

except pywikibot.exceptions.Error as e: 

pywikibot.error('Error occured with monument %s: %s' % ( 

monumentId, str(e))) 

return (newcats, categorisation_method) 

def replace_default_cat_with_new_categories_in_image( 

page, base_category, new_categories, comment, verbose=False): 

old_text = page.get() 

old_categories = list(page.categories()) 

# ensure base category is never re-added 

old_categories.append(base_category) 

categories_to_add = filter_out_categories_to_add(new_categories, 

old_categories) 

try: 

final_text = replace_default_cat_with_new_categories_in_image_text( 

old_text, base_category, categories_to_add) 

except NoCategoryToAddException: 

return False 

if verbose: 

pywikibot.showDiff(old_text, final_text) 

try: 

common.save_to_wiki_or_local(page, comment, final_text) 

return True 

except pywikibot.EditConflict: 

pywikibot.log( 

'Got an edit conflict. Someone else beat me to it at %s' % page.title()) 

return False 

def replace_default_cat_with_new_categories_in_image_text( 

old_text, base_category, new_categories): 

"""Add new categories to page text and remove any base_category.""" 

if not new_categories: 

# No categories to add. We do not want to remove the base one, 

raise NoCategoryToAddException() 

# Remove base category 

page_text_without_base_category = textlib.replaceCategoryInPlace( 

old_text, base_category, None) 

final_text = textlib.replaceCategoryLinks( 

page_text_without_base_category, new_categories, addOnly=True) 

return final_text 

def filter_out_categories_to_add(new_categories, unwanted_categories): 

""" 

Ensure unwanted categories are not added. 

This included hidden, duplicate, ignored or already present categories. 

Requires the inputs to be lists of pywikibot.Category. 

""" 

candidate_categories = set(new_categories) - set(unwanted_categories) 

candidate_categories -= set(_load_ignored_categories()) 

final_categories = [cat for cat in list(candidate_categories) if not cat.isHiddenCategory()] 

return final_categories 

def getMonData(countrycode, lang, monumentId, conn, cursor): 

"""Get monument name and source from db.""" 

query = ( 

"SELECT `name`, `commonscat`, `monument_article`, `source`, `project` " 

"FROM monuments_all " 

"WHERE (country=%s AND lang=%s AND id=%s) " 

"LIMIT 1") 

cursor.execute(query, (countrycode, lang, monumentId)) 

try: 

row = cursor.fetchone() 

return row 

except TypeError: 

pywikibot.warning('Didn\'t find anything for id %s' % (monumentId,)) 

return False 

def getList(lang, project, monumentSource): 

"""Get the listpage, if not harvested from a sparql query.""" 

if monumentSource: 

try: 

page_title, found_site = common.get_source_page(monumentSource) 

except ValueError: 

return False 

if (project, lang) != found_site: 

return False 

site = pywikibot.Site(lang, project) 

return pywikibot.Page(site, page_title) 

else: 

return False 

def is_template_present_in_page(template, page): 

"""Return whether a template is present in a page (as in directly transcluded.)""" 

return template in [x[0] for x in page.templatesWithParams()] 

def get_categories_from_source_page(page, commonsCatTemplates): 

""" 

Get Commons categories based on page. 

1. If page contains a Commonscat template, use that category 

2. Else, try getting it from Wikidata 

3. Else pull Commonscat links from upper categories 

""" 

new_categories = set() 

categorisation_method = '' 

for commonsCatTemplateName in commonsCatTemplates: 

commonsCatTemplate = pywikibot.Page(page.site, 'Template:%s' % commonsCatTemplateName) 

if is_template_present_in_page(commonsCatTemplate, page): 

(new_cat, _) = getCategoryFromCommonscat(page, commonsCatTemplates) 

new_categories.add(new_cat) 

categorisation_method = 'C1: CommonsCat on the monument list page' 

if not len(new_categories): 

try: 

site = pywikibot.Site('commons', 'commons') 

cat_title = get_Commons_category_via_Wikidata(page) 

cat = pywikibot.Category(site, cat_title) 

new_categories.add(cat) 

categorisation_method = 'C2: via Wikidata on the monument list page' 

except NoCommonsCatFromWikidataItemException: 

pass 

if not len(new_categories): 

new_categories = get_categories_from_upper_categories(page, commonsCatTemplates) 

categorisation_method = 'D: from upper categories of monument list page' 

return (new_categories, categorisation_method) 

def get_categories_from_upper_categories(page, commonsCatTemplates): 

new_categories = set() 

for cat in page.categories(): 

for commonsCatTemplateName in commonsCatTemplates: 

commonsCatTemplate = pywikibot.Page(page.site, 'Template:%s' % commonsCatTemplateName) 

if is_template_present_in_page(commonsCatTemplate, cat): 

(new_cat, method) = getCategoryFromCommonscat(cat, commonsCatTemplates) 

new_categories.add(new_cat) 

try: 

site = pywikibot.Site('commons', 'commons') 

new_cat_title = get_Commons_category_via_Wikidata(cat) 

new_cat = pywikibot.Category(site, new_cat_title) 

new_categories.add(new_cat) 

except NoCommonsCatFromWikidataItemException: 

pass 

return new_categories 

def getCategoryFromCommonscat(page, commonsCatTemplates): 

""" 

Get a Commons category based on a page. 

1. Get category from commonscat template on page 

2. Get category from commonscat property or Commons category sitelink on 

the Wikidata object corresponding to the page 

3. Get category with same name as page 

""" 

cat_title = None 

categorisation_method = '1' # By 'default', we do not rely on Wikidata 

(template, params) = get_commonscat_template_in_page(page, commonsCatTemplates) 

if len(params) >= 1: 

cat_title = params[0] 

if not cat_title: 

try: 

cat_title = get_Commons_category_via_Wikidata(page) 

categorisation_method = '2' 

except NoCommonsCatFromWikidataItemException: 

pass 

if not cat_title: 

categorisation_method = '3' 

cat_title = page.title(with_ns=False) 

site = pywikibot.Site('commons', 'commons') 

cat = pywikibot.Category(site, cat_title) 

return (cat, categorisation_method) 

def get_commonscat_template_in_page(page, commonsCatTemplates): 

for (template, params) in page.templatesWithParams(): 

if template.title(with_ns=False) in commonsCatTemplates: 

return (template, params) 

def get_Commons_category_via_Wikidata(page): 

""" 

Get Commons Category from the linked Wikidata item and P373. 

Raises: NoCommonsCatFromWikidataItemException if either there is no linked 

item or it does not bear P373 or a sitelink to a category page on 

Commons. 

""" 

try: 

data_item = page.data_item() 

claims = data_item.get()['claims'] 

if 'P373' in claims: 

return 'Category:' + claims['P373'][0].getTarget() 

else: 

commons_site = pywikibot.Site('commons', 'commons') 

commons_page = data_item.getSitelink(commons_site) 

if commons_page.startswith('Category:'): 

return commons_page 

else: 

raise NoCommonsCatFromWikidataItemException(page) 

except (pywikibot.NoPage, KeyError, NoCommonsCatFromWikidataItemException): 

raise NoCommonsCatFromWikidataItemException(page) 

def processCountry(countryconfig, commonsCatTemplates, conn, cursor, 

overridecat=None): 

"""Work on a single country.""" 

if not countryconfig.get('commonsTemplate'): 

# No template found, just skip silently. 

basecat = None 

if countryconfig.get('commonsCategoryBase'): 

basecat = countryconfig.get('commonsCategoryBase') 

return { 

'code': countryconfig.get('country'), 

'lang': countryconfig.get('lang'), 

'cat': basecat, 

'cmt': 'skipped: no template' 

} 

if not countryconfig.get('commonsCategoryBase') and not overridecat: 

# No template found, just skip silently. 

commonsTemplate = countryconfig.get('commonsTemplate') 

return { 

'code': countryconfig.get('country'), 

'lang': countryconfig.get('lang'), 

'template': commonsTemplate, 

'cmt': 'skipped: no base category' 

} 

if (not commonsCatTemplates): 

# No commonsCatTemplates found, just skip. 

pywikibot.warning( 

'Language: {0} has no commonsCatTemplates set!'.format( 

countryconfig.get('lang'))) 

return False 

totalImages = 0 

categorizedImages = 0 

site = pywikibot.Site('commons', 'commons') 

generator = None 

commonsTemplate = countryconfig.get('commonsTemplate') 

harvest_type = countryconfig.get('type') 

category_name = overridecat or countryconfig.get('commonsCategoryBase') 

commonsCategoryBase = pywikibot.Category( 

site, "%s:%s" % (site.namespace(14), category_name)) 

generator = pagegenerators.CategorizedPageGenerator(commonsCategoryBase) 

# Get a preloading generator with only images 

pgenerator = pagegenerators.PreloadingGenerator( 

pagegenerators.NamespaceFilterPageGenerator(generator, [6])) 

for page in pgenerator: 

totalImages += 1 

success = False 

if not totalImages >= 10000: 

success = categorizeImage( 

countryconfig.get('country'), countryconfig.get('lang'), 

commonsTemplate, commonsCategoryBase, commonsCatTemplates, 

page, conn, cursor, harvest_type) 

if success: 

categorizedImages += 1 

return { 

'code': countryconfig.get('country'), 

'lang': countryconfig.get('lang'), 

'cat': commonsCategoryBase.title(with_ns=False), 

'template': commonsTemplate, 

'total_images': totalImages, 

'cat_images': categorizedImages 

} 

def outputStatistics(statistics): 

"""Output the results of the bot as a nice wikitable.""" 

site = pywikibot.Site('commons', 'commons') 

page = pywikibot.Page( 

site, 'Commons:Monuments database/Categorization/Statistics') 

title_column = OrderedDict([ 

('code', 'country'), 

('lang', '[[:en:List of ISO 639-1 codes|lang]]'), 

('cat', 'Base category'), 

('template', 'Template'), 

('total_images', 'Total images'), 

('cat_images', 'Categorized images'), 

('leftover', 'Images left'), 

('pages_in_cat', 'Current image count') 

]) 

numeric_columns = ('total_images', 'cat_images', 'leftover', 

'pages_in_cat') 

table = StatisticsTable(title_column, numeric_columns) 

total_images_sum = 0 

categorized_images_sum = 0 

leftover_images_sum = 0 

for row in statistics: 

leftover = None 

cat_link = None 

pages_in_cat = None 

template_link = None 

total_images = row.get('total_images') 

cat_images_or_cmt = row.get('cat_images') 

if row.get('cat_images') is not None: 

leftover = row['total_images'] - row['cat_images'] 

else: 

cat_images_or_cmt = row.get('cmt') 

total_images = None 

if row.get('cat'): 

cat_link = '[[:Category:{0}]]'.format(row['cat']) 

pages_in_cat = '{{PAGESINCATEGORY:%s|files}}' % row['cat'] 

if row.get('template'): 

template_link = '{{tl|%s}}' % row['template'] 

table.add_row({ 

'code': row['code'], 

'lang': row['lang'], 

'cat': cat_link, 

'template': template_link, 

'total_images': total_images, 

'cat_images': cat_images_or_cmt, 

'leftover': leftover, 

'pages_in_cat': pages_in_cat 

}) 

# we want pages_in_cat numerically sorted but not summed 

table.totals['pages_in_cat'] = None 

output = table.to_wikitext() 

summary = ( 

'Updating categorization statistics. ' 

'Total: {total_images} Categorized: {cat_images} ' 

'Leftover: {leftover}'.format(**table.get_sum())) 

common.save_to_wiki_or_local(page, summary, output) 

def getCommonscatTemplates(lang=None, project=None): 

""" 

Get the template name in a language on a project. 

Expects the language code and project. 

Return as list containing the primary template and it's alternatives 

""" 

project = project or 'wikipedia' # default to wikipedia 

wikipedia_commonscat_templates = _load_wikipedia_commonscat_templates() 

result = [] 

if project == 'wikipedia' and lang in wikipedia_commonscat_templates: 

(prim, backups) = wikipedia_commonscat_templates[lang] 

else: 

(prim, backups) = wikipedia_commonscat_templates['_default'] 

result.append(prim) 

result += backups 

return result 

def custom_output_statistics_message(country_config, message): 

"""Return a outputStatistics compatible summary for a missing dataset (skipped or failed).""" 

site = pywikibot.Site('commons', 'commons') 

commons_category_base = pywikibot.Category(site, '{ns}:{cat}'.format( 

ns=site.namespace(14), cat=country_config.get('commonsCategoryBase'))) 

return { 

'code': country_config.get('country'), 

'lang': country_config.get('lang'), 

'cat': commons_category_base.title(with_ns=False), 

'template': country_config.get('commonsTemplate'), 

'cmt': message 

} 

def main(): 

countrycode = '' 

lang = '' 

overridecat = '' 

skip_wd = False 

conn = None 

cursor = None 

for arg in pywikibot.handleArgs(): 

option, sep, value = arg.partition(':') 

if option == '-countrycode': 

countrycode = value 

elif option == '-langcode': 

lang = value 

elif option == '-overridecat': 

overridecat = value 

elif option == '-skip_wd': 

skip_wd = True 

else: 

raise Exception( 

'Bad parameters. Expected "-countrycode", "-langcode", ' 

'"-overridecat", "-skip_wd" or pywikibot args. ' 

'Found "{}"'.format(option)) 

if countrycode and lang: 

if not mconfig.countries.get((countrycode, lang)): 

pywikibot.warning( 

'I have no config for countrycode "%s" in language "%s"' % (countrycode, lang)) 

return False 

pywikibot.log( 

'Working on countrycode "%s" in language "%s"' % (countrycode, lang)) 

countryconfig = mconfig.countries.get((countrycode, lang)) 

commonsCatTemplates = getCommonscatTemplates( 

lang, countryconfig.get('project')) 

# print commonsCatTemplates 

(conn, cursor) = connect_to_monuments_database() 

processCountry(countryconfig, commonsCatTemplates, conn, cursor, 

overridecat=overridecat) 

close_database_connection(conn, cursor) 

elif countrycode or lang: 

raise Exception('The "countrycode" and "langcode" arguments must ' 

'be used together.') 

else: 

statistics = [] 

for (countrycode, lang), countryconfig in mconfig.filtered_countries( 

skip_wd=skip_wd): 

if (countrycode, lang) in SKIP_LIST: 

pywikibot.log( 

'Skipping countrycode "%s" in language "%s"' % (countrycode, lang)) 

statistics.append(custom_output_statistics_message(countryconfig, 'skipped: on the skip-list')) 

continue 

pywikibot.log( 

'Working on countrycode "%s" in language "%s"' % (countrycode, lang)) 

commonsCatTemplates = getCommonscatTemplates( 

lang, countryconfig.get('project')) 

(conn, cursor) = connect_to_monuments_database() 

try: 

result = processCountry( 

countryconfig, commonsCatTemplates, conn, cursor) 

except Exception as e: 

pywikibot.error( 

'Unknown error occurred when processing country ' 

'{0} in lang {1}\n{2}'.format(countrycode, lang, str(e))) 

statistics.append(custom_output_statistics_message(countryconfig, 'failed: unexpected error during processing')) 

continue 

finally: 

close_database_connection(conn, cursor) 

if result: 

statistics.append(result) 

outputStatistics(statistics) 

if __name__ == "__main__": 

pywikibot.log('Start of %s' % __file__) 

main()