source: ValBot/Python/check_interwiki_links.py@ 1192

Last change on this file since 1192 was 1192, checked in by iritscen, 7 weeks ago

ValBot: Corrected syntax of two strings with regex patterns.
File size: 12.5 KB
Line 
1# Check Interwiki Links
2# by iritscen@yahoo.com
3# Looks at each link on a page (or in all the pages in a category) which uses a registered
4# interwiki prefix and loads the linked page, verifying that it exists and that any section
5# link, if present, is valid as well. The output will use the word "ERROR" when it cannot
6# validate the interwiki link.
7# Recommended viewing width:
8# |---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---|
9
10import os
11
12from urllib.parse import urljoin
13
14import pywikibot
15import bs4
16import re
17import requests # for listing members with dir()
18
19from pywikibot.bot import QuitKeyboardInterrupt
20from pywikibot import pagegenerators
21from pywikibot.tools.formatter import color_format
22from pywikibot.comms.http import fetch
23from pywikibot.specialbots import UploadRobot
24from bs4 import BeautifulSoup
25
26# Parallel arrays based on https://wiki.oni2.net/Special:Interwiki
27interwiki_prefixes = ('acronym', 'cache', 'commons', 'dictionary', 'google', 'metawikimedia', 'mw', 'wikibooks', 'wikidata', 'wikimedia', 'wikinews', 'wikipedia', 'wikiquote', 'wikisource', 'wikispecies', 'wikiversity', 'wikivoyage', 'wikt', 'wiktionary', 'wp')
28
29interwiki_urls = ('http://www.acronymfinder.com/~/search/af.aspx?string=exact&Acronym=', 'http://www.google.com/search?q=cache:', 'https://commons.wikimedia.org/wiki/', 'http://www.dict.org/bin/Dict?Database=*&Form=Dict1&Strategy=*&Query=', 'http://www.google.com/search?q=', 'https://meta.wikimedia.org/wiki/', 'https://www.mediawiki.org/wiki/', 'https://en.wikibooks.org/wiki/', 'https://www.wikidata.org/wiki/', 'https://foundation.wikimedia.org/wiki/', 'https://en.wikinews.org/wiki/', 'https://en.wikipedia.org/wiki/', 'https://en.wikiquote.org/wiki/', 'https://wikisource.org/wiki/', 'https://species.wikimedia.org/wiki/', 'https://en.wikiversity.org/wiki/', 'https://en.wikivoyage.org/wiki/', 'https://en.wiktionary.org/wiki/', 'https://en.wiktionary.org/wiki/', 'https://en.wikipedia.org/wiki/')
30
31# Initialize globals
32debug = 0
33pages_checked = 0
34iw_found = 0
35errors_issued = 0
36name_printed = 0
37
38# Prints the name of a page on which something occurred, if it has not been printed before
39def possibly_print(page_name):
40 global debug
41 global name_printed
42
43 if not name_printed and not debug:
44 pywikibot.stdout('')
45 pywikibot.stdout('From page "{}":'.format(page_name))
46 name_printed = 1
47
48# Search a page for the section specified in the link
49def find_section(page_text, page_name, page_slug, prefix, print_result):
50 global errors_issued
51
52 # Isolate section link
53 target_page_name, anchor_name = page_slug.split('#')
54 target_page_name_human = target_page_name.replace('_', ' ')
55
56 # Convert dot-notation hex entities to proper characters
57 anchor_name = anchor_name.replace('.22', '"')
58 anchor_name = anchor_name.replace('.27', '\'')
59 anchor_name = anchor_name.replace('.28', '(')
60 anchor_name = anchor_name.replace('.29', ')')
61
62 # Read linked page to see if it really has this anchor link
63 soup = BeautifulSoup(page_text, 'html.parser')
64 found_section = False
65 for the_tag in soup.findAll('span'): # search for span with ID matching the section name
66 tag_name = the_tag.get('id', None)
67 if tag_name == anchor_name:
68 found_section = True
69 break
70 if found_section == False:
71 for the_tag in soup.findAll('div'): # search for div with ID matching the section name
72 tag_name = the_tag.get('id', None)
73 if tag_name == anchor_name:
74 found_section = True
75 break
76 if found_section == False:
77 for the_tag in soup.findAll('h2'): # search for h2 with ID matching the section name
78 tag_name = the_tag.get('id', None)
79 if tag_name == anchor_name:
80 found_section = True
81 break
82 if found_section == False:
83 for the_tag in soup.findAll('h3'): # search for h3 with ID matching the section name
84 tag_name = the_tag.get('id', None)
85 if tag_name == anchor_name:
86 found_section = True
87 break
88 if found_section == False:
89 for the_tag in soup.findAll('h4'): # search for h4 with ID matching the section name
90 tag_name = the_tag.get('id', None)
91 if tag_name == anchor_name:
92 found_section = True
93 break
94 if found_section == False:
95 possibly_print(page_name)
96 pywikibot.stdout(' ERROR: Could not find section "{0}" on {1} page "{2}".'.format(anchor_name, prefix, target_page_name_human))
97 errors_issued = errors_issued + 1
98 elif print_result == True:
99 pywikibot.stdout(' The section "{0}" was found on {1} page "{2}".'.format(anchor_name, prefix, target_page_name_human))
100
101# For a link that redirected us to another page, extract the name of the target page from
102# the target page's source
103def find_canonical_link(page_text, page_name, page_slug, prefix, prefix_url):
104 # Extract link from this markup which contains name of redirected-to page:
105 # <link rel="canonical" href="https://en.wikipedia.org/wiki/Page_name"/>
106 canonical_name = page_text.split('<link rel="canonical" href="')[-1]
107 prefix_length = len(prefix_url)
108 canonical_name = canonical_name[prefix_length:]
109 tag_end = canonical_name.find('">')
110
111 if tag_end == -1:
112 pywikibot.stdout(' ERROR: The {0} link "{1}" is a redirect page, but this script could not isolate the target page name.'.format(prefix, page_slug))
113 errors_issued = errors_issued + 1
114 else:
115 canonical_name = canonical_name[:tag_end]
116 if len(canonical_name) > 100:
117 # Certain things can cause the trim to fail; report error and avoid slamming the
118 # output with massive page source from a failed trim
119 pywikibot.stdout(' ERROR: The {0} link "{1}" is a redirect to "{2}…" (string overflow).'.format(prefix, page_slug, canonical_name[:100]))
120 errors_issued = errors_issued + 1
121 else:
122 canonical_name = canonical_name.replace('_', ' ')
123 if '#' in page_slug:
124 _, anchor_name = page_slug.split('#')
125 pywikibot.stdout(' The {0} link "{1}" is a redirect to "{2}#{3}", which is a valid page. Checking section link….'.format(prefix, page_slug, canonical_name, anchor_name))
126 find_section(page_text, page_name, page_slug, prefix, True)
127 else:
128 pywikibot.stdout(' The {0} link "{1}" is a redirect to "{2}", which is a valid page.'.format(prefix, page_slug, canonical_name))
129
130# Test an interwiki link and look for a section link if applicable
131def test_interwiki_link(prefix, prefix_url, iw_url, page_name, page_slug):
132 global errors_issued
133
134 response = fetch(iw_url)
135
136 # One way we tell that a redirect occurred is by checking fetch's history, as it
137 # automatically follows redirects. This will catch formal redirects which come from pages
138 # such as Special:PermanentLink.
139 if response.history != []:
140 possibly_print(page_name)
141
142 if page_slug.startswith('WP:') and page_slug == page_slug.upper():
143 pywikibot.stdout(' Got redirection code "{0}" for {1} link "{2}". This appears to be a deliberate use of a Wikipedia shortcut. Checking the target page….'.format(response.history[0], prefix, page_slug))
144 find_canonical_link(response.text, page_name, page_slug, prefix, prefix_url)
145 else:
146 permalink1 = 'Special:PermanentLink/'.lower()
147 permalink2 = 'Special:Permalink/'.lower()
148 page_slug_lower = page_slug.lower()
149 if page_slug_lower.startswith(permalink1) or page_slug_lower.startswith(permalink2):
150 pywikibot.stdout(' Got redirection code "{0}" for {1} permanent revision link "{2}". Checking the target page….'.format(response.history[0], prefix, page_slug))
151 find_canonical_link(response.text, page_name, page_slug, prefix, prefix_url)
152 else:
153 pywikibot.stdout(' ERROR: Unrecognized type of redirection (code "{0}") for {1} link "{2}". You should check the link manually.'.format(response.history[0], prefix, page_slug))
154 errors_issued = errors_issued + 1
155 elif response.status_code != 200:
156 possibly_print(page_name)
157 pywikibot.stdout(' ERROR: Got response code {0} for {1} link "{2}". The page may not exist.'.format(response.status_code, prefix, page_slug))
158 errors_issued = errors_issued + 1
159 # However the usual way that a redirect occurs is that MediaWiki redirects us sneakily
160 # using JavaScript, while returning code OK 200 as if the link was correct; this happens
161 # when a redirect page is accessed. We must detect these soft redirects by looking at the
162 # page source to find the redirect note inserted at the top of the page for the reader.
163 elif 'Redirected from <a' in response.text:
164 possibly_print(page_name)
165 pywikibot.stdout(' Got silently redirected by {0} link "{1}". Checking the target page….'.format(prefix, page_slug))
166 find_canonical_link(response.text, page_name, page_slug, prefix, prefix_url)
167 elif '#' in page_slug:
168 find_section(response.text, page_name, page_slug, prefix, False)
169
170# Searches the given page text for interwiki links
171def scan_for_interwiki_links(page_text, page_name):
172 global debug
173 global pages_checked
174 global iw_found
175 global name_printed
176 pages_checked = pages_checked + 1
177 cur_prefix = 0
178 name_printed = 0
179
180 for prefix in interwiki_prefixes:
181 # Isolate strings that start with "[[prefix:" and end with "|" or "]"
182 iw_link = r"\[\[" + prefix + r":[^|\]]*(\||\])"
183 for match in re.finditer(iw_link, page_text):
184 # Extract just the page title from this regex match
185 s = match.start() + 2 + len(prefix) + 1
186 e = match.end() - 1
187
188 # Commonly we use spaces instead of underscores, so fix that before querying
189 page_slug = page_text[s:e].replace(' ', '_')
190
191 # But use spaces for title when printing it
192 page_title_human = page_slug.replace('_', ' ')
193 if debug: pywikibot.stdout(' Validating {0} link "{1}"'.format(prefix, page_title_human))
194 iw_found = iw_found + 1
195
196 # Construct full URL for the particular wiki
197 iw_url = interwiki_urls[cur_prefix] + page_slug
198
199 # Adjust URL if this is a foreign-language WP link
200 if re.match("^[a-zA-Z]{2}:", page_slug):
201 lang_code = page_slug[0:2] + "."
202 # "wp:" is the Wikipedia: namespace, not a language
203 if lang_code != "wp." and lang_code != "WP.":
204 iw_url = iw_url.replace('en.', lang_code)
205 iw_url = iw_url.replace(page_slug[0:3], '')
206
207 # Test the URL
208 test_interwiki_link(prefix, interwiki_urls[cur_prefix], iw_url, page_name, page_slug)
209 cur_prefix = cur_prefix + 1
210
211# Print a wrap-up message
212def print_summary():
213 global pages_checked
214 global iw_found
215 global errors_issued
216
217 page_str = "pages"
218 if pages_checked == 1:
219 page_str = "page"
220
221 link_str = "links"
222 if iw_found == 1:
223 link_str = "link"
224
225 pywikibot.stdout('Checked {0} {1} and found {2} interwiki {3}.'.format(pages_checked, page_str, iw_found, link_str))
226
227 error_str = "errors were"
228 if errors_issued == 1:
229 error_str = "error was"
230
231 pywikibot.stdout('{0} {1} encountered in validating these links.'.format(errors_issued, error_str))
232
233# Main function
234def main(*args):
235 global debug
236 search_cat = ''
237 search_page = ''
238
239 # Process arguments
240 local_args = pywikibot.handle_args(args)
241 for arg in local_args:
242 if arg.startswith('-cat:'):
243 search_cat = arg[5:]
244 elif arg.startswith('-page:'):
245 search_page = arg[6:]
246 elif arg == '-dbg':
247 debug = 1
248 else:
249 pywikibot.stdout('Unknown argument "{}". Exiting.'.format(arg))
250 return
251
252 #pywikibot.stdout('The members of the requests.models.Response class are:')
253 #pywikibot.stdout(format(dir(requests.models.Response)))
254 #return
255
256 # Check specified page or loop through specified category and check all pages
257 site = pywikibot.Site()
258 if search_cat != '':
259 cat_obj = pywikibot.Category(site, search_cat)
260 generator = pagegenerators.CategorizedPageGenerator(cat_obj, recurse=True)
261 for page in pagegenerators.PreloadingGenerator(generator, 100):
262 if debug: pywikibot.stdout('Checking page "{}"'.format(page.title()))
263 scan_for_interwiki_links(page.text, page.title())
264 elif search_page != '':
265 page = pywikibot.Page(site, search_page)
266 if debug: pywikibot.stdout('Checking page "{}"'.format(page.title()))
267 scan_for_interwiki_links(page.text, page.title())
268
269 # Print the results
270 print_summary()
271
272if __name__ == '__main__':
273 main()
Note: See TracBrowser for help on using the repository browser.