source: ValBot/Python/check_intrawiki_section_links.py@ 1194

Last change on this file since 1194 was 1194, checked in by iritscen, 4 months ago

ValBot: check_intrawiki_section_links.py now understands text fragment directives.
File size: 19.4 KB
Line 
1# Check Intrawiki Section Links
2# by iritscen@yahoo.com
3# Looks at each wikilink on a page (or in all the pages in a category) for a section link ('#'),
4# and loads the linked page and verifies that the named section actually exists. It also
5# understands section links generated through a call to Template:SectionLink.
6# Recommended viewing width:
7# |---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- --|
8
9import os
10
11from urllib.parse import urljoin
12
13import pywikibot
14import re
15
16from pywikibot.bot import QuitKeyboardInterrupt
17from pywikibot import pagegenerators
18from pywikibot.tools.formatter import color_format
19from pywikibot.comms.http import fetch
20from pywikibot.specialbots import UploadRobot
21from bs4 import BeautifulSoup
22
23# Tuple of OniGalore's namespaces
24intrawiki_prefixes = ('Image', 'Special', 'Talk', 'User', 'User_talk', 'OniGalore', 'OniGalore_talk', 'File', 'File_talk', 'MediaWiki', 'MediaWiki_talk', 'Template', 'Template_talk', 'Help', 'Help_talk', 'Category', 'Category_talk', 'BSL', 'BSL_talk', 'OBD', 'OBD_talk', 'AE', 'AE_talk', 'Oni2', 'Oni2_talk', 'XML', 'XML_talk')
25
26# URL for main namespace of our wiki
27onigalore_url = 'https://wiki.oni2.net/'
28
29# Tuple of interwiki prefixes, for recognizing and passing over such links
30interwiki_prefixes = ('acronym', 'cache', 'commons', 'dictionary', 'google', 'metawikimedia', 'mw', 'wikibooks', 'wikidata', 'wikimedia', 'wikinews', 'wikipedia', 'wikiquote', 'wikisource', 'wikispecies', 'wikiversity', 'wikivoyage', 'wikt', 'wiktionary', 'wp')
31
32# List of chapter names, for substitution into links that use "{{Cn}}" transclusion
33chapter_names = ['CHAPTER_00_._COMBAT_TRAINING', 'CHAPTER_01_._TRIAL_RUN', 'CHAPTER_02_._ENGINES_OF_EVIL', 'CHAPTER_03_._PUZZLE_PIECES', 'CHAPTER_04_._TIGER_BY_THE_TAIL', 'CHAPTER_05_._HOT_PURSUIT', 'CHAPTER_06_._COUNTERATTACK', 'CHAPTER_07_._A_FRIEND_IN_NEED', 'CHAPTER_08_._AN_INNOCENT_LIFE', 'CHAPTER_09_._TRUTH_AND_CONSEQUENCES', 'CHAPTER_10_._CAT_AND_MOUSE', 'CHAPTER_11_._DREAM_DIVER', 'CHAPTER_12_._SINS_OF_THE_FATHER', 'CHAPTER_13_._PHOENIX_RISING', 'CHAPTER_14_._DAWN_OF_THE_CHRYSALIS']
34
35# Tuple of patterns for recognizing wikilinks
36# Pattern 1: Detect "[[anything]]", "[[any:thing]]", "[[any|thing]]", "[[any:thi|ng]]"
37# Pattern 2: Detect "{{SectionLink|Page|Section name}}", "{{SectionLink||Section name}}"
38link_patterns = (r"\[\[[^|\]]*(\||\])", r"\{\{SectionLink\|[^|\}]*\|[^|\}]*\}\}")
39
40# Initialize globals
41debug = 0
42pages_checked = 0
43iw_found = 0
44advice_issued = 0
45errors_issued = 0
46name_printed = 0
47
48# Prints the name of a page on which something occurred, if it has not been printed before
49def possibly_print(page_name):
50 global debug
51 global name_printed
52
53 if not name_printed and not debug:
54 pywikibot.stdout('')
55 pywikibot.stdout('From page "{}":'.format(page_name))
56 name_printed = 1
57
58# Search a page for the section specified in the link
59def find_section(page_text, page_name, page_slug, print_result):
60 global errors_issued
61 found_section = False
62
63 # Isolate section link or text fragment link
64 target_page_name, anchor_name = page_slug.split('#', 1)
65 target_page_name_human = target_page_name.replace('_', ' ')
66
67 # First check if this is a text fragment directive, and look for it if so
68 if anchor_name.startswith(':~:text='):
69 if debug: pywikibot.stdout(' Found text fragment directive {} from URL {}.'.format(anchor_name, page_slug))
70 anchor_name = anchor_name[8:]
71 # We're only checking the first text directive, so strip add'l ones if present
72 addl_fragment = anchor_name.find('&text=')
73 if addl_fragment != -1:
74 anchor_name = anchor_name[:addl_fragment]
75 search_terms = anchor_name.split(',')
76 # Delete prefix and suffix terms because they aren't needed
77 if search_terms[0].endswith('-'):
78 search_terms.pop(0)
79 if search_terms[-1].startswith('-'):
80 search_terms.pop()
81 # Remake text directive with the terms separated by spaces as they should be in the page text
82 newSep = ' '
83 search_string = newSep.join(search_terms)
84 if debug: pywikibot.stdout(' Converted text fragment to string "{}".'.format(search_string))
85 if search_string in page_text:
86 found_section = True
87 if debug and not print_result: pywikibot.stdout(' Found text fragment!')
88
89 # If we're still here, it's a section link; read linked page to see if it really has this
90 # anchor link
91 if found_section == False:
92 if debug: pywikibot.stdout(' Searching for section link {} on page.'.format(anchor_name))
93 soup = BeautifulSoup(page_text, 'html.parser')
94 # Search for a span with this ID
95 for span_tag in soup.findAll('span'):
96 span_name = span_tag.get('id', None)
97 if span_name == anchor_name:
98 if debug and not print_result: pywikibot.stdout(' Found section in a span!')
99 found_section = True
100 break
101 if found_section == False:
102 # Search for a div with this ID
103 for span_tag in soup.findAll('div'):
104 span_name = span_tag.get('id', None)
105 if span_name == anchor_name:
106 if debug and not print_result: pywikibot.stdout(' Found section in a div!')
107 found_section = True
108 break
109 if found_section == False:
110 possibly_print(page_name)
111 pywikibot.stdout(' ERROR: Could not find section "{0}" on page {1}!'.format(anchor_name, target_page_name_human))
112 errors_issued += 1
113 elif debug and print_result:
114 pywikibot.stdout(' The section "{0}" was found on page "{1}".'.format(anchor_name, target_page_name_human))
115
116# For a link that redirected us to another page, extract the name of the target page from
117# the target page's source
118def find_canonical_link(page_text, page_name, page_slug):
119 # Extract link from this markup which contains name of redirected-to page:
120 # <link rel="canonical" href="https://en.wikipedia.org/wiki/Page_name"/>
121 # "wgPageName":"Namespace:Page_name",
122 canonical_name = page_text.split('"wgPageName":"')[-1]
123 tag_end = canonical_name.find('",')
124
125 if tag_end == -1:
126 pywikibot.stdout(' ERROR: The link "{}" is a redirect page, but this script could not isolate the target page name.'.format(page_slug))
127 errors_issued = errors_issued + 1
128 else:
129 canonical_name = canonical_name[:tag_end]
130 if len(canonical_name) > 100:
131 # Certain things can cause the trim to fail; report error and avoid slamming the
132 # output with massive page source from a failed trim
133 pywikibot.stdout(' ERROR: The link "{}" is a redirect to "{2}…" (string overflow).'.format(page_slug, canonical_name[:100]))
134 errors_issued = errors_issued + 1
135 else:
136 canonical_name = canonical_name.replace('_', ' ')
137 if '#' in page_slug:
138 _, anchor_name = page_slug.split('#')
139 if debug: pywikibot.stdout(' The link "{0}" is a redirect to "{1}#{2}", which is a valid page. Checking section link….'.format(page_slug, canonical_name, anchor_name))
140 find_section(page_text, page_name, page_slug, True)
141 else:
142 pywikibot.stdout(' The link "{0}" is a redirect to "{1}", which is a valid page.'.format(page_slug, canonical_name))
143
144# Test an intrawiki link and look for a section link if applicable
145def test_intrawiki_link(iw_url, page_name, page_slug):
146 global advice_issued
147 global errors_issued
148
149 response = fetch(iw_url)
150
151 # One way we tell that a redirect occurred is by checking fetch's history, as it
152 # automatically follows redirects. This will catch formal redirects which come from pages
153 # such as Special:PermanentLink.
154 if response.history != []:
155 permalink1 = 'Special:PermanentLink/'.lower()
156 permalink2 = 'Special:Permalink/'.lower()
157 page_slug_lower = page_slug.lower()
158 if page_slug_lower.startswith(permalink1) or page_slug_lower.startswith(permalink2):
159 if debug:
160 possibly_print(page_name)
161 pywikibot.stdout(' Got redirection code "{0}" for permanent revision link "{1}". Checking the target page….'.format(response.history[0], page_slug))
162 find_canonical_link(response.text, page_name, page_slug)
163 else:
164 possibly_print(page_name)
165 pywikibot.stdout(' ERROR: Unrecognized type of redirection (code "{0}") for link "{1}". You should check the link manually.'.format(response.history[0], page_slug))
166 advice_issued += 1
167 elif response.status_code != 200:
168 possibly_print(page_name)
169 pywikibot.stdout(' ERROR: Got response code {0} on URL {1}. The target page may not exist.'.format(response.status_code, iw_url))
170 errors_issued += 1
171 # However the usual way that a redirect occurs is that MediaWiki redirects us sneakily
172 # using JavaScript, while returning code OK 200 as if the link was correct; this happens
173 # when a redirect page is accessed. We must detect these soft redirects by looking at the
174 # page source to find the redirect note inserted at the top of the page for the reader.
175 elif 'Redirected from <a' in response.text:
176 if debug:
177 possibly_print(page_name)
178 pywikibot.stdout(' Got silently redirected by link "{}". Checking the target page….'.format(page_slug))
179 find_canonical_link(response.text, page_name, page_slug)
180 else: # URL is OK, so proceed
181 find_section(response.text, page_name, page_slug, False)
182
183# Searches the given page text for intrawiki links with section links in them
184def scan_for_intrawiki_links(page_text, page_name):
185 global debug
186 global pages_checked
187 global iw_found
188 global advice_issued
189 global errors_issued
190 global name_printed
191 pages_checked += 1
192 name_printed = 0
193
194 for i, the_pattern in enumerate(link_patterns):
195 if debug:
196 if i == 0:
197 pywikibot.stdout(' Checking page for wikilinks with section names.')
198 elif i == 1:
199 pywikibot.stdout(' Checking page for {{SectionLink}} calls.')
200
201 for match in re.finditer(the_pattern, page_text):
202 found_iw_match = False
203 iw_url = ""
204 page_name2 = page_name
205
206 # Cut out the matched text from the page, isolating just the page+section name
207 target_start = 2 # "[["
208 target_end = 1 # "|" or "]" (we only match the first ending bracket)
209 if i == 1:
210 target_start = 14 # "{{SectionLink|"
211 target_end = 2 # "}}"
212 s = match.start() + target_start # remove the link-opening markup
213 e = match.end() - target_end # remove the link-ending markup
214 page_slug = page_text[s:e]
215
216 # The second link type will look like "Page|Section" or "|Section", so fix that pipe
217 if i == 1:
218 page_slug = page_slug.replace('|', '#')
219
220 # Sometimes we use a space char. instead of a '_', so fix that before querying
221 page_slug = page_slug.replace(' ', '_')
222 if debug: pywikibot.stdout(' Found link {0}.'.format(page_slug))
223
224 # If this link doesn't have a section link in it, then we don't care about it, as
225 # MediaWiki takes care of checking basic intrawiki links
226 if not '#' in page_slug:
227 if debug: pywikibot.stdout(' Link doesn\'t have a section anchor in it. Skipping.')
228 continue
229
230 # If this link has an interwiki prefix, it can be ignored; see check_interwiki_links.py
231 # for the task of checking interwiki page+section links
232 is_interwiki = False
233 if found_iw_match == False:
234 for prefix in interwiki_prefixes:
235 if prefix + ":" in page_slug:
236 if debug: pywikibot.stdout(' Skipping link {} because it is an interwiki link.'.format(page_slug))
237 is_interwiki = True
238 break
239 if is_interwiki:
240 continue
241
242 # If there is a '{' in the link, then probably it's a link built on transcluded text.
243 # If it's a chapter template transclusion like "Quotes/Diary#{{C3}}", expand it using
244 # our "chapter_names" array. If it's another type of transclusion, punt it to the user.
245 if '{' in page_slug:
246 ch_link_pattern = re.compile(r"{{C[0-9]*}}")
247 ch_link = ch_link_pattern.search(page_slug)
248 if debug: pywikibot.stdout(' Found transclusion in link: "{}".'.format(ch_link.group(0)))
249 if ch_link:
250 ch_link_match = ch_link.group(0)
251 ch_num_pattern = re.compile("[0-9]+")
252 ch_num = ch_num_pattern.search(ch_link_match)
253 if ch_num:
254 ch_num_match = int(ch_num.group(0))
255 if ch_num_match >= 0 and ch_num_match <= 14:
256 ch_name = chapter_names[ch_num_match]
257 replace_pattern = re.compile(r"{{C" + ch_num.group(0) + r"}}")
258 page_slug = replace_pattern.sub(ch_name, page_slug)
259 if debug: pywikibot.stdout(' After performing transclusion, link is now "{}".'.format(page_slug))
260 else:
261 possibly_print(page_name)
262 pywikibot.stdout(' ERROR: Link {0} transcludes a chapter name using an out-of-range number, {1}.'.format(page_slug, ch_num_match))
263 errors_issued += 1
264 continue
265 else:
266 possibly_print(page_name)
267 pywikibot.stdout(' ADVICE: Link {} seems to be transcluding a chapter name, but this script couldn\'t read it.'.format(page_slug))
268 advice_issued += 1
269 continue
270 else:
271 possibly_print(page_name)
272 pywikibot.stdout(' ADVICE: Link {0} seems to use transclusion. This script can understand chapter name transclusions such as "{1}" but it doesn\'t recognize this one so it can\'t be verified. You should check the link manually.'.format(page_slug, "{{C7}}"))
273 advice_issued += 1
274 continue
275
276 # If this is a relative "/" link, use the current page as the basis for the URL. Note
277 # that only a leading slash is looked for, so if there's multiple steps down ("/x/y"),
278 # we're out of luck.
279 if page_slug.startswith('/'):
280 page_slug = page_name + page_slug
281 if debug: pywikibot.stdout(' Changed page_slug to {} on account of "/".'.format(page_slug))
282
283 # If this is a relative "../" link, find the parent page, set ourselves to that page,
284 # then remove the relative portion of the link. Note that this is only performed once,
285 # so if there's multiple steps back ("../../"), we're out of luck.
286 if page_slug.startswith('../'):
287 last_slash = page_name.rfind('/')
288 page_name2 = page_name[0:last_slash]
289 if debug: pywikibot.stdout(' Changed page_name to {} on account of "../".'.format(page_name2))
290 page_slug = page_slug[3:len(page_slug)]
291 if debug: pywikibot.stdout(' Changed page_slug to {} on account of "../".'.format(page_slug))
292 # If this is now going to be a bare section link for the parent page, don't add a
293 # slash, otherwise do because we are drilling down to another subpage
294 if page_slug.startswith('#'):
295 page_slug = page_name2 + page_slug
296 else:
297 page_slug = page_name2 + '/' + page_slug
298
299 # If this is a bare section link, build URL based on this page
300 if page_slug.startswith('#'):
301 iw_url = onigalore_url + page_name2
302 iw_found += 1
303 if debug: pywikibot.stdout(' Found link to this very page, {}.'.format(page_slug))
304 found_iw_match = True
305 page_slug = page_name2 + page_slug
306
307 # If there's no ":" in the link (before the section link, where a colon would just be
308 # part of the text) then it's a Main namespace article; proceed with building URL
309 if found_iw_match == False:
310 if not re.search(":.*#", page_slug):
311 iw_url = onigalore_url + page_slug
312 iw_found += 1
313 if debug: pywikibot.stdout(' Link is to a Main namespace page.')
314 found_iw_match = True
315
316 # If there is a ":", match the prefix against the intrawiki prefixes on OniGalore
317 # before building URL
318 if found_iw_match == False:
319 for prefix in intrawiki_prefixes:
320 if prefix + ":" in page_slug:
321 iw_url = onigalore_url + page_slug
322 if debug: pywikibot.stdout(' Identified namespace {}.'.format(prefix))
323 iw_found += 1
324 found_iw_match = True
325 break
326
327 # If we still haven't turned this match into a URL, something's gone wrong
328 if (found_iw_match == False) or (iw_url == ""):
329 possibly_print(page_name)
330 pywikibot.stdout(' ERROR: Couldn\'t figure out link {}.'.format(page_slug))
331 continue
332
333 # Test the URL
334 iw_url = iw_url.replace(' ', '_')
335 if debug: pywikibot.stdout(' Reading page at {}….'.format(iw_url))
336 test_intrawiki_link(iw_url, page_name, page_slug)
337
338# Print a wrap-up message
339def print_summary():
340 global pages_checked
341 global iw_found
342 global advice_issued
343 global errors_issued
344
345 page_str = "pages"
346 if pages_checked == 1:
347 page_str = "page"
348
349 link_str = "links"
350 if iw_found == 1:
351 link_str = "link"
352
353 pywikibot.stdout('Checked {0} {1} and found {2} intrawiki {3}.'.format(pages_checked, page_str, iw_found, link_str))
354 pywikibot.stdout('While attempting to follow section links….')
355
356 if advice_issued == 0:
357 pywikibot.stdout(' No advice on potential problems was issued.')
358 elif advice_issued == 1:
359 pywikibot.stdout(' 1 piece of advice on a potential problem was issued.')
360 else:
361 pywikibot.stdout(' {} pieces of advice on potential problems were issued.'.format(advice_issued))
362
363 error_str = "errors were"
364 if errors_issued == 1:
365 error_str = "error was"
366 pywikibot.stdout(' {0} {1} encountered.'.format(errors_issued, error_str))
367
368# Main function
369def main(*args):
370 global debug
371 search_cat = ''
372 search_page = ''
373
374 # Process arguments
375 local_args = pywikibot.handle_args(args)
376 for arg in local_args:
377 if arg.startswith('-cat:'):
378 search_cat = arg[5:]
379 elif arg.startswith('-page:'):
380 search_page = arg[6:]
381 elif arg == '-dbg':
382 debug = 1
383 else:
384 pywikibot.stdout('Unknown argument "{}".'.format(arg))
385 return
386
387 site = pywikibot.Site()
388
389 # This line of code enumerates the methods in the 'page' class
390 #pywikibot.stdout(format(dir(page)))
391
392 # Check specified page or loop through specified category and check all pages
393 if search_cat != '':
394 cat_obj = pywikibot.Category(site, search_cat)
395 generator = pagegenerators.CategorizedPageGenerator(cat_obj, recurse=True)
396 for page in pagegenerators.PreloadingGenerator(generator, 100):
397 if debug: pywikibot.stdout('Checking page {0}'.format(page.title()))
398 scan_for_intrawiki_links(page.text, page.title())
399 elif search_page != '':
400 page = pywikibot.Page(site, search_page)
401 if debug: pywikibot.stdout('Checking page {0}'.format(page.title()))
402 scan_for_intrawiki_links(page.text, page.title())
403
404 # Print the results
405 print_summary()
406
407if __name__ == '__main__':
408 main()
Note: See TracBrowser for help on using the repository browser.