source: ValBot/Python/check_intrawiki_section_links.py@ 1189

Last change on this file since 1189 was 1188, checked in by iritscen, 10 months ago

ValBot: Removed line converting slashes to HTML notation that now seems to cause a problem rather than solve one.
File size: 18.0 KB
RevLine 
[1171]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 ('#'),
[1179]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.
[1171]6# Recommended viewing width:
7# |---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- --|
8
[1169]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
[1179]23# Tuple of OniGalore's namespaces
[1169]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
[1185]29# Tuple of interwiki prefixes, for recognizing and passing over such links
[1169]30interwiki_prefixes = ('acronym', 'cache', 'commons', 'dictionary', 'google', 'metawikimedia', 'mw', 'wikibooks', 'wikidata', 'wikimedia', 'wikinews', 'wikipedia', 'wikiquote', 'wikisource', 'wikispecies', 'wikiversity', 'wikivoyage', 'wikt', 'wiktionary', 'wp')
31
[1179]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 = ("\[\[[^|\]]*(\||\])", "\{\{SectionLink\|[^|\}]*\|[^|\}]*\}\}")
39
40# Initialize globals
41debug = 0
[1169]42pages_checked = 0
43iw_found = 0
[1171]44advice_issued = 0
45errors_issued = 0
[1185]46name_printed = 0
[1169]47
[1185]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
62 # Isolate section link
63 target_page_name, anchor_name = page_slug.split('#', 1)
64 target_page_name_human = target_page_name.replace('_', ' ')
65 if debug: pywikibot.stdout(' Searching for section link {} on page.'.format(anchor_name))
66
67 # Read linked page to see if it really has this anchor link
68 soup = BeautifulSoup(page_text, 'html.parser')
69 found_section = False
70 for span_tag in soup.findAll('span'):
71 span_name = span_tag.get('id', None)
72 if span_name == anchor_name:
73 if debug and not print_result: pywikibot.stdout(' Found section in a span!')
74 found_section = True
75 break
76 if found_section == False:
77 # Search for a div with this ID
78 for span_tag in soup.findAll('div'):
79 span_name = span_tag.get('id', None)
80 if span_name == anchor_name:
81 if debug and not print_result: pywikibot.stdout(' Found section in a div!')
82 found_section = True
83 break
84 if found_section == False:
85 possibly_print(page_name)
[1186]86 pywikibot.stdout(' ERROR: Could not find section "{0}" on page {1}!'.format(anchor_name, target_page_name_human))
[1185]87 errors_issued += 1
88 elif debug and print_result:
89 pywikibot.stdout(' The section "{0}" was found on page "{1}".'.format(anchor_name, target_page_name_human))
90
91# For a link that redirected us to another page, extract the name of the target page from
92# the target page's source
93def find_canonical_link(page_text, page_name, page_slug):
94 # Extract link from this markup which contains name of redirected-to page:
95 # <link rel="canonical" href="https://en.wikipedia.org/wiki/Page_name"/>
96 # "wgPageName":"Namespace:Page_name",
97 canonical_name = page_text.split('"wgPageName":"')[-1]
98 tag_end = canonical_name.find('",')
99
100 if tag_end == -1:
101 pywikibot.stdout(' ERROR: The link "{}" is a redirect page, but this script could not isolate the target page name.'.format(page_slug))
102 errors_issued = errors_issued + 1
103 else:
104 canonical_name = canonical_name[:tag_end]
105 if len(canonical_name) > 100:
106 # Certain things can cause the trim to fail; report error and avoid slamming the
107 # output with massive page source from a failed trim
108 pywikibot.stdout(' ERROR: The link "{}" is a redirect to "{2}…" (string overflow).'.format(page_slug, canonical_name[:100]))
109 errors_issued = errors_issued + 1
110 else:
111 canonical_name = canonical_name.replace('_', ' ')
112 if '#' in page_slug:
113 _, anchor_name = page_slug.split('#')
114 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))
115 find_section(page_text, page_name, page_slug, True)
116 else:
117 pywikibot.stdout(' The link "{0}" is a redirect to "{1}", which is a valid page.'.format(page_slug, canonical_name))
118
119# Test an intrawiki link and look for a section link if applicable
120def test_intrawiki_link(iw_url, page_name, page_slug):
121 global advice_issued
122 global errors_issued
123
124 response = fetch(iw_url)
125
126 # One way we tell that a redirect occurred is by checking fetch's history, as it
127 # automatically follows redirects. This will catch formal redirects which come from pages
128 # such as Special:PermanentLink.
129 if response.history != []:
130
131 permalink1 = 'Special:PermanentLink/'.lower()
132 permalink2 = 'Special:Permalink/'.lower()
133 page_slug_lower = page_slug.lower()
134 if page_slug_lower.startswith(permalink1) or page_slug_lower.startswith(permalink2):
135 if debug:
136 possibly_print(page_name)
137 pywikibot.stdout(' Got redirection code "{0}" for permanent revision link "{1}". Checking the target page….'.format(response.history[0], page_slug))
138 find_canonical_link(response.text, page_name, page_slug)
139 else:
140 possibly_print(page_name)
141 pywikibot.stdout(' ERROR: Unrecognized type of redirection (code "{0}") for link "{1}". You should check the link manually.'.format(response.history[0], page_slug))
142 advice_issued += 1
143 elif response.status_code != 200:
144 possibly_print(page_name)
145 pywikibot.stdout(' ERROR: Got response code {0} on URL {1}. The target page may not exist.'.format(response.status_code, iw_url))
146 errors_issued += 1
147 # However the usual way that a redirect occurs is that MediaWiki redirects us sneakily
148 # using JavaScript, while returning code OK 200 as if the link was correct; this happens
149 # when a redirect page is accessed. We must detect these soft redirects by looking at the
150 # page source to find the redirect note inserted at the top of the page for the reader.
151 elif 'Redirected from <a' in response.text:
152 if debug:
153 possibly_print(page_name)
154 pywikibot.stdout(' Got silently redirected by link "{}". Checking the target page….'.format(page_slug))
155 find_canonical_link(response.text, page_name, page_slug)
156 else: # URL is OK, so proceed
157 find_section(response.text, page_name, page_slug, False)
158
[1169]159# Searches the given page text for intrawiki links with section links in them
[1179]160def scan_for_intrawiki_links(page_text, page_name):
161 global debug
162 global pages_checked
163 global iw_found
164 global advice_issued
165 global errors_issued
[1185]166 global name_printed
[1179]167 pages_checked += 1
168 name_printed = 0
[1169]169
[1179]170 for i, the_pattern in enumerate(link_patterns):
171 if debug:
172 if i == 0:
173 pywikibot.stdout(' Checking page for wikilinks with section names.')
174 elif i == 1:
175 pywikibot.stdout(' Checking page for {{SectionLink}} calls.')
176
177 for match in re.finditer(the_pattern, page_text):
178 found_iw_match = False
179 iw_url = ""
180 page_name2 = page_name
181
182 # Cut out the matched text from the page, isolating just the page+section name
183 target_start = 2 # "[["
184 target_end = 1 # "|" or "]" (we only match the first ending bracket)
185 if i == 1:
186 target_start = 14 # "{{SectionLink|"
187 target_end = 2 # "}}"
188 s = match.start() + target_start # remove the link-opening markup
189 e = match.end() - target_end # remove the link-ending markup
[1185]190 page_slug = page_text[s:e]
[1179]191
192 # The second link type will look like "Page|Section" or "|Section", so fix that pipe
193 if i == 1:
[1185]194 page_slug = page_slug.replace('|', '#')
[1169]195
[1179]196 # Sometimes we use a space char. instead of a '_', so fix that before querying
[1185]197 page_slug = page_slug.replace(' ', '_')
198 if debug: pywikibot.stdout(' Found link {0}.'.format(page_slug))
[1179]199
200 # If this link doesn't have a section link in it, then we don't care about it, as
201 # MediaWiki takes care of checking basic intrawiki links
[1185]202 if not '#' in page_slug:
[1179]203 if debug: pywikibot.stdout(' Link doesn\'t have a section anchor in it. Skipping.')
[1169]204 continue
[1176]205
[1179]206 # If this link has an interwiki prefix, it can be ignored; see check_interwiki_links.py
207 # for the task of checking interwiki page+section links
208 is_interwiki = False
209 if found_iw_match == False:
[1176]210 for prefix in interwiki_prefixes:
[1185]211 if prefix + ":" in page_slug:
212 if debug: pywikibot.stdout(' Skipping link {} because it is an interwiki link.'.format(page_slug))
[1179]213 is_interwiki = True
214 break
215 if is_interwiki:
[1176]216 continue
[1179]217
[1185]218 # If there is a '{' in the link, then probably it's a link built on transcluded text.
219 # If it's a chapter template transclusion like "Quotes/Diary#{{C3}}", expand it using
220 # our "chapter_names" array. If it's another type of transclusion, punt it to the user.
221 if '{' in page_slug:
[1179]222 ch_link_pattern = re.compile(r"{{C[0-9]*}}")
[1185]223 ch_link = ch_link_pattern.search(page_slug)
[1179]224 if debug: pywikibot.stdout(' Found transclusion in link: "{}".'.format(ch_link.group(0)))
225 if ch_link:
226 ch_link_match = ch_link.group(0)
227 ch_num_pattern = re.compile("[0-9]+")
228 ch_num = ch_num_pattern.search(ch_link_match)
229 if ch_num:
230 ch_num_match = int(ch_num.group(0))
231 if ch_num_match >= 0 and ch_num_match <= 14:
232 ch_name = chapter_names[ch_num_match]
233 replace_pattern = re.compile(r"{{C" + ch_num.group(0) + r"}}")
[1185]234 page_slug = replace_pattern.sub(ch_name, page_slug)
235 if debug: pywikibot.stdout(' After performing transclusion, link is now "{}".'.format(page_slug))
[1179]236 else:
[1185]237 possibly_print(page_name)
238 pywikibot.stdout(' ERROR: Link {0} transcludes a chapter name using an out-of-range number, {1}.'.format(page_slug, ch_num_match))
239 errors_issued += 1
[1179]240 continue
241 else:
[1185]242 possibly_print(page_name)
243 pywikibot.stdout(' ADVICE: Link {} seems to be transcluding a chapter name, but this script couldn\'t read it.'.format(page_slug))
[1179]244 advice_issued += 1
245 continue
246 else:
[1185]247 possibly_print(page_name)
248 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}}"))
[1179]249 advice_issued += 1
250 continue
[1171]251
[1179]252 # If this is a relative "/" link, use the current page as the basis for the URL. Note
253 # that only a leading slash is looked for, so if there's multiple steps down ("/x/y"),
254 # we're out of luck.
[1185]255 if page_slug.startswith('/'):
256 page_slug = page_name + page_slug
257 if debug: pywikibot.stdout(' Changed page_slug to {} on account of "/".'.format(page_slug))
[1179]258
259 # If this is a relative "../" link, find the parent page, set ourselves to that page,
260 # then remove the relative portion of the link. Note that this is only performed once,
261 # so if there's multiple steps back ("../../"), we're out of luck.
[1185]262 if page_slug.startswith('../'):
[1169]263 last_slash = page_name.rfind('/')
264 page_name2 = page_name[0:last_slash]
[1179]265 if debug: pywikibot.stdout(' Changed page_name to {} on account of "../".'.format(page_name2))
[1185]266 page_slug = page_slug[3:len(page_slug)]
267 if debug: pywikibot.stdout(' Changed page_slug to {} on account of "../".'.format(page_slug))
[1171]268 # If this is now going to be a bare section link for the parent page, don't add a
269 # slash, otherwise do because we are drilling down to another subpage
[1185]270 if page_slug.startswith('#'):
271 page_slug = page_name2 + page_slug
[1169]272 else:
[1185]273 page_slug = page_name2 + '/' + page_slug
[1179]274
275 # If this is a bare section link, build URL based on this page
[1185]276 if page_slug.startswith('#'):
[1169]277 iw_url = onigalore_url + page_name2
[1179]278 iw_found += 1
[1185]279 if debug: pywikibot.stdout(' Found link to this very page, {}.'.format(page_slug))
[1169]280 found_iw_match = True
[1185]281 page_slug = page_name2 + page_slug
[1179]282
283 # If there's no ":" in the link (before the section link, where a colon would just be
284 # part of the text) then it's a Main namespace article; proceed with building URL
285 if found_iw_match == False:
[1185]286 if not re.search(":.*#", page_slug):
287 iw_url = onigalore_url + page_slug
[1179]288 iw_found += 1
289 if debug: pywikibot.stdout(' Link is to a Main namespace page.')
290 found_iw_match = True
291
292 # If there is a ":", match the prefix against the intrawiki prefixes on OniGalore
293 # before building URL
294 if found_iw_match == False:
[1169]295 for prefix in intrawiki_prefixes:
[1185]296 if prefix + ":" in page_slug:
297 iw_url = onigalore_url + page_slug
[1179]298 if debug: pywikibot.stdout(' Identified namespace {}.'.format(prefix))
299 iw_found += 1
300 found_iw_match = True
301 break
302
303 # If we still haven't turned this match into a URL, something's gone wrong
304 if (found_iw_match == False) or (iw_url == ""):
[1185]305 possibly_print(page_name)
306 pywikibot.stdout(' ERROR: Couldn\'t figure out link {}.'.format(page_slug))
[1173]307 continue
[1169]308
[1179]309 # Test the URL
310 iw_url = iw_url.replace(' ', '_')
[1185]311 if debug: pywikibot.stdout(' Reading page at {}….'.format(iw_url))
312 test_intrawiki_link(iw_url, page_name, page_slug)
[1169]313
[1185]314# Print a wrap-up message
315def print_summary():
[1179]316 global pages_checked
317 global iw_found
318 global advice_issued
319 global errors_issued
[1185]320
321 page_str = "pages"
322 if pages_checked == 1:
323 page_str = "page"
324
325 link_str = "links"
326 if iw_found == 1:
327 link_str = "link"
328
329 pywikibot.stdout('Checked {0} {1} and found {2} intrawiki {3}.'.format(pages_checked, page_str, iw_found, link_str))
330 pywikibot.stdout('While attempting to follow section links….')
331
332 if advice_issued == 0:
333 pywikibot.stdout(' No advice on potential problems was issued.')
334 elif advice_issued == 1:
335 pywikibot.stdout(' 1 piece of advice on a potential problem was issued.')
336 else:
337 pywikibot.stdout(' {} pieces of advice on potential problems were issued.'.format(advice_issued))
338
339 error_str = "errors were"
340 if errors_issued == 1:
341 error_str = "error was"
342 pywikibot.stdout(' {0} {1} encountered.'.format(errors_issued, error_str))
343
344# Main function
345def main(*args):
346 global debug
[1179]347 search_cat = ''
348 search_page = ''
[1169]349
[1185]350 # Process arguments
[1179]351 local_args = pywikibot.handle_args(args)
352 for arg in local_args:
353 if arg.startswith('-cat:'):
354 search_cat = arg[5:]
355 elif arg.startswith('-page:'):
356 search_page = arg[6:]
357 elif arg == '-dbg':
358 debug = 1
359 else:
360 pywikibot.stdout('Unknown argument "{}".'.format(arg))
361 return
[1169]362
[1179]363 site = pywikibot.Site()
[1169]364
[1179]365 # This line of code enumerates the methods in the 'page' class
366 #pywikibot.stdout(format(dir(page)))
[1169]367
[1185]368 # Check specified page or loop through specified category and check all pages
[1179]369 if search_cat != '':
370 cat_obj = pywikibot.Category(site, search_cat)
371 generator = pagegenerators.CategorizedPageGenerator(cat_obj, recurse=True)
372 for page in pagegenerators.PreloadingGenerator(generator, 100):
373 if debug: pywikibot.stdout('Checking page {0}'.format(page.title()))
374 scan_for_intrawiki_links(page.text, page.title())
375 elif search_page != '':
376 page = pywikibot.Page(site, search_page)
377 if debug: pywikibot.stdout('Checking page {0}'.format(page.title()))
378 scan_for_intrawiki_links(page.text, page.title())
[1169]379
[1185]380 # Print the results
381 print_summary()
[1169]382
383if __name__ == '__main__':
[1179]384 main()
Note: See TracBrowser for help on using the repository browser.