ViewVC Help

View File | Revision Log | View Changeset | Root Listing

root/Oni2/ValBot/Python/check_interwiki_links.py

Comparing ValBot/Python/check_interwiki_links.py (file contents):
Revision 1180 by iritscen, Fri Apr 28 00:54:21 2023 UTC vs.
Revision 1197 by iritscen, Fri Aug 29 03:52:17 2025 UTC

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

Diff Legend

–	Removed lines
+	Added lines
<	Changed lines (old)
>	Changed lines (new)