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 1192 by iritscen, Mon Sep 16 23:08:26 2024 UTC

#	Line 12 | Line 12 | import os
12		from urllib.parse import urljoin
13
14		import pywikibot
15	+	import bs4
16		import re
17		import requests # for listing members with dir()
18
#	Line 32 | Line 33 | debug = 0
33		pages_checked = 0
34		iw_found = 0
35		errors_issued = 0
36	+	name_printed = 0
37	+
38	+	# Prints the name of a page on which something occurred, if it has not been printed before
39	+	def 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
49	+	def 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
103	+	def 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
131	+	def 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
171		def scan_for_interwiki_links(page_text, page_name):
172		global debug
173		global pages_checked
174		global iw_found
175	<	global errors_issued
175	>	global name_printed
176		pages_checked = pages_checked + 1
177	<	cur = 0
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 = "\[\[" + prefix + ":[^|\]]*(\||\])"
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	<	# Sometimes we used a space char. instead of a '_', so fix that before querying
189	<	page_title = page_text[s:e].replace(' ', '_')
188	>	# Commonly we use spaces instead of underscores, so fix that before querying
189	>	page_slug = page_text[s:e].replace(' ', '_')
190
191	<	# Use only spaces for title when printing it
192	<	page_title_human = page_title.replace('_', ' ')
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] + page_title
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_title):
201	<	lang_code = page_title[0:2] + "."
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_title[0:3], '')
205	>	iw_url = iw_url.replace(page_slug[0:3], '')
206
207		# Test the URL
208	<	response = fetch(iw_url)
208	>	test_interwiki_link(prefix, interwiki_urls[cur_prefix], iw_url, page_name, page_slug)
209	>	cur_prefix = cur_prefix + 1
210
211	<	# One way we tell that a redirect occurred is by checking the history
212	<	if response.history != []:
213	<	if not name_printed and not debug:
214	<	pywikibot.stdout('From page "{}":'.format(page_name))
215	<	name_printed = 1
216	<	if page_title.startswith('WP:') and page_title == page_title.upper():
217	<	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))
218	<	else:
219	<	pywikibot.stdout('   ERROR: Got redirection code ({0}) for {1} link "{2}". You should check the link manually.'.format(response.history[0], prefix, page_title))
220	<	errors_issued = errors_issued + 1
221	<	elif response.status_code != 200:
222	<	if not name_printed and not debug:
223	<	pywikibot.stdout('From page "{}":'.format(page_name))
224	<	name_printed = 1
225	<	pywikibot.stdout('   ERROR: Got response code {0} for {1} link "{2}". The page may not exist.'.format(response.status_code, prefix, page_title))
226	<	errors_issued = errors_issued + 1
227	<	# The usual way that a redirect occurs is that MediaWiki redirects us sneakily
228	<	# using JavaScript, while returning code OK 200 as if the link was correct; we
229	<	# must detect this from the page source
230	<	elif 'Redirected from <a' in response.text:
231	<	if not name_printed and not debug:
97	<	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
211	>	# Print a wrap-up message
212	>	def 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
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)