#!/usr/bin/python
# -*- coding: utf-8 -*-
"""
This script can be used to update links transcluded using the {{PDFlink}} template.
Syntax: python pdfbot.py [-ref: TemplateName]
Command line options:
-file: Update pages listed in a text file.
-ref: Update pages transcluding from a given page.
-cat: Update pages from the given category.
-links: Update pages linked from a given page.
-page: Update that page.
"""
#
# (c) Dispenser, 2007
#
import re, sys, httplib, time
import wikipedia, pagegenerators, login, config, catlib
from urllib2 import urlparse
# Define global constants
readDelay = 20 # seconds
writeDelay = 60 # seconds
prefix = ['bytes', '[[Kibibyte|KiB]]', '[[Mebibyte|MiB]]', '[[Gibibyte|GiB]]']
urlpattern = re.compile(r'http[s]?://[^][>< \n|]*', re.IGNORECASE)
userAgent = 'PDFbot (http://en.wikipedia.org/wiki/User:PDFbot)'
# Edit summary messages
msg_added = {
'de': u'BOT: hinzufuegen der Dateigroesse markiert als {{PDFlink}}',
'en': u'Added file size for external links tagged with {{[[Template:PDFlink|PDFlink]]}}',
}
msg_updated = {
'de': u'BOT: Aktualisieren der Dateigroesse mit Vorlageeinbindung',
'en': u'Updated file size of transcluded {{[[Template:PDFlink|PDFlink]]}}',
}
msg_fixed = {
'de': u'BOT: Korrigierte Benutzung der Vorlage {{PDFlink}}',
'en': u'Corrected usage of {{[[Template:PDFlink|PDFlink]]}}',
}
msg_removed_cite = {
'de': u'BOT: Entfernen {{PDFlink}} entsprechend der zitierten Vorlage',
'en': u'Remove {{PDFlink}} from citation template.',
}
def checkLink(___location, redirectCounter = 5):
try:
while (redirectCounter > 0 and ___location is not None):
(scheme, site, path, args, query, frag) = urlparse.urlparse(___location)
if scheme == "http":
conn = httplib.HTTPConnection(site)
elif scheme == "https":
conn = httplib.HTTPSConnection(site)
#conn.set_debuglevel(1)
conn.putrequest('HEAD', path + args + query)
conn.putheader('User-Agent', userAgent)
conn.endheaders()
response = conn.getresponse()
redirect = response.msg.getheader('___location')
content_length = response.msg.getheader('content-length')
content_type = response.msg.getheader('content-type')
response_code = response.status
conn.close()
redirectCounter -= 1
if(redirect is not None):
if(redirect[:4] != "http"):
___location = urlparse.urljoin(___location, redirect)
else:
___location = redirect
wikipedia.output( u'STATUS: HTTP %s Moved: %s' % (response_code, ___location) )
else:
___location = None
return [___location, response_code, content_length, content_type]
except httplib.error, arg:
wikipedia.output(u'HTTP Error: %s %s' % (arg, ___location))
return [___location, 7, None, None]
except:
wikipedia.output(u'Error with URL: %s' % ___location)
return [___location, 6, None, None]
# Convert the byte count to a human readable value
def binary_notation(size):
a = float(size)
exponent = 0
while(a >= 1000.):
a /= 1024.
exponent += 3
# Truncate and remove trailing dot
byteSigs = str(a)[:4]
if (byteSigs.endswith('.')):
byteSigs = byteSigs[:3]
return byteSigs + ' ' + prefix[exponent / 3]
def fix_broken_links(hypertext):
#This function attempts to fix multipule broken link using its dictionary
# Moving of resources
hypertext = re.sub(r'virginiadot.org/infoservice/resources/', r'virginiadot.org/info/resources/', hypertext)
hypertext = re.sub(r'ncdot.org/transit/aviation/ncairports/locations/pdf/', r'ncdot.org/transit/aviation/download/ncairports/', hypertext)
# 301 Permanent Redirects
hypertext = re.sub(r'transportation.ky.gov/planning/', r'www.planning.kytc.ky.gov/', hypertext)
hypertext = re.sub(r'official-documents.co.uk/', r'official-documents.gov.uk/', hypertext)
return hypertext
def update_size_paramter(template_text):
# following char sperate url from title: []"<>\ \n
# | is included since we're in a template
fixed_text = fix_broken_links(template_text)
___location = urlpattern.search(fixed_text).group(0)
prefix_text = re.search(r'(\{\{[^|]*\|[^|]*).*\}\}', fixed_text).group(1)
if (re.findall(r'=', template_text)):
parameter_prefix = '|2='
else:
parameter_prefix = '|'
# Parse indirect HTML character references
___location = re.sub(r'&\#61;', r'=', ___location)
___location = re.sub(r'&', r'&', ___location)
(redirect, response, content_length, content_type) = checkLink(___location)
if (content_type is not None and content_length is not None and int(content_length) > 16):
# I should really put in 404 error handling code, but this has been working just fine.
if re.findall(r'pdf|octet-stream', content_type):
return prefix_text + parameter_prefix + binary_notation(content_length) + '<!-- ' + content_type + ', ' + str(content_length) + ' bytes -->}}'
else:
wikipedia.output(u'Unusual content_type: %s, code: %s' % (content_type, response))
return template_text
# If anything else return template_text back
def process_article(page):
wikitext = page.get()
# Fix Casing (Reduces the number of possible expressions)
wikitext = re.compile(r'\{\{\s*(template:|)pdf', re.IGNORECASE).sub(r'{{PDF', wikitext)
# State point. Count any changes as needing an update if they're after this line
state0 = wikitext
# Convert hard coded pdf links (ex: [http link] (pdf) )
wikitext = re.sub(r'(\[http[^]]*\]) *\((\[\[[^|\]]*|)(PDF|pdf)(\]\]|)\)', r'{{PDFlink|\1}}', wikitext)
# Convert from the old style to the new style (ex: [http link] {{PDF}} or {{PDF}} [http link] )
wikitext = re.sub(r'(\(|)\{\{(PDFlink|PDF)\}\}(\)|) *(\[http[^]]*\])', r'{{\2|\4}}', wikitext)
wikitext = re.sub(r'(("|)\[http[^]]*\]("|))([^a-zA-Z(]*) *(\(|)\{\{(PDFlink|PDF)\}\}(\)|)', r'{{\6|\1}}\4', wikitext)
# Experimental: Convert with with tag at the end of a bullet list (ex: * [http link] some text ([[PDF]]) )
wikitext = re.compile(r'(\n\*[^\n:/]*)(\[http[^]]*\])([^\n:/]*) *\((\[\[|\{\{|)(Portable Document Format\|PDF|PDFlink).?(file|format|datei|)(\}\}|\]\]|)\)', re.IGNORECASE).sub(r'\1{{PDF|\2}}\3', wikitext)
wikitext = re.sub(r'(\n\*[^\n:/]*)(\[http[^]]*\])([^\n:/]*) *\{\{(PDFlink|PDF)\}\}', r'\1{{\4|\2}}\3', wikitext)
# Remove PDFlink from citation templates
wikitext = re.sub(r'(format *= *)(PDF|pdf|)(\(|)\{\{PDF[^{}]*\}\}(\)|)', r'\1PDF', wikitext)
wikitext = re.sub(r'(\{\{(Cite|cite)[^}]*)(}}[^a-zA-Z]*)(\(|)\{\{(PDF|PDFlink)\}\}(\)|)', r'\1 |format=PDF\3', wikitext)
wikitext = re.sub(r'(\{\{.ite web[^}]*) *(\(|)\{\{(PDF|PDFlink)\}\}(\)|) *([^}]*\}\})', r'\1 |format=PDF \5', wikitext)
state1 = wikitext
m = re.findall(r'\{\{PDF[link]{0,4}\|[^}]*\}\}', wikitext)
for s in m:
if re.findall(r'http[s]?://', s):
replacetext = update_size_paramter(s)
wikitext = re.sub(re.escape(s), replacetext, wikitext)
# Uncomment the bellow line to see the replacement text
# wikipedia.output(replacetext)
# Fix equal sign problem
wikitext = re.sub(r'\{\{(PDF|PDFlink)\|(1=|)(.{2}[^{|}]+=[^{|}]+)', r'{{\1|1=\3', wikitext)
# Test to see if file sizes parameter was untouched
if wikitext == state1:
if len(wikitext) - len(state1) <= 4:
# 4 or more bytes removed typically indicate a embed citation removal
EditMsg = msg_removed_cite
else:
EditMsg = msg_fixed
else:
if len(wikitext) - len(state1) > 34:
# Minimum of 34 bytes to add file size information
EditMsg = msg_added
else:
EditMsg = msg_updated
wikipedia.setAction(wikipedia.translate(page.site().language(), EditMsg))
# If the text has changed at all since the state point, upload it
if (wikitext != state0):
try:
wikipedia.output(u'Page change by %s bytes. Writing new version.' % str(len(wikitext)-len(state0)))
page.put(wikitext)
except:
wikipedia.output(u'------- Write error ------')
# Pause to reduce load on the servers
time.sleep(writeDelay)
else:
time.sleep(readDelay)
def main():
site = wikipedia.getSite()
gen = None
for arg in wikipedia.handleArgs():
if arg.startswith('-ref:'):
referredPage = wikipedia.Page(site, arg[5:])
gen = pagegenerators.ReferringPageGenerator(referredPage)
elif arg.startswith('-file:'):
gen = pagegenerators.TextfilePageGenerator(arg[6:])
elif arg.startswith('-cat:'):
cat = catlib.Category(site, arg[5:])
gen = pagegenerators.CategorizedPageGenerator(cat)
elif arg.startswith('-links:'):
pagelinks = wikipedia.Page(wikipedia.getSite(), arg[7:])
gen = pagegenerators.LinkedPageGenerator(pagelinks)
elif arg.startswith('-page:'):
page = wikipedia.Page(wikipedia.getSite(), unicode(arg[6:]))
gen = iter([page])
if gen is None:
wikipedia.showHelp(u'pdfbot')
return
wikipedia.output(u'Read delay is %s seconds.' % readDelay)
wikipedia.output(u'Write delay is %s seconds.\n' % writeDelay)
# Only process pages from the main namespace
gen = pagegenerators.NamespaceFilterPageGenerator(gen, [0])
gen = pagegenerators.RedirectFilterPageGenerator(gen)
for page in gen:
process_article(page)
wikipedia.output(u'\nOperation Complete.\n')
if __name__ == "__main__":
try:
main()
finally:
wikipedia.stopme()
