1 | #! /usr/bin/env python |
---|
2 | |
---|
3 | ## Copyright (c) 1999 - 2003 L. C. Rees. All rights reserved. |
---|
4 | ## See COPYRIGHT file for license terms. |
---|
5 | |
---|
6 | from __future__ import generators |
---|
7 | |
---|
8 | |
---|
9 | class Spider: |
---|
10 | |
---|
11 | '''HTTP and FTP crawling, reporting, and checking''' |
---|
12 | |
---|
13 | import os as _os |
---|
14 | import urllib as _ulib |
---|
15 | import urlparse as _uparse |
---|
16 | from os import path as _path |
---|
17 | from ftplib import FTP as _ftp |
---|
18 | from time import strftime as _formtime |
---|
19 | from time import localtime as _localtime |
---|
20 | from ftplib import error_perm as _ftperr |
---|
21 | from sgmllib import SGMLParseError as _sperror |
---|
22 | from robotparser import RobotFileParser as _rparser |
---|
23 | # Use threads if available |
---|
24 | try: from threading import Thread as _thread |
---|
25 | except ImportError: pass |
---|
26 | _bdsig, _bfsig, _session, _newparser = None, None, None, None |
---|
27 | # HTML tags with URLs |
---|
28 | _urltags = {'a':1, 'img':1, 'link':1, 'script':1, 'iframe':1, 'object':1, |
---|
29 | 'embed':1, 'area':1, 'frame':1, 'applet':1, 'input':1, 'base':1, |
---|
30 | 'div':1, 'layer':1, 'ilayer':1, 'bgsound':1} |
---|
31 | # Supported protocols |
---|
32 | _supported = {'HTTP':1, 'http':1, 'HTTPS':1, 'https':1, 'FTP':1, 'ftp':1} |
---|
33 | # HTML attributes with URLs |
---|
34 | _urlattrs = {'href':1, 'src':1, 'data':1} |
---|
35 | |
---|
36 | def __init__(self, base=None, width=None, depth=None): |
---|
37 | '''Initializes a Spider instance and its base attributes |
---|
38 | |
---|
39 | Arguments: |
---|
40 | base -- URL to crawl (default: None) |
---|
41 | width -- maximum resources to crawl (default: None) |
---|
42 | depth -- how deep in a hierarchy to crawl (default: None)''' |
---|
43 | if base: self.base = base |
---|
44 | else: self.base = None |
---|
45 | if width: self.width = width |
---|
46 | else: self.width = None |
---|
47 | if depth: self.depth = depth |
---|
48 | else: self.depth = None |
---|
49 | |
---|
50 | def _ftpopen(self, base, name='anonymous', password=None, attempts=3): |
---|
51 | '''Returns FTP client session |
---|
52 | |
---|
53 | Arguments: |
---|
54 | base -- FTP server URL |
---|
55 | name -- login name (default: 'anonymous') |
---|
56 | password -- login password (default: None) |
---|
57 | attempts -- number of login attempts to try (default: 3)''' |
---|
58 | |
---|
59 | def ftpprompt(tries=0): |
---|
60 | '''Prompts for FTP username and password |
---|
61 | |
---|
62 | Arguments: |
---|
63 | tries -- number of login attempts''' |
---|
64 | tries += tries |
---|
65 | try: |
---|
66 | self._name = raw_input('Enter login name: ') |
---|
67 | self._password = raw_input('Enter password: ') |
---|
68 | session = ftp(base, self._name, self._password) |
---|
69 | return session |
---|
70 | # If login attempt fails, retry login |
---|
71 | except ftperr: |
---|
72 | if attempts >= tries: |
---|
73 | session = ftpprompt(tries) |
---|
74 | return session |
---|
75 | # Too many login attempts? End program |
---|
76 | elif attempts <= tries: |
---|
77 | raise IOError, 'Permission denied.' |
---|
78 | import sys |
---|
79 | sys.exit(0) |
---|
80 | |
---|
81 | # Assignments |
---|
82 | self._name, self._password, ftperr = name, password, self._ftperr |
---|
83 | su, ftp = self._uparse.urlsplit(base), self._ftp |
---|
84 | # Set URL, path, and strip 'ftp://' off |
---|
85 | base, path = su[1], '/'.join([su[2], '']) |
---|
86 | try: session = ftp(base, name, password) |
---|
87 | # Prompt for username, password if initial arguments are incorrect |
---|
88 | except ftperr: session = ftpprompt() |
---|
89 | # Change to remote path if it exits |
---|
90 | if path: session.cwd(path) |
---|
91 | return session |
---|
92 | |
---|
93 | def ftpmirror(self, l, t=None, b=None, w=200, d=6, n='anonymous', p=None): |
---|
94 | '''Mirrors an FTP site on a local filesystem |
---|
95 | |
---|
96 | Arguments: |
---|
97 | l -- local filesystem path (default: None) |
---|
98 | b -- FTP server URL (default: None) |
---|
99 | t -- number of download threads (default: None) |
---|
100 | w -- maximum amount of resources to crawl (default: 200) |
---|
101 | d -- depth in hierarchy to crawl (default: 6) |
---|
102 | n -- login username (default: 'anonymous') |
---|
103 | p -- login password (default: None)''' |
---|
104 | if b: self.ftpspider(b, w, d, n, p) |
---|
105 | return self._mirror((self.paths, self.urls), l, t) |
---|
106 | |
---|
107 | def ftppaths(self, b=None, w=200, d=6, n='anonymous', p=None): |
---|
108 | '''Returns a list of FTP paths. |
---|
109 | |
---|
110 | Arguments: |
---|
111 | b -- FTP server URL (default: None) |
---|
112 | w -- maximum amount of resources to crawl (default: 200) |
---|
113 | d -- depth in hierarchy to crawl (default: 6) |
---|
114 | n -- login username (default: 'anonymous') |
---|
115 | p -- login password (default: None)''' |
---|
116 | |
---|
117 | def sortftp(rdir): |
---|
118 | '''Returns a list of entries marked as files or directories |
---|
119 | |
---|
120 | Arguments: |
---|
121 | rdir -- remote directory list''' |
---|
122 | rlist = [] |
---|
123 | rappend = rlist.append |
---|
124 | for rl in rdir: |
---|
125 | # Split remote file based on whitespace |
---|
126 | ri = rl.split()[-1] |
---|
127 | # Add tuple of remote item type, permissions & name to rlist |
---|
128 | if ri not in ('.', '..'): rappend((rl[0], rl[7], ri)) |
---|
129 | return rlist |
---|
130 | |
---|
131 | def visitftp(): |
---|
132 | '''Extracts contents of an FTP directory''' |
---|
133 | wd = pwd() |
---|
134 | if wd[-1] != '/': wd = '/'.join([wd, '']) |
---|
135 | # Add present working directory to visited directories |
---|
136 | dirs[wd], rlist = None, [] |
---|
137 | # Get list of current directory's contents |
---|
138 | retr('LIST -a', rlist.append) |
---|
139 | for url in sortftp(rlist): |
---|
140 | # Test if remote item is a file (indicated by '-') |
---|
141 | if url[0] == '-': |
---|
142 | # Resolve path of file |
---|
143 | purl = ''.join([wd, url[2]]) |
---|
144 | # Ensure file list don't exceed max number of resources |
---|
145 | if len(files) >= width: return None |
---|
146 | # Add files to file dictionary |
---|
147 | elif purl not in files: files[purl] = None |
---|
148 | # Test if it's a directory ('d') and allows scanning ('-') |
---|
149 | elif url[0] == 'd': |
---|
150 | if url[1] != '-': |
---|
151 | # Resolve path of directory |
---|
152 | purl = ''.join([wd, url[2], '/']) |
---|
153 | # Ensure no recursion beyond depth allowed |
---|
154 | if len(purl.split('/')) >= depth: dirs[purl] = None |
---|
155 | # Visit directory if it hasn't been visited yet |
---|
156 | elif purl not in dirs: |
---|
157 | # Change to new directory |
---|
158 | cwd(purl) |
---|
159 | # Run 'visitftp' on new directory |
---|
160 | visitftp() |
---|
161 | |
---|
162 | # Use classwide attributes if set |
---|
163 | if b: self.base = b |
---|
164 | else: b = self.base |
---|
165 | # Use classwide width if different from method default |
---|
166 | if self.width and w == 200: width = self.width |
---|
167 | else: width = w |
---|
168 | # Use classwide depth if different from method default |
---|
169 | if self.depth and d == 6: depth = self.depth + 1 |
---|
170 | else: depth = d + 1 |
---|
171 | # File and directory dicts |
---|
172 | files, dirs = {}, {} |
---|
173 | # Use existing FTP client session if present |
---|
174 | if self._session: ftp = self._session |
---|
175 | # Create new FTP client session if necessary |
---|
176 | else: |
---|
177 | ftp = self._ftpopen(b, n, p) |
---|
178 | self._session = ftp |
---|
179 | # Avoid outside namespace lookups |
---|
180 | cwd, pwd, retr = ftp.cwd, ftp.pwd, ftp.retrlines |
---|
181 | # Walk FTP site |
---|
182 | visitftp() |
---|
183 | # Make path list out of files' keys and return it |
---|
184 | self.paths = files.keys() |
---|
185 | self.paths.sort() |
---|
186 | return self.paths |
---|
187 | |
---|
188 | def ftpspider(self, b=None, w=200, d=6, n='anonymous', p=None): |
---|
189 | '''Returns lists of URLs and paths plus a live FTP client session |
---|
190 | |
---|
191 | Arguments: |
---|
192 | b -- FTP server URL (default: None) |
---|
193 | w -- maximum amount of resources to crawl (default: 200) |
---|
194 | d -- depth in hierarchy to crawl (default: 6) |
---|
195 | n -- login username (default: 'anonymous') |
---|
196 | p -- login password (default: None)''' |
---|
197 | if b: ftppaths(b, w, d, n, p) |
---|
198 | return self.paths, ftpurls(), self._session |
---|
199 | |
---|
200 | def ftpurls(self, b=None, w=200, d=6, n='anonymous', p=None): |
---|
201 | '''Returns a list of FTP URLs |
---|
202 | |
---|
203 | Arguments: |
---|
204 | b -- FTP server URL (default: None) |
---|
205 | w -- maximum amount of resources to crawl (default: 200) |
---|
206 | d -- depth in hierarchy to crawl (default: 6) |
---|
207 | n -- login username (default: 'anonymous') |
---|
208 | p -- login password (default: None)''' |
---|
209 | if b: |
---|
210 | ftppaths(b, w, d, n, p) |
---|
211 | # Get rid of trailing '/' in base if present before joining |
---|
212 | if b[-1] == '/': base = b[:-1] |
---|
213 | else: |
---|
214 | base = self.base |
---|
215 | # Get rid of trailing '/' in base if present before joining |
---|
216 | if base[-1] == '/': base = self.base[:-1] |
---|
217 | paths = self.paths |
---|
218 | # Add FTP URL |
---|
219 | self.urls = [''.join([base, i]) for i in paths] |
---|
220 | return self.urls |
---|
221 | |
---|
222 | def _parserpick(self, old=None): |
---|
223 | '''Returns a class using the sgmllib parser or the sgmlop parser |
---|
224 | |
---|
225 | Arguments: |
---|
226 | old -- use classic sgmllib SGMLParser''' |
---|
227 | # Assignments |
---|
228 | urltags, urlattrs = self._urltags, self._urlattrs |
---|
229 | # Lists for bad file and bad directory signatures |
---|
230 | self._bfsig, self._bdsig = [], [] |
---|
231 | bfsig, bdsig = self._bfsig, self._bdsig |
---|
232 | # Use faster SGMLParser if available |
---|
233 | try: |
---|
234 | from sgmlop import SGMLParser as newparser |
---|
235 | self._newparser = newparser |
---|
236 | # If unavailable, use classic SGML parser |
---|
237 | except ImportError: |
---|
238 | from sgmllib import SGMLParser as oldparser |
---|
239 | old = 1 |
---|
240 | # Classes using classic sgmllib SGML Parser |
---|
241 | if old: |
---|
242 | from sgmllib import SGMLParser as oldparser |
---|
243 | # Remove sgmlop parser if present |
---|
244 | self._newparser = None |
---|
245 | # UrlExtract class using classic parser |
---|
246 | class UrlExtract(oldparser): |
---|
247 | '''Extracts URLs from a SGMLish document''' |
---|
248 | def reset(self): |
---|
249 | '''Resets SGML parser and clears lists''' |
---|
250 | oldparser.reset(self) |
---|
251 | self.urls, self.text, self.badurl = [], [], None |
---|
252 | def handle_data(self, data): |
---|
253 | '''Handles non-markup data''' |
---|
254 | # Get first 5 lines of non-markup data |
---|
255 | if len(self.text) <= 5: self.text.append(data) |
---|
256 | # Compare signature of known bad URL to a new web page |
---|
257 | if self.text == bfsig: self.badurl = 1 |
---|
258 | elif self.text == bdsig: self.badurl = 1 |
---|
259 | def finish_starttag(self, tag, attrs): |
---|
260 | '''Extracts URL bearing tags''' |
---|
261 | if tag in urltags: |
---|
262 | # Get key, vale in attributes if they match |
---|
263 | url = [v for k, v in attrs if k in urlattrs] |
---|
264 | if url: self.urls.extend(url) |
---|
265 | # BadUrl class using classic parser |
---|
266 | class BadUrl(oldparser): |
---|
267 | '''Collects results of intentionally incorrect URLs''' |
---|
268 | def reset(self): |
---|
269 | '''Resets SGML parser and clears lists''' |
---|
270 | oldparser.reset(self) |
---|
271 | self.text = [] |
---|
272 | def handle_data(self, data): |
---|
273 | '''Collects lines to profile bad URLs''' |
---|
274 | # Adds first 5 lines of non-markup data to text |
---|
275 | if len(self.text) <= 5: self.text.append(data) |
---|
276 | # If no old flag, use SGMLParser from sgmlop and related classes |
---|
277 | else: |
---|
278 | # UrlExtract class using sgmlop parser |
---|
279 | class UrlExtract: |
---|
280 | '''Extracts URLs from a SGMLish document''' |
---|
281 | def __init__(self): |
---|
282 | '''Resets SGML parser and clears lists''' |
---|
283 | self.urls, self.text, self.badurl = [], [], None |
---|
284 | def handle_data(self, data): |
---|
285 | '''Handles non-markup data''' |
---|
286 | # Get first 5 lines of non-markup data |
---|
287 | if len(self.text) <= 5: self.text.append(data) |
---|
288 | # Compare signature of known bad URL to a new web page |
---|
289 | if self.text == bfsig: self.badurl = 1 |
---|
290 | elif self.text == bdsig: self.badurl = 1 |
---|
291 | def finish_starttag(self, tag, attrs): |
---|
292 | '''Extracts URL bearing tags''' |
---|
293 | if tag in urltags: |
---|
294 | # Get key, vale in attributes if they match |
---|
295 | url = [v for k, v in attrs if k in urlattrs] |
---|
296 | if url: self.urls.extend(url) |
---|
297 | # BadUrl class using sgmlop parser |
---|
298 | class BadUrl: |
---|
299 | '''Collects results of intentionally incorrect URLs''' |
---|
300 | def __init__(self): |
---|
301 | '''Resets SGML parser and clears lists''' |
---|
302 | self.text = [] |
---|
303 | def handle_data(self, data): |
---|
304 | '''Collects lines to profile not found responses''' |
---|
305 | # Adds first 5 lines of non-markup data to list 'text' |
---|
306 | if len(self.text) <= 5: self.text.append(data) |
---|
307 | # Make resulting classes available class wide |
---|
308 | self._UrlExtract, self._BadUrl = UrlExtract, BadUrl |
---|
309 | |
---|
310 | def _webtest(self): |
---|
311 | '''Generates signatures for identifying bad URLs''' |
---|
312 | |
---|
313 | def badurl(url): |
---|
314 | '''Returns first 5 lines of a bad URL |
---|
315 | |
---|
316 | Arguments: |
---|
317 | url -- Bad URL to open and parse''' |
---|
318 | # Use different classes if faster SGML Parser is available |
---|
319 | if self._newparser: |
---|
320 | # sgmlop parser must have a handler passed to it |
---|
321 | parser, urlget = self._newparser(), BadUrl() |
---|
322 | # Pass handler (sgmlop cannot be subclassed) |
---|
323 | parser.register(urlget) |
---|
324 | parser.feed(urlopen(url).read()) |
---|
325 | parser.close() |
---|
326 | # Use classic parser |
---|
327 | else: |
---|
328 | urlget = BadUrl() |
---|
329 | urlget.feed(urlopen(url).read()) |
---|
330 | urlget.close() |
---|
331 | # Return singature of bad URL |
---|
332 | return urlget.text |
---|
333 | |
---|
334 | # Make globals local |
---|
335 | base, urljoin = self.base, self._uparse.urljoin |
---|
336 | urlopen, BadUrl = self._ulib.urlopen, self._BadUrl |
---|
337 | # Generate random string of jibber |
---|
338 | from string import letters, digits |
---|
339 | from random import choice, randint |
---|
340 | jibber = ''.join([letters, digits]) |
---|
341 | ru = ''.join([choice(jibber) for x in range(randint(1, 30))]) |
---|
342 | # Builds signature of a bad URL for a file |
---|
343 | self._bfsig.extend(badurl(urljoin(base, '%s.html' % ru))) |
---|
344 | # Builds signature of a bad URL for a directory |
---|
345 | self._bdsig.extend(badurl(urljoin(base,'%s/' % ru))) |
---|
346 | |
---|
347 | def _webparser(self, html): |
---|
348 | '''Parses HTML and returns bad URL indicator and extracted URLs |
---|
349 | |
---|
350 | Arguments: |
---|
351 | html -- HTML data''' |
---|
352 | # Use different classes if faster SGML Parser is available |
---|
353 | if self._newparser: |
---|
354 | # Make instances of SGML parser and URL extracting handler |
---|
355 | parser, urlget = self._newparser(), self._UrlExtract() |
---|
356 | # Pass handler to parser |
---|
357 | parser.register(urlget) |
---|
358 | # Feed data to parser |
---|
359 | parser.feed(html) |
---|
360 | parser.close() |
---|
361 | # Return bad URL indicator and extracted URLs |
---|
362 | else: |
---|
363 | urlget = self._UrlExtract() |
---|
364 | urlget.feed(html) |
---|
365 | urlget.close() |
---|
366 | # Return badurl marker and list of child URLS |
---|
367 | return urlget.badurl, urlget.urls |
---|
368 | |
---|
369 | def _webopen(self, base): |
---|
370 | '''Verifies URL and returns actual URL and extracted child URLs |
---|
371 | |
---|
372 | Arguments: |
---|
373 | base -- tuple containing a URL and its referring URL''' |
---|
374 | # Assignments |
---|
375 | good, cbase = self._good, base[0] |
---|
376 | try: |
---|
377 | # If webspiders can access URL, open it |
---|
378 | if self._robot.can_fetch('*', cbase): |
---|
379 | url = self._ulib.urlopen(cbase) |
---|
380 | # Otherwise, mark as visited and abort |
---|
381 | else: |
---|
382 | self._visited[cbase] = 1 |
---|
383 | return False |
---|
384 | # If HTTP error, log bad URL and abort |
---|
385 | except IOError: |
---|
386 | self._visited[cbase] = 1 |
---|
387 | self.badurls.append((base[1], cbase)) |
---|
388 | return False |
---|
389 | # Get real URL |
---|
390 | newbase = url.geturl() |
---|
391 | # Change URL if different from old URL |
---|
392 | if newbase != cbase: cbase, base = newbase, (newbase, base[1]) |
---|
393 | # URLs with mimetype 'text/html" scanned for URLs |
---|
394 | if url.headers.type == 'text/html': |
---|
395 | # Feed parser |
---|
396 | contents = url.read() |
---|
397 | try: badurl, urls = self._webparser(contents) |
---|
398 | # Log URL if SGML parser can't parse it |
---|
399 | except self._sperror: |
---|
400 | self._visited[cbase], self.badhtm[cbase] = 1, 1 |
---|
401 | return False |
---|
402 | url.close() |
---|
403 | # Return URL and extracted urls if it's good |
---|
404 | if not badurl: return cbase, urls |
---|
405 | # If the URL is bad (after BadUrl), stop processing and log URL |
---|
406 | else: |
---|
407 | self._visited[cbase] = 1 |
---|
408 | self.badurls.append((base[1], cbase)) |
---|
409 | return False |
---|
410 | # Return URL of non-HTML resources and empty list |
---|
411 | else: |
---|
412 | url.close() |
---|
413 | return cbase, [] |
---|
414 | |
---|
415 | def _genverify(self, urls, base): |
---|
416 | '''Verifies a list of full URL relative to a base URL |
---|
417 | |
---|
418 | Arguments: |
---|
419 | urls -- list of raw URLs |
---|
420 | base -- referring URL''' |
---|
421 | # Assignments |
---|
422 | cache, visit, urlverify = self._cache, self._visited, self._urlverify |
---|
423 | # Strip file off base URL for joining |
---|
424 | newbase = base.replace(base.split('/')[-1], '') |
---|
425 | for url in urls: |
---|
426 | # Get resolved url and raw child URLs |
---|
427 | url, rawurls = urlverify(url, base, newbase) |
---|
428 | # Handle any child URLs |
---|
429 | if rawurls: |
---|
430 | newurls = {} |
---|
431 | # Eliminate duplicate URLs |
---|
432 | for rawurl in rawurls: |
---|
433 | # Eliminate known visited URLs |
---|
434 | if rawurl not in visit: newurls[rawurl] = 1 |
---|
435 | # Put new URLs in cache if present |
---|
436 | if newurls: cache[url] = newurls |
---|
437 | # Yield new URL |
---|
438 | if url: yield url |
---|
439 | |
---|
440 | def _multiverify(self, url, base): |
---|
441 | '''Verifies a full URL relative to a base URL |
---|
442 | |
---|
443 | Arguments: |
---|
444 | url -- a raw URLs |
---|
445 | base -- referring URL''' |
---|
446 | # Assignments |
---|
447 | cache, visited = self._cache, self._visited |
---|
448 | # Strip file off base URL for joining |
---|
449 | newbase = base.replace(base.split('/')[-1], '') |
---|
450 | # Get resolved url and raw child URLs |
---|
451 | url, rawurls = self._urlverify(url, base, newbase) |
---|
452 | # Handle any child URLs |
---|
453 | if rawurls: |
---|
454 | # Eliminate known visited URLs and duplicates |
---|
455 | for rawurl in rawurls: |
---|
456 | # Put new URLs in cache if present |
---|
457 | if rawurl not in visited: cache[rawurl] = url |
---|
458 | # Put URL in list of good URLs |
---|
459 | if url: self._good[url] = 1 |
---|
460 | |
---|
461 | def _urlverify(self, url, base, newbase): |
---|
462 | '''Returns a full URL relative to a base URL |
---|
463 | |
---|
464 | Arguments: |
---|
465 | urls -- list of raw URLs |
---|
466 | base -- referring URL |
---|
467 | newbase -- temporary version of referring URL for joining''' |
---|
468 | # Assignments |
---|
469 | visited, webopen, other = self._visited, self._webopen, self.other |
---|
470 | sb, depth, urljoin = self._sb[2], self.depth, self._uparse.urljoin |
---|
471 | urlsplit, urldefrag = self._uparse.urlsplit, self._uparse.urldefrag |
---|
472 | outside, redirs, supported = self.outside, self.redirs, self._supported |
---|
473 | if url not in visited: |
---|
474 | # Remove whitespace from URL |
---|
475 | if url.find(' ') != -1: |
---|
476 | visited[url], url = 1, url.replace(' ', '') |
---|
477 | if url in visited: return 0, 0 |
---|
478 | # Remove fragments i.e. 'http:foo/bar#frag' |
---|
479 | if url.find('#') != -1: |
---|
480 | visited[url], url = 1, urldefrag(url)[0] |
---|
481 | if url in visited: return 0, 0 |
---|
482 | # Process full URLs i.e. 'http://foo/bar |
---|
483 | if url.find(':') != -1: |
---|
484 | urlseg = urlsplit(url) |
---|
485 | # Block non-FTP, HTTP URLs |
---|
486 | if urlseg[0] not in supported: |
---|
487 | # Log as non-FTP/HTTP URL |
---|
488 | other[url], visited[url] = 1, 1 |
---|
489 | return 0, 0 |
---|
490 | # If URL is not in root domain, block it |
---|
491 | if urlseg[1] not in sb: |
---|
492 | visited[url], outside[url] = 1, 1 |
---|
493 | return 0, 0 |
---|
494 | # Block duplicate root URLs |
---|
495 | elif not urlseg[2] and urlseg[1] == sb: |
---|
496 | visited[url] = 1 |
---|
497 | return 0, 0 |
---|
498 | # Handle relative URLs i.e. ../foo/bar |
---|
499 | elif url.find(':') == -1: |
---|
500 | # Join root domain and relative URL |
---|
501 | visited[url], url = 1, urljoin(newbase, url) |
---|
502 | if url in visited: return 0, 0 |
---|
503 | # Test URL by attempting to open it |
---|
504 | rurl = webopen((url, base)) |
---|
505 | if rurl and rurl[0] not in visited: |
---|
506 | # Get URL |
---|
507 | turl, rawurls = rurl |
---|
508 | visited[url], visited[turl] = 1, 1 |
---|
509 | # If URL resolved to a different URL, process it |
---|
510 | if turl != url: |
---|
511 | urlseg = urlsplit(turl) |
---|
512 | # If URL is not in root domain, block it |
---|
513 | if urlseg[1] not in sb: |
---|
514 | # Log as a redirected internal URL |
---|
515 | redirs[(url, turl)] = 1 |
---|
516 | return 0, 0 |
---|
517 | # Block duplicate root URLs |
---|
518 | elif not urlseg[2] and urlseg[1] == sb: return 0, 0 |
---|
519 | # If URL exceeds depth, don't process |
---|
520 | if len(turl.split('/')) >= depth: return 0, 0 |
---|
521 | # Otherwise return URL |
---|
522 | else: |
---|
523 | if rawurls: return turl, rawurls |
---|
524 | else: return turl, [] |
---|
525 | else: return 0,0 |
---|
526 | else: return 0, 0 |
---|
527 | |
---|
528 | def _onewalk(self): |
---|
529 | '''Yields good URLs from under a base URL''' |
---|
530 | # Assignments |
---|
531 | cache, genverify = self._cache, self._genverify |
---|
532 | # End processing if cache is empty |
---|
533 | while cache: |
---|
534 | # Fetch item from cache |
---|
535 | base, urls = cache.popitem() |
---|
536 | # If item has child URLs, process them and yield good URLs |
---|
537 | if urls: |
---|
538 | for url in genverify(urls, base): yield url |
---|
539 | |
---|
540 | def _multiwalk(self, threads): |
---|
541 | '''Extracts good URLs from under a base URL |
---|
542 | |
---|
543 | Arguments: |
---|
544 | threads -- number of threads to run''' |
---|
545 | |
---|
546 | def urlthread(url, base): |
---|
547 | '''Spawns a thread containing a multiverify function |
---|
548 | |
---|
549 | Arguments: |
---|
550 | |
---|
551 | url -- URL to verify |
---|
552 | base -- referring URL''' |
---|
553 | # Create instance of Thread |
---|
554 | dthread = Thread(target=multiverify, args=(url, base)) |
---|
555 | # Put in pool |
---|
556 | pool.append(dthread) |
---|
557 | |
---|
558 | # Assignments |
---|
559 | pool, cache, multiverify = [], self._cache, self._multiverify |
---|
560 | Thread, width, good = self._thread, self.width, self._good |
---|
561 | # End processing if cache is empty |
---|
562 | while cache: |
---|
563 | # Process URLs as long as width not exceeded |
---|
564 | if len(good) <= width: |
---|
565 | # Fetch item from cache |
---|
566 | url, base = cache.popitem() |
---|
567 | # Make thread |
---|
568 | if url: urlthread(url, base) |
---|
569 | # Run threads once pool size is reached |
---|
570 | if len(pool) == threads or threads >= len(cache): |
---|
571 | # Start threads |
---|
572 | for thread in pool: thread.start() |
---|
573 | # Empty thread pool as threads complete |
---|
574 | while pool: |
---|
575 | for thread in pool: |
---|
576 | if not thread.isAlive(): pool.remove(thread) |
---|
577 | # End if width reached |
---|
578 | elif len(good) >= width: break |
---|
579 | |
---|
580 | def weburls(self, base=None, width=200, depth=5, thread=None): |
---|
581 | '''Returns a list of web paths. |
---|
582 | |
---|
583 | Arguments: |
---|
584 | base -- base web URL (default: None) |
---|
585 | width -- amount of resources to crawl (default: 200) |
---|
586 | depth -- depth in hierarchy to crawl (default: 5) |
---|
587 | thread -- number of threads to run (default: None)''' |
---|
588 | # Assignments |
---|
589 | self._visited, self._good, self._cache, self.badurls = {}, {}, {}, [] |
---|
590 | self.redirs, self.outside, self.badhtm, self.other = {}, {}, {}, {} |
---|
591 | onewalk, good, self._robot = self._onewalk, self._good, self._rparser() |
---|
592 | uparse, robot, multiwalk = self._uparse, self._robot, self._multiwalk |
---|
593 | cache = self._cache |
---|
594 | # Assign width |
---|
595 | if self.width and width == 200: width = self.width |
---|
596 | else: self.width = width |
---|
597 | # sgmlop crashes Python after too many iterations |
---|
598 | if width > 5000: self._parserpick(1) |
---|
599 | else: self._parserpick() |
---|
600 | # Use global base if present |
---|
601 | if not base: base = self.base |
---|
602 | # Verify URL and get child URLs |
---|
603 | newbase, rawurls = self._webopen((base, '')) |
---|
604 | if newbase: |
---|
605 | # Change base URL if different |
---|
606 | if newbase != base: base = newbase |
---|
607 | # Ensure there's a trailing '/' in base URL |
---|
608 | if base[-1] != '/': |
---|
609 | url = list(uparse.urlsplit(base)) |
---|
610 | url[1] = ''.join([url[1], '/']) |
---|
611 | base = uparse.urlunsplit(url) |
---|
612 | # Eliminate duplicates and put raw URLs in cache |
---|
613 | newurls = {} |
---|
614 | for rawurl in rawurls: newurls[rawurl] = 1 |
---|
615 | if newurls: |
---|
616 | # Cache URLs individually if threads are desired |
---|
617 | if thread: |
---|
618 | for newurl in newurls: cache[newurl] = base |
---|
619 | # Cache in group if no threads |
---|
620 | else: cache[base] = newurls |
---|
621 | # Make base URL, get split, and put in verified URL list |
---|
622 | self.base, self._sb = base, base.split('/') |
---|
623 | self._visited[base], good[base] = 1, 1 |
---|
624 | # If URL is bad, abort and raise error |
---|
625 | else: raise IOError, "URL is invalid" |
---|
626 | # Adjust dept to length of base URL |
---|
627 | if self.depth and depth == 6: self.depth += len(self._sb) |
---|
628 | else: self.depth = depth + len(self._sb) |
---|
629 | # Get robot limits |
---|
630 | robot.set_url(''.join([base, 'robots.txt'])) |
---|
631 | robot.read() |
---|
632 | # Get signature of bad URL |
---|
633 | self._webtest() |
---|
634 | # Get good URLs as long as total width isn't exceeded |
---|
635 | try: |
---|
636 | # Multiwalk if threaded |
---|
637 | if thread: self._multiwalk(thread) |
---|
638 | # Otherwise, use single thread |
---|
639 | else: |
---|
640 | for item in onewalk(): |
---|
641 | # Don't exceed maximum width |
---|
642 | if len(good) <= width: good[item] = 1 |
---|
643 | elif len(good) >= width: break |
---|
644 | # If user interrupts crawl, return what's done |
---|
645 | except KeyboardInterrupt: pass |
---|
646 | # Get URLs, sort them, and return list |
---|
647 | self.urls = good.keys() |
---|
648 | self.urls.sort() |
---|
649 | return self.urls |
---|
650 | |
---|
651 | def webpaths(self, b=None, w=200, d=5, t=None): |
---|
652 | '''Returns a list of web paths. |
---|
653 | |
---|
654 | Arguments: |
---|
655 | b -- base web URL (default: None) |
---|
656 | w -- amount of resources to crawl (default: 200) |
---|
657 | d -- depth in hierarchy to crawl (default: 5) |
---|
658 | t -- number of threads (default: None)''' |
---|
659 | |
---|
660 | def pathize(): |
---|
661 | '''Strips base URL from full URLs to produce paths''' |
---|
662 | for url in urls: |
---|
663 | # Remove base URL from path list |
---|
664 | url = url.replace(self.base, '') |
---|
665 | # Add default name 'index.html' to root URLs and directories |
---|
666 | if not url: url = 'index.html' |
---|
667 | elif url[-1] == '/': url = ''.join([url, 'index.html']) |
---|
668 | # Verify removal of base URL and remove it if found |
---|
669 | if url.find(':') != -1: url = urlsplit(url)[2:][0] |
---|
670 | yield url |
---|
671 | |
---|
672 | # Assignments |
---|
673 | urlsplit = self._uparse.urlsplit |
---|
674 | # Run weburls if base passed as an argument |
---|
675 | if b: self.weburls(b, w, d, t) |
---|
676 | # Strip off trailing resource or query from base URL |
---|
677 | if self.base[-1] != '/': self.base = '/'.join(self._sb[:-1]) |
---|
678 | urls = self.urls |
---|
679 | # Return path list after stripping base URL |
---|
680 | self.paths = list(pathize()) |
---|
681 | return self.paths |
---|
682 | |
---|
683 | def webmirror(self, root=None, t=None, base=None, width=200, depth=5): |
---|
684 | '''Mirrors a website on a local filesystem |
---|
685 | |
---|
686 | Arguments: |
---|
687 | root -- local filesystem path (default: None) |
---|
688 | t -- number of threads (default: None) |
---|
689 | base -- base web URL (default: None) |
---|
690 | width -- amount of resources to crawl (default: 200) |
---|
691 | depth -- depth in hierarchy to crawl (default: 5)''' |
---|
692 | if base: self.webspider(base, width, depth, t) |
---|
693 | return self._mirror((self.paths, self.urls), root, t) |
---|
694 | |
---|
695 | def webspider(self, b=None, w=200, d=5, t=None): |
---|
696 | '''Returns two lists of child URLs and paths |
---|
697 | |
---|
698 | Arguments: |
---|
699 | b -- base web URL (default: None) |
---|
700 | w -- amount of resources to crawl (default: 200) |
---|
701 | d -- depth in hierarchy to crawl (default: 5) |
---|
702 | t -- number of threads (default: None)''' |
---|
703 | if b: self.weburls(b, w, d, t) |
---|
704 | return self.webpaths(), self.urls |
---|
705 | |
---|
706 | def badurlreport(self, f=None, b=None, w=200, d=5, t=None): |
---|
707 | '''Pretties up a list of bad URLs |
---|
708 | |
---|
709 | Arguments: |
---|
710 | f -- output file for report (default: None) |
---|
711 | b -- base web URL (default: None) |
---|
712 | w -- amount of resources to crawl (default: 200) |
---|
713 | d -- depth in hierarchy to crawl (default: 5) |
---|
714 | t -- number of threads (default: None)''' |
---|
715 | if b: self.weburls(b, w, d, t) |
---|
716 | # Format report if information is available |
---|
717 | if self.badurls: |
---|
718 | # Number of bad URLs |
---|
719 | amount = str(len(self.badurls)) |
---|
720 | header = '%s broken URLs under %s on %s:\n' |
---|
721 | # Print referring URL pointing to bad URL |
---|
722 | body = '\n'.join([' -> '.join([i[0], i[1]]) for i in self.badurls]) |
---|
723 | report = self._formatreport(amount, header, body, f) |
---|
724 | # Return if just getting string |
---|
725 | if report: return report |
---|
726 | |
---|
727 | def badhtmreport(self, f=None, b=None, w=200, d=5, t=None): |
---|
728 | '''Pretties up a list of unparsed HTML URLs |
---|
729 | |
---|
730 | Arguments: |
---|
731 | f -- output file for report (default: None) |
---|
732 | b -- base web URL (default: None) |
---|
733 | w -- amount of resources to crawl (default: 200) |
---|
734 | d -- depth in hierarchy to crawl (default: 5) |
---|
735 | t -- number of threads (default: None)''' |
---|
736 | if b: self.weburls(b, w, d, t) |
---|
737 | # Format report if information is available |
---|
738 | if self.badhtm: |
---|
739 | amount = str(len(self.badhtm)) |
---|
740 | header = '%s unparsable HTML URLs under %s on %s:\n' |
---|
741 | body = '\n'.join(self.badhtm) |
---|
742 | report = self._formatreport(amount, header, body, f) |
---|
743 | # Return if just getting string |
---|
744 | if report: return report |
---|
745 | |
---|
746 | def redireport(self, f=None, b=None, w=200, d=5, t=None): |
---|
747 | '''Pretties up a list of URLs redirected to an external URL |
---|
748 | |
---|
749 | Arguments: |
---|
750 | f -- output file for report (default: None) |
---|
751 | b -- base web URL (default: None) |
---|
752 | w -- amount of resources to crawl (default: 200) |
---|
753 | d -- depth in hierarchy to crawl (default: 5) |
---|
754 | t -- number of threads (default: None)''' |
---|
755 | if b: self.weburls(b, w, d, t) |
---|
756 | # Format report if information is available |
---|
757 | if self.redirs: |
---|
758 | amount = str(len(self.redirs)) |
---|
759 | header = '%s redirects to external URLs under %s on %s:\n' |
---|
760 | # Print referring URL pointing to new URL |
---|
761 | body = '\n'.join([' -> '.join([i[0], i[1]]) for i in self.redirs]) |
---|
762 | report = self._formatreport(amount, header, body, f) |
---|
763 | # Return if just getting string |
---|
764 | if report: return report |
---|
765 | |
---|
766 | def outreport(self, f=None, b=None, w=200, d=5, t=None): |
---|
767 | '''Pretties up a list of outside URLs referenced under the base URL |
---|
768 | |
---|
769 | Arguments: |
---|
770 | f -- output file for report (default: None) |
---|
771 | b -- base web URL (default: None) |
---|
772 | w -- amount of resources to crawl (default: 200) |
---|
773 | d -- depth in hierarchy to crawl (default: 5) |
---|
774 | t -- number of threads (default: None)''' |
---|
775 | if b: self.weburls(b, w, d, t) |
---|
776 | # Format report if information is available |
---|
777 | if self.outside: |
---|
778 | amount = str(len(self.outside)) |
---|
779 | header = '%s links to external URLs under %s on %s:\n' |
---|
780 | body = '\n'.join(self.outside) |
---|
781 | report = self._formatreport(amount, header, body, f) |
---|
782 | # Return if just getting string |
---|
783 | if report: return report |
---|
784 | |
---|
785 | def othereport(self, f=None, b=None, w=200, d=5, t=None): |
---|
786 | '''Pretties up a list of non-HTTP/FTP URLs |
---|
787 | |
---|
788 | Arguments: |
---|
789 | f -- output file for report (default: None) |
---|
790 | b -- base web URL (default: None) |
---|
791 | w -- amount of resources to crawl (default: 200) |
---|
792 | d -- depth in hierarchy to crawl (default: 5) |
---|
793 | t -- number of threads (default: None)''' |
---|
794 | if b: self.weburls(b, w, d, t) |
---|
795 | # Format report if information is available |
---|
796 | if self.other: |
---|
797 | amount = str(len(self.other)) |
---|
798 | header = '%s non-FTP/non-HTTP URLs under %s on %s:\n' |
---|
799 | body = '\n'.join(self.other) |
---|
800 | report = self._formatreport(amount, header, body, f) |
---|
801 | # Return if just getting string |
---|
802 | if report: return report |
---|
803 | |
---|
804 | def urlreport(self, f=None, b=None, w=200, d=5, t=None): |
---|
805 | '''Pretties up a list of all URLs under a URL |
---|
806 | |
---|
807 | Arguments: |
---|
808 | f -- output file for report (default: None) |
---|
809 | b -- base web URL (default: None) |
---|
810 | w -- amount of resources to crawl (default: 200) |
---|
811 | d -- depth in hierarchy to crawl (default: 5) |
---|
812 | t -- number of threads (default: None)''' |
---|
813 | if b: self.weburls(b, w, d, t) |
---|
814 | # Format report if information is available |
---|
815 | if self.urls: |
---|
816 | amount = str(len(self.urls)) |
---|
817 | header = '%s verified URLs under %s on %s:\n' |
---|
818 | body = '\n'.join(self.urls) |
---|
819 | report = self._formatreport(amount, header, body, f) |
---|
820 | # Return if just getting string |
---|
821 | if report: return report |
---|
822 | |
---|
823 | def webreport(self, f=None, b=None, w=200, d=5, t=None, *vargs): |
---|
824 | '''Pretties up a list of logged information under a URL |
---|
825 | |
---|
826 | Arguments: |
---|
827 | f -- output file for report (default: None) |
---|
828 | b -- base web URL (default: None) |
---|
829 | w -- amount of resources to crawl (default: 200) |
---|
830 | d -- depth in hierarchy to crawl (default: 5) |
---|
831 | t -- number of threads (default: None) |
---|
832 | vargs -- report sections to include or exclude |
---|
833 | To override defaults: |
---|
834 | To include a section add 'badhtm', 'redirs', 'outside', or 'other' |
---|
835 | To exclude a section add 'badurls' or "urls"''' |
---|
836 | if b: self.weburls(b, w, d, t) |
---|
837 | # Defaults for report |
---|
838 | badurls, badhtm, redirs, urls, outside, other = 1, 0, 0, 1, 0, 0 |
---|
839 | # Create compilation list |
---|
840 | compile = [] |
---|
841 | # Override default report settings if argument is passed to vargs |
---|
842 | for arg in vargs: |
---|
843 | if arg == 'badurls': badurls = 0 |
---|
844 | elif arg == 'badhtm': badhtm = 1 |
---|
845 | elif arg == 'redirs': redirs = 1 |
---|
846 | elif arg == 'urls': urls = 0 |
---|
847 | elif arg == 'outside': outside = 1 |
---|
848 | elif arg == 'other': other = 1 |
---|
849 | # Compile report |
---|
850 | if badurls: |
---|
851 | badurls = self.badurlreport() |
---|
852 | if badurls: compile.append(badurls) |
---|
853 | if urls: |
---|
854 | urls = self.urlreport() |
---|
855 | if urls: compile.append(urls) |
---|
856 | if outside: |
---|
857 | outside = self.outreport() |
---|
858 | if outside: compile.append(outside) |
---|
859 | if redirs: |
---|
860 | redirs = self.redireport() |
---|
861 | if redirs: compile.append(redirs) |
---|
862 | if badhtm: |
---|
863 | badhtm = self.badhtmreport() |
---|
864 | if badhtm: compile.append(badhtm) |
---|
865 | if other: |
---|
866 | other = self.othereport() |
---|
867 | if other: compile.append(other) |
---|
868 | # Make report |
---|
869 | report = '\n\n'.join(compile) |
---|
870 | # Write to file if argument present |
---|
871 | if file: open(f, 'w').write(report) |
---|
872 | # Or return string |
---|
873 | else: return report |
---|
874 | |
---|
875 | def _formatreport(self, amount, header, body, file=None): |
---|
876 | '''Generic prettifier with date/time stamper |
---|
877 | |
---|
878 | Arguments: |
---|
879 | header -- title of report |
---|
880 | body -- body of report |
---|
881 | file -- output file for report (default: None)''' |
---|
882 | # Get current time |
---|
883 | localtime, strftime = self._localtime, self._formtime |
---|
884 | curtime = strftime('%A, %B %d, %Y at %I:%M %p', localtime()) |
---|
885 | # Make section header |
---|
886 | header = header % (amount, self.base, curtime) |
---|
887 | # Add header to body |
---|
888 | report = '\n'.join([header, body]) |
---|
889 | # Write to file if argument present |
---|
890 | if file: open(file, 'w').write(report) |
---|
891 | # Or return string |
---|
892 | else: return report |
---|
893 | |
---|
894 | def _mirror(self, lists, root=None, threads=None): |
---|
895 | '''Mirrors a site on a local filesystem based on lists passed to it |
---|
896 | |
---|
897 | Argument: |
---|
898 | lists -- lists of URLs and paths |
---|
899 | root -- local filesystem path (default: None) |
---|
900 | threads -- number of threads (default: None)''' |
---|
901 | |
---|
902 | def download(url, np, op): |
---|
903 | '''Downloads files that need to be mirrored.''' |
---|
904 | # If ftp... |
---|
905 | if self._path.exists(np): |
---|
906 | return None |
---|
907 | if url[:3] == 'ftp': |
---|
908 | # Open local file |
---|
909 | local = open(np, 'wb') |
---|
910 | # Download using FTP session |
---|
911 | ftp = ftpopen(base, name, password) |
---|
912 | ftp.retrbinary('RETR %s' % op, local.write) |
---|
913 | ftp.close() |
---|
914 | # Close local file |
---|
915 | local.close() |
---|
916 | # Use normal urlretrieve if no FTP required |
---|
917 | else: ulib.urlretrieve(url, np) |
---|
918 | |
---|
919 | def dlthread(url, np, op): |
---|
920 | '''Spawns a thread containing the download function''' |
---|
921 | # Create thread |
---|
922 | dthread = Thread(target=download, args=(url, np, op)) |
---|
923 | # Add to thread pool |
---|
924 | pool.append(dthread) |
---|
925 | |
---|
926 | # Extract path and URL lists |
---|
927 | paths, urls = lists |
---|
928 | # Avoid outside namespace lookups |
---|
929 | ulib, makedirs, sep = self._ulib, self._os.makedirs, self._os.sep |
---|
930 | normcase, split = self._path.normcase, self._path.split |
---|
931 | exists, isdir = self._path.exists, self._path.isdir |
---|
932 | ftpopen = self._ftpopen |
---|
933 | # Create local names for thread class and thread pool |
---|
934 | if threads: Thread, pool = self._thread, [] |
---|
935 | # Localize name and password if exists |
---|
936 | try: base, name, password = self.base, self._name, self._password |
---|
937 | except AttributeError: pass |
---|
938 | # Change to directory if given... |
---|
939 | if root: |
---|
940 | if exists(root): |
---|
941 | if isdir(root): self._os.chdir(root) |
---|
942 | # Create root if it doesn't exist |
---|
943 | else: |
---|
944 | makedirs(root) |
---|
945 | self._os.chdir(root) |
---|
946 | # Otherwise use current directory |
---|
947 | else: root = self._os.getcwd() |
---|
948 | # Iterate over paths and download files |
---|
949 | for oldpath in paths: |
---|
950 | # Sync with the URL for oldpath |
---|
951 | url = urls[paths.index(oldpath)] |
---|
952 | # Create name of local copy |
---|
953 | newpath = normcase(oldpath).lstrip(sep) |
---|
954 | # Get directory name |
---|
955 | dirname = split(newpath)[0] |
---|
956 | # If the directory exists, download the file directly |
---|
957 | if exists(dirname): |
---|
958 | if isdir(dirname): |
---|
959 | if threads: dlthread(url, newpath, oldpath) |
---|
960 | else: download(url, newpath, oldpath) |
---|
961 | # Don't create local directory if path in root of remote URL |
---|
962 | elif not dirname: |
---|
963 | if threads: dlthread(url, newpath, oldpath) |
---|
964 | else: download(url, newpath, oldpath) |
---|
965 | # Make local directory if it doesn't exist, then dowload file |
---|
966 | else: |
---|
967 | makedirs(dirname) |
---|
968 | if threads: dlthread(url, newpath, oldpath) |
---|
969 | else: download(url, newpath, oldpath) |
---|
970 | # Run threads if they've hit the max number of threads allowed |
---|
971 | if threads: |
---|
972 | # Run if max threads or final thread reached |
---|
973 | if len(pool) == threads or paths[-1] == oldpath: |
---|
974 | # Start all threads |
---|
975 | for thread in pool: thread.start() |
---|
976 | # Clear the thread pool as they finish |
---|
977 | while pool: |
---|
978 | for thread in pool: |
---|
979 | if not thread.isAlive(): pool.remove(thread) |
---|
980 | |
---|
981 | |
---|
982 | # Instance of Spider enables exporting Spider's methods as standalone functions |
---|
983 | _inst = Spider() |
---|
984 | ftpurls = _inst.ftpurls |
---|
985 | weburls = _inst.weburls |
---|
986 | ftppaths = _inst.ftppaths |
---|
987 | webpaths = _inst.webpaths |
---|
988 | ftpmirror = _inst.ftpmirror |
---|
989 | ftpspider = _inst.ftpspider |
---|
990 | webmirror = _inst.webmirror |
---|
991 | webspider = _inst.webspider |
---|
992 | webreport = _inst.webreport |
---|
993 | urlreport = _inst.urlreport |
---|
994 | outreport = _inst.outreport |
---|
995 | redireport = _inst.redireport |
---|
996 | othereport = _inst.othereport |
---|
997 | badurlreport = _inst.badurlreport |
---|
998 | badhtmreport = _inst.badhtmreport |
---|