2 # -*- coding: utf-8 -*-
4 # Copyright 1999, 2000 Martin Pool <mbp@humbug.org.au>
5 # Copyright 2002 Gerardo Poggiali
6 # Copyright 2007, 2008 Bernardo Innocenti <bernie@codewiz.org>
8 # This program is free software: you can redistribute it and/or modify
9 # it under the terms of the GNU General Public License as published by
10 # the Free Software Foundation, either version 3 of the License, or
11 # (at your option) any later version.
13 # This program is distributed in the hope that it will be useful, but
14 # WITHOUT ANY WARRANTY; without even the implied warranty of
15 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
16 # General Public License for more details.
18 # You should have received a copy of the GNU General Public License
19 # along with this program. If not, see <http://www.gnu.org/licenses/>.
21 __version__ = '$Id$'[4:12]
23 from time import clock
26 import cgi, sys, os, re, errno, stat
27 from os import path, environ
29 # Regular expression defining a WikiWord
30 # (but this definition is also assumed in other places)
31 word_re = re.compile(r"^\b((([A-Z][a-z]+){2,}/)*([A-Z][a-z]+){2,})\b$")
32 # FIXME: we accept stuff like foo/../bar and we shouldn't
33 file_re = re.compile(r"^\b([A-Za-z0-9_\-][A-Za-z0-9_\.\-/]*)\b$")
34 img_re = re.compile(r"^.*\.(png|gif|jpg|jpeg)$", re.IGNORECASE)
35 url_re = re.compile(r"^[a-z]{3,8}://[^\s'\"]+\S$")
40 # CGI stuff ---------------------------------------------------------
43 return environ.get('SCRIPT_NAME', '')
45 def privileged_path():
46 return privileged_url or script_name()
49 user = environ.get('REMOTE_USER', '')
50 if user is None or user == '' or user == 'anonymous':
51 user = 'AnonymousCoward'
55 return environ.get('REMOTE_ADDR', '')
57 def get_hostname(addr):
59 from socket import gethostbyaddr
60 return gethostbyaddr(addr)[0] + ' (' + addr + ')'
64 def relative_url(pathname, privileged=False):
65 if not (url_re.match(pathname) or pathname.startswith('/')):
67 url = privileged_path()
70 pathname = url + '/' + pathname
73 # Formatting stuff --------------------------------------------------
75 def emit_header(mime_type="text/html"):
76 print "Content-type: " + mime_type + "; charset=utf-8\n"
78 def send_guru(msg_text, msg_type):
79 if not msg_text: return
80 print '<pre id="guru" onclick="this.style.display = \'none\'" class="' + msg_type + '">'
81 if msg_type == 'error':
82 print ' Software Failure. Press left mouse button to continue.\n'
84 if msg_type == 'error':
85 print ' Guru Meditation #DEADBEEF.ABADC0DE'
87 # FIXME: This little JS snippet is harder to pass than ACID 3.0
89 <script language="JavaScript" type="text/javascript">
90 var guru = document.getElementById('guru');
91 // Firefox 2.0 doesn't take border-color, but returns border-top-color fine
92 var color = document.defaultView.getComputedStyle(guru,null).getPropertyValue('border-top-color');
95 guru.style.setProperty('border-color', color, '');
96 setTimeout('guruOff()', 1000);
99 guru.style.setProperty('border-color', '#000000', '');
100 setTimeout('guruOn()', 1000);
102 // Safari 2.0 returns this rgba crap
103 // Konqueror 3.5.6 doesn't seem to support computed properties
104 if (color && color != 'rgba(0, 0, 0, 0)') {
105 //window.alert("enabled! color='" + color + "'");
110 def send_title(name, text="Limbo", msg_text=None, msg_type='error'):
112 if title_done: return
116 print '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'
117 print ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">'
118 print '<html xmlns="http://www.w3.org/1999/xhtml" lang="en" xml:lang="en">'
120 site_name = globals().get('site_name', 'Unconfigured Site')
121 print "<head><title>%s: %s</title>" % (site_name, text)
122 print ' <meta http-equiv="Content-Type" content="application/xhtml+xml; charset=utf-8" />'
124 print ' <meta name="robots" content="noindex,nofollow" />'
126 print ' <link rel="stylesheet" type="text/css" href="%s" />' % relative_url(css)
130 if name and privileged_url is not None:
131 print '<body ondblclick="location.href=\'' + privileged_path() + '?edit=' + name + '\'">'
136 send_guru(msg_text, msg_type)
139 print '<div class="nav">'
141 print ' <b>' + link_tag('?fullsearch=' + name, text, 'navlink') + '</b> '
143 print ' <b>' + text + '</b> '
144 print ' | ' + link_tag('FrontPage', 'Front Page', 'navlink')
145 print ' | ' + link_tag('FindPage', 'Find Page', 'navlink')
146 if 'history_url' in globals():
147 print ' | <a href="' + history_url + '" class="navlink">Recent Changes</a>'
149 print ' | <a href="' + history_url + '?a=history;f=' + name + '" class="navlink">Page History</a>'
152 print ' | ' + link_tag('?raw=' + name, 'Raw Text', 'navlink')
153 if privileged_url is not None:
154 print ' | ' + link_tag('?edit=' + name, 'Edit Page', 'navlink', privileged=True)
156 print ' | <i>Immutable Page</i>'
159 if user != 'AnonymousCoward':
160 print ' | <span class="login"><i>logged in as <b>' + cgi.escape(user) + '</b></i></span>'
164 def link_tag(params, text=None, ss_class=None, privileged=False):
166 text = params # default
169 classattr += 'class="%s" ' % ss_class
170 # Prevent crawlers from following links potentially added by spammers or to generated pages
171 if ss_class == 'external' or ss_class == 'navlink':
172 classattr += 'rel="nofollow" '
173 elif url_re.match(params):
174 classattr += 'rel="nofollow" '
175 return '<a %shref="%s">%s</a>' % (classattr, relative_url(params, privileged=privileged), text)
177 # Search ---------------------------------------------------
179 def handle_fullsearch(needle):
180 send_title(None, 'Full text search for "%s"' % (needle))
182 needle_re = re.compile(needle, re.IGNORECASE)
184 all_pages = page_list()
185 for page_name in all_pages:
186 body = Page(page_name).get_raw_body()
187 count = len(needle_re.findall(body))
189 hits.append((count, page_name))
191 # The default comparison for tuples compares elements in order,
192 # so this sorts by number of hits
197 for (count, page_name) in hits:
198 print '<li><p>' + Page(page_name).link_to()
199 print ' . . . . ' + `count`
200 print ['match', 'matches'][count != 1]
204 print_search_stats(len(hits), len(all_pages))
206 def handle_titlesearch(needle):
207 # TODO: check needle is legal -- but probably we can just accept any RE
208 send_title(None, "Title search for \"" + needle + '"')
210 needle_re = re.compile(needle, re.IGNORECASE)
211 all_pages = page_list()
212 hits = filter(needle_re.search, all_pages)
215 for filename in hits:
216 print '<li><p>' + Page(filename).link_to() + "</p></li>"
219 print_search_stats(len(hits), len(all_pages))
221 def print_search_stats(hits, searched):
222 print "<p>%d hits out of %d pages searched.</p>" % (hits, searched)
224 def handle_raw(pagename):
225 Page(pagename).send_raw()
227 def handle_edit(pagename):
230 if form['file'].value:
231 pg.save(form['file'].file.read())
233 pg.save(form['savetext'].value.replace('\r\n', '\n'))
235 elif 'cancel' in form:
236 pg.msg_text = 'Editing canceled'
237 pg.msg_type = 'notice'
239 else: # preview or edit
241 if 'preview' in form:
242 text = form['savetext'].value
245 def make_index_key():
246 links = map(lambda ch: '<a href="#%s">%s</a>' % (ch, ch), 'abcdefghijklmnopqrstuvwxyz')
247 return '<p><center>'+ ' | '.join(links) + '</center></p>'
250 return filter(word_re.match, os.listdir(data_dir))
252 def send_footer(name, mod_string=None):
253 if globals().get('debug_cgi', False):
254 cgi.print_arguments()
257 print '<div id="footer"><hr />'
258 print ('<p class="copyright"><span class="benchmark">generated in %0.3fs</span>' +
259 ' by <a href="http://www.codewiz.org/wiki/GeekiGeeki">GeekiGeeki</a>' +
260 ' version %s</p>') % (clock() - start_time, __version__)
262 print '<p class="modified">last modified %s</p>' % mod_string
263 print '</div></body></html>'
265 # Macros ------------------------------------------------------------
266 def _macro_TitleSearch(*vargs):
267 return _macro_search("titlesearch")
269 def _macro_FullSearch(*vargs):
270 return _macro_search("fullsearch")
272 def _macro_search(type):
275 default = form['value'].value
276 return """<form method="get"><input name="%s" size="30" value="%s" /><input type="submit" value="Search" /></form>""" % (type, default)
278 def _macro_WordIndex(*vargs):
280 pages = list(page_list())
282 word_re = re.compile('[A-Z][a-z]+')
284 for word in word_re.findall(name):
286 map[word].append(name)
290 all_words = map.keys()
294 for word in all_words:
295 letter = word[0].lower()
296 if letter != last_letter:
297 s = s + '; <a name="%s"><h3>%s</h3></a>' % (letter, letter)
300 s = s + '<b>%s</b><ul>' % word
305 if name == last_page: continue
306 s = s + '<li>' + Page(name).link_to()
311 def _macro_TitleIndex(*vargs):
313 pages = list(page_list())
315 current_letter = None
317 letter = name[0].lower()
318 if letter != current_letter:
319 s += '<a name="%s"><h3>%s</h3></a>' % (letter, letter)
320 current_letter = letter
323 s += Page(name).link_to()
327 """Object that turns Wiki markup into HTML.
329 All formatting commands can be parsed one line at a time, though
330 some state is carried over between lines.
332 def __init__(self, raw):
335 self.in_pre = self.in_table = False
336 self.in_header = True
337 self.list_indents = []
348 "^^": ["sup", False],
352 def _b_repl(self, word):
353 style = self.styles[word]
354 style[1] = not style[1]
355 return ['</', '<'][style[1]] + style[0] + '>'
357 def _tit_repl(self, word):
359 result = '</h%d>' % self.h_level
362 self.h_level = len(word) - 1
363 abridged = re.sub('[^A-Za-z0-9_]', '', self.line).lower()
364 result = '<h%d id="%s"><a class="heading" href="#%s">¶</a> ' % (self.h_level, abridged, abridged)
367 def _br_repl(self, word):
370 def _rule_repl(self, word):
371 return self._undent() + '\n<hr size="%d" noshade="noshade" />\n' % (len(word) - 2)
373 def _word_repl(self, word):
374 return Page(word).link_to()
376 def _img_repl(self, word):
377 path = relative_url(word)
378 return '<a href="%s"><img border="0" src="%s" /></a>' % (path, path)
380 def _url_repl(self, word):
381 if img_re.match(word):
382 return '<a href="%s"><img border="0" src="%s" /></a>' % (word, word)
384 return '<a href="%s" rel="nofollow" class="external">%s</a>' % (word, word)
386 def _hurl_repl(self, word):
387 m = re.compile("\[\[([^ \t\n\r\f\v\|]+)(?:\s*\|\s*([^\]]+)|)\]\]").match(word)
389 descr = m.group(2) or name
391 macro = globals().get('_macro_' + name)
393 return macro(name, descr)
394 elif img_re.match(name):
395 name = relative_url(name)
396 # The "extthumb" nonsense works around a limitation of the HTML block model
397 return '<div class="extthumb"><div class="thumb"><a href="%s"><img border="0" src="%s" alt="%s" /></a><div class="caption">%s</div></div></div>' % (name, name, descr, descr)
399 if img_re.match(descr):
400 descr = '<img border="0" src="' + descr + '" />'
402 return link_tag(name, descr, 'wikilink')
404 def _email_repl(self, word):
405 return '<a href="mailto:%s">%s</a>' % (word, word)
407 def _html_repl(self, word):
408 return word; # Pass through
410 def _ent_repl(self, s):
411 return {'&': '&',
415 def _li_repl(self, match):
418 def _pre_repl(self, word):
419 if word == '{{{' and not self.in_pre:
427 def _hi_repl(self, word):
428 return '<strong class="highlight ' + word + '">' + word + '</strong>'
430 def _tr_repl(self, word):
432 if not self.in_table:
435 out = '</p><table><tbody>\n'
437 out = out + '<tr class="' + ['even', 'odd'][self.tr_cnt % 2] + '">'
438 return out + ['<td>', '<th>'][word.strip() == '||=']
440 def _td_repl(self, word):
442 return ['</td><td>', '</th><th>'][word.strip() == '||=']
445 def _tre_repl(self, word):
447 return ['</td></tr>', '</th></tr>'][word.strip() == '||=']
450 def _indent_level(self):
451 return len(self.list_indents) and self.list_indents[-1]
453 def _indent_to(self, new_level):
454 if self._indent_level() == new_level:
457 while self._indent_level() > new_level:
458 del(self.list_indents[-1])
460 while self._indent_level() < new_level:
461 self.list_indents.append(new_level)
468 res += '</ul>' * len(self.list_indents)
470 self.list_indents = []
473 def replace(self, match):
474 for type, hit in match.groupdict().items():
476 return getattr(self, '_' + type + '_repl')(hit)
478 raise "Can't handle match " + `match`
480 def print_html(self):
481 print '<div class="wiki"><p>'
483 # For each line, we scan through looking for magic
484 # strings, outputting verbatim any intervening text
485 # TODO: highlight search words (look at referrer)
486 scan_re = re.compile(
489 + r"(?P<b>\*\*|'''|//|''|##|``|__|\^\^|,,)"
490 + r"|(?P<tit>\={2,6})"
492 + r"|(?P<rule>^-{3,})"
493 + r"|(?P<html><(/|)(div|span|iframe)[^<>]*>)"
495 + r"|(?P<hi>\b(FIXME|TODO|DONE)\b)"
498 + r"|(?P<img>\b[a-zA-Z0-9_-]+\.(png|gif|jpg|jpeg|bmp))"
499 + r"|(?P<word>\b(?:[A-Z][a-z]+){2,}\b)"
500 + r"|(?P<hurl>\[\[([^ \t\n\r\f\v\|]+)(?:\s*\|\s*([^\]]+)|)\]\])"
501 + r"|(?P<url>(http|https|ftp|mailto)\:[^\s'\"]+\S)"
502 + r"|(?P<email>[-\w._+]+\@[\w.-]+)"
505 + r"|(?P<li>^\s+[\*#] +)"
506 + r"|(?P<pre>\{\{\{|\s*\}\}\})"
509 + r"|(?P<tr>^\s*\|\|(=|)\s*)"
510 + r"|(?P<tre>\s*\|\|(=|)\s*$)"
511 + r"|(?P<td>\s*\|\|(=|)\s*)"
515 + r"(?P<pre>\s*\}\}\})"
518 blank_re = re.compile(r"^\s*$")
519 indent_re = re.compile(r"^\s*")
520 tr_re = re.compile(r"^\s*\|\|")
521 eol_re = re.compile(r"\r?\n")
522 for self.line in eol_re.split(self.raw.expandtabs()):
525 if self.line.startswith('#'):
527 self.in_header = False
530 print re.sub(pre_re, self.replace, self.line)
532 if self.in_table and not tr_re.match(self.line):
533 self.in_table = False
534 print '</tbody></table><p>'
536 if blank_re.match(self.line):
539 indent = indent_re.match(self.line)
540 print self._indent_to(len(indent.group(0)))
541 print re.sub(scan_re, self.replace, self.line)
543 if self.in_pre: print '</pre>'
544 if self.in_table: print '</tbody></table><p>'
549 def __init__(self, page_name):
550 self.page_name = page_name
552 self.msg_type = 'error'
554 def split_title(self):
555 # look for the end of words and the start of a new word and insert a space there
556 return re.sub('([a-z])([A-Z])', r'\1 \2', self.page_name)
558 def _text_filename(self):
559 return path.join(data_dir, self.page_name)
561 def _tmp_filename(self):
562 return path.join(data_dir, ('#' + self.page_name.replace('/','_') + '.' + `os.getpid()` + '#'))
566 os.stat(self._text_filename())
569 if er.errno == errno.ENOENT:
574 word = self.page_name
576 return link_tag(word, word, 'wikilink')
578 return link_tag(word, nonexist_pfx + word, 'nonexistent')
580 def get_raw_body(self):
582 return open(self._text_filename(), 'rb').read()
584 if er.errno == errno.ENOENT:
585 return '' # just doesn't exist, use default
586 if er.errno == errno.EISDIR:
591 if 'attrs' in self.__dict__:
595 file = open(self._text_filename(), 'rt')
596 attr_re = re.compile(r"^#(\S*)(.*)$")
598 m = attr_re.match(line)
601 self.attrs[m.group(1)] = m.group(2).strip()
602 #print "bernie: attrs[" + m.group(1) + "] = " + m.group(2) + "<br>\n"
604 if er.errno != errno.ENOENT:
608 def get_attr(self, name, default):
609 return self.get_attrs().get(name, default)
611 def can(self, action, default=True):
613 #acl SomeUser:read,write All:read
614 acl = self.get_attr("acl", None)
615 for rule in acl.split():
616 (user, perms) = rule.split(':')
617 if user == remote_user() or user == "All":
618 return action in perms.split(',')
620 except Exception, er:
622 self.msg_text = 'Illegal acl line: ' + acl
626 return self.can("write", True)
629 return self.can("read", True)
634 page_name = self.page_name
638 css_url = self.get_attr("css", "").split() + css_url
640 send_title(page_name, self.split_title(), msg_text=self.msg_text, msg_type=self.msg_type)
642 PageFormatter(self.get_raw_body()).print_html()
644 send_guru("Read access denied by ACLs", "notice")
645 send_footer(page_name, self._last_modified())
647 def _last_modified(self):
649 from time import localtime, strftime
650 modtime = localtime(os.stat(self._text_filename())[stat.ST_MTIME])
652 if er.errno != errno.ENOENT:
655 return strftime(datetime_fmt, modtime)
657 def send_editor(self, preview=None):
658 send_title(None, 'Edit ' + self.split_title(), msg_text=self.msg_text, msg_type=self.msg_type)
659 if not self.can_write():
660 send_guru("Write access denied by ACLs", "error")
665 file = form['file'].value
667 print ('<p><b>Editing ' + self.page_name
668 + ' for ' + cgi.escape(remote_user())
669 + ' from ' + cgi.escape(get_hostname(remote_host()))
671 print '<div class="editor"><form method="post" enctype="multipart/form-data" action="%s">' % relative_url(self.page_name)
672 print '<input type="hidden" name="edit" value="%s">' % (self.page_name)
673 print '<textarea wrap="off" spellcheck="true" id="editor" name="savetext" rows="17" cols="100">%s</textarea>' % (preview or self.get_raw_body())
674 print 'Or upload a file: <input type="file" name="file" value="%s" />' % file
677 <input type="submit" name="save" value="Save" />
678 <input type="submit" name="preview" value="Preview" />
679 <input type="reset" value="Reset" />
680 <input type="submit" name="cancel" value="Cancel" />
683 print "<p>" + Page('EditingTips').link_to() + "</p>"
685 print "<div class='preview'>"
686 PageFormatter(preview).print_html()
688 send_footer(self.page_name)
690 def send_raw(self, mimetype='text/plain'):
692 body = self.get_raw_body()
693 emit_header(mimetype)
696 send_title(None, msg_text='Read access denied by ACLs', msg_type='notice')
698 def _write_file(self, data):
699 tmp_filename = self._tmp_filename()
700 open(tmp_filename, 'wb').write(data)
701 name = self._text_filename()
703 # Bad Bill! POSIX rename ought to replace. :-(
707 if er.errno != errno.ENOENT: raise er
708 os.rename(tmp_filename, name)
710 def save(self, newdata):
711 if not self.can_write():
712 self.msg_text = 'Write access denied by ACLs'
713 self.msg_type = 'error'
716 self._write_file(newdata)
719 # FIXME: what's the std way to perform shell quoting in python?
720 cmd = ( post_edit_hook
721 + " '" + data_dir + '/' + self.page_name
722 + "' '" + remote_user()
723 + "' '" + remote_host() + "'"
729 self.msg_text += "Post-editing hook returned %d.\n" % rc
730 self.msg_text += 'Command was: ' + cmd + '\n'
732 self.msg_text += 'Output follows:\n' + output
734 self.msg_text = 'Thank you for your contribution. Your attention to detail is appreciated.'
735 self.msg_type = 'success'
737 # Main ---------------------------------------------------------------
739 execfile("geekigeeki.conf.py")
740 form = cgi.FieldStorage()
743 handler = globals().get('handle_' + cmd)
745 handler(form[cmd].value)
748 path_info = environ.get('PATH_INFO', '')
749 if len(path_info) and path_info[0] == '/':
750 query = path_info[1:] or 'FrontPage'
752 query = environ.get('QUERY_STRING', '') or 'FrontPage'
754 if file_re.match(query):
755 if word_re.match(query):
756 Page(query).send_page()
758 from mimetypes import MimeTypes
759 type, encoding = MimeTypes().guess_type(query)
760 type = type or 'text/plain'
761 Page(query).send_raw(mimetype=type)
763 print "Status: 404 Not Found"
764 send_title(None, msg_text='Can\'t work out query: ' + query)
768 msg_text = traceback.format_exc()
770 send_guru(msg_text, "error")
772 send_title(None, msg_text=msg_text)