2 # -*- coding: utf-8 -*-
4 # Copyright 1999, 2000 Martin Pool <mbp@humbug.org.au>
5 # Copyright 2002 Gerardo Poggiali
6 # Copyright 2007, 2008 Bernie 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-z0-9]+){2,}/)*([A-Z][a-z0-9]+){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
74 return re.sub(' ', '-', re.sub('[^a-z0-9_ ]', '', s.lower()).strip())
76 # Formatting stuff --------------------------------------------------
77 def emit_header(mime_type="text/html"):
78 print "Content-type: " + mime_type + "; charset=utf-8\n"
80 def send_guru(msg_text, msg_type):
81 if not msg_text: return
82 print '<pre id="guru" onclick="this.style.display = \'none\'" class="' + msg_type + '">'
83 if msg_type == 'error':
84 print ' Software Failure. Press left mouse button to continue.\n'
86 if msg_type == 'error':
87 print ' Guru Meditation #DEADBEEF.ABADC0DE'
89 # FIXME: This little JS snippet is harder to pass than ACID 3.0
91 <script language="JavaScript" type="text/javascript">
92 var guru = document.getElementById('guru');
93 // Firefox 2.0 doesn't take border-color, but returns border-top-color fine
94 var color = document.defaultView.getComputedStyle(guru,null).getPropertyValue('border-top-color');
97 guru.style.setProperty('border-color', color, '');
98 setTimeout('guruOff()', 1000);
101 guru.style.setProperty('border-color', '#000000', '');
102 setTimeout('guruOn()', 1000);
104 // Safari 2.0 returns this rgba crap
105 // Konqueror 3.5.6 doesn't seem to support computed properties
106 if (color && color != 'rgba(0, 0, 0, 0)') {
107 //window.alert("enabled! color='" + color + "'");
112 def send_title(name, text="Limbo", msg_text=None, msg_type='error'):
114 if title_done: return
118 print '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'
119 print ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">'
120 print '<html xmlns="http://www.w3.org/1999/xhtml" lang="en" xml:lang="en">'
122 site_name = globals().get('site_name', 'Unconfigured Site')
123 print "<head><title>%s: %s</title>" % (site_name, text)
124 print ' <meta http-equiv="Content-Type" content="application/xhtml+xml; charset=utf-8" />'
126 print ' <meta name="robots" content="noindex,nofollow" />'
128 print ' <link rel="stylesheet" type="text/css" href="%s" />' % relative_url(css)
132 if name and privileged_url is not None:
133 print '<body ondblclick="location.href=\'' + privileged_path() + '?edit=' + name + '\'">'
138 send_guru(msg_text, msg_type)
141 print '<div class="nav">'
143 print ' <b>' + link_tag('?fullsearch=' + name, text, 'navlink') + '</b> '
145 print ' <b>' + text + '</b> '
146 print ' | ' + link_tag('FrontPage', 'Front Page', 'navlink')
147 print ' | ' + link_tag('FindPage', 'Find Page', 'navlink')
148 if 'history_url' in globals():
149 print ' | <a href="' + history_url + '" class="navlink">Recent Changes</a>'
151 print ' | <a href="' + history_url + '?a=history;f=' + name + '" class="navlink">Page History</a>'
154 print ' | ' + link_tag('?raw=' + name, 'Raw Text', 'navlink')
155 if privileged_url is not None:
156 print ' | ' + link_tag('?edit=' + name, 'Edit Page', 'navlink', privileged=True)
158 print ' | <i>Immutable Page</i>'
161 if user != 'AnonymousCoward':
162 print ' | <span class="login"><i>logged in as <b>' + cgi.escape(user) + '</b></i></span>'
166 def link_tag(params, text=None, ss_class=None, privileged=False):
168 text = params # default
171 classattr += 'class="%s" ' % ss_class
172 # Prevent crawlers from following links potentially added by spammers or to generated pages
173 if ss_class == 'external' or ss_class == 'navlink':
174 classattr += 'rel="nofollow" '
175 elif url_re.match(params):
176 classattr += 'rel="nofollow" '
177 return '<a %shref="%s">%s</a>' % (classattr, relative_url(params, privileged=privileged), text)
179 # Search ---------------------------------------------------
181 def handle_fullsearch(needle):
182 send_title(None, 'Full text search for "%s"' % (needle))
184 needle_re = re.compile(needle, re.IGNORECASE)
186 all_pages = page_list()
187 for page_name in all_pages:
188 body = Page(page_name).get_raw_body()
189 count = len(needle_re.findall(body))
191 hits.append((count, page_name))
193 # The default comparison for tuples compares elements in order,
194 # so this sorts by number of hits
199 for (count, page_name) in hits:
200 print '<li><p>' + Page(page_name).link_to()
201 print ' . . . . ' + `count`
202 print ['match', 'matches'][count != 1]
206 print_search_stats(len(hits), len(all_pages))
208 def handle_titlesearch(needle):
209 # TODO: check needle is legal -- but probably we can just accept any RE
210 send_title(None, "Title search for \"" + needle + '"')
212 needle_re = re.compile(needle, re.IGNORECASE)
213 all_pages = page_list()
214 hits = filter(needle_re.search, all_pages)
217 for filename in hits:
218 print '<li><p>' + Page(filename).link_to() + "</p></li>"
221 print_search_stats(len(hits), len(all_pages))
223 def print_search_stats(hits, searched):
224 print "<p>%d hits out of %d pages searched.</p>" % (hits, searched)
226 def handle_raw(pagename):
227 Page(pagename).send_raw()
229 def handle_edit(pagename):
232 if form['file'].value:
233 pg.save(form['file'].file.read())
235 pg.save(form['savetext'].value.replace('\r\n', '\n'))
237 elif 'cancel' in form:
238 pg.msg_text = 'Editing canceled'
239 pg.msg_type = 'notice'
241 else: # preview or edit
243 if 'preview' in form:
244 text = form['savetext'].value
247 def make_index_key():
248 links = map(lambda ch: '<a href="#%s">%s</a>' % (ch, ch), 'abcdefghijklmnopqrstuvwxyz')
249 return '<p><center>'+ ' | '.join(links) + '</center></p>'
251 def page_list(dir = None, re = word_re):
252 return filter(re.match, os.listdir(dir or data_dir))
254 def send_footer(name, mod_string=None):
255 if globals().get('debug_cgi', False):
256 cgi.print_arguments()
259 print '<div id="footer"><hr />'
260 print ('<p class="copyright"><span class="benchmark">generated in %0.3fs</span>' +
261 ' by <a href="http://www.codewiz.org/wiki/GeekiGeeki">GeekiGeeki</a>' +
262 ' version %s</p>') % (clock() - start_time, __version__)
264 print '<p class="modified">last modified %s</p>' % mod_string
265 print '</div></body></html>'
268 """Object that turns Wiki markup into HTML.
270 All formatting commands can be parsed one line at a time, though
271 some state is carried over between lines.
273 def __init__(self, raw):
276 self.in_pre = self.in_html = self.in_table = self.in_li = False
277 self.in_header = True
278 self.list_indents = []
289 "^^": ["sup", False],
293 def _b_repl(self, word):
294 style = self.styles[word]
295 style[1] = not style[1]
296 return ['</', '<'][style[1]] + style[0] + '>'
298 def _tit_repl(self, word):
300 result = '</h%d><p>\n' % self.h_level
303 self.h_level = len(word) - 1
304 link = permalink(self.line)
305 result = '\n</p><h%d id="%s"><a class="heading" href="#%s">¶</a> ' % (self.h_level, link, link)
308 def _br_repl(self, word):
311 def _rule_repl(self, word):
312 return self._undent() + '\n<hr size="%d" noshade="noshade" />\n' % (len(word) - 2)
314 def _word_repl(self, word):
315 return Page(word).link_to()
317 def _img_repl(self, word):
318 path = relative_url(word)
319 return '<a href="%s"><img border="0" src="%s" /></a>' % (path, path)
321 def _url_repl(self, word):
322 if img_re.match(word):
323 return '<a href="%s"><img border="0" src="%s" /></a>' % (word, word)
325 return '<a href="%s" rel="nofollow" class="external">%s</a>' % (word, word)
327 def _macro_repl(self, word):
328 m = re.compile("\<\<([^\s\|\>]+)(?:\s*\|\s*([^\>]+)|)\>\>").match(word)
332 argv.extend(m.group(2).split('|'))
333 argv = map(str.strip, argv)
335 macro = globals().get('_macro_' + name)
337 execfile("macros/" + name + ".py", globals())
338 macro = globals().get('_macro_' + name)
342 return '<strong class="error"><<' + '|'.join(argv) + '>></strong>'
344 def _hurl_repl(self, word):
345 m = re.compile("\[\[([^\s\|]+)(?:\s*\|\s*([^\]]+)|)\]\]").match(word)
347 descr = m.group(2) or name
349 if img_re.match(name):
350 name = relative_url(name)
351 # The "extthumb" nonsense works around a limitation of the HTML block model
352 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)
354 if img_re.match(descr):
355 descr = '<img border="0" src="' + descr + '" />'
357 return link_tag(name, descr, 'wikilink')
359 def _email_repl(self, word):
360 return '<a href="mailto:%s">%s</a>' % (word, word)
362 def _html_repl(self, word):
364 return word; # Pass through
366 def _ent_repl(self, s):
367 if self.in_html and s == '>':
370 return {'&': '&',
374 def _li_repl(self, match):
381 def _pre_repl(self, word):
382 if word == '{{{' and not self.in_pre:
390 def _hi_repl(self, word):
391 return '<strong class="highlight ' + word + '">' + word + '</strong>'
393 def _tr_repl(self, word):
395 if not self.in_table:
398 out = '</p><table><tbody>\n'
400 out = out + '<tr class="' + ['even', 'odd'][self.tr_cnt % 2] + '">'
401 return out + ['<td>', '<th>'][word.strip() == '||=']
403 def _td_repl(self, word):
405 return ['</td><td>', '</th><th>'][word.strip() == '||=']
408 def _tre_repl(self, word):
410 return ['</td></tr>', '</th></tr>'][word.strip() == '||=']
413 def _indent_level(self):
414 return len(self.list_indents) and self.list_indents[-1]
416 def _indent_to(self, new_level):
417 if self._indent_level() == new_level:
420 while self._indent_level() > new_level:
421 del(self.list_indents[-1])
424 self.in_li = False # FIXME
426 while self._indent_level() < new_level:
427 self.list_indents.append(new_level)
434 res += '</ul>' * len(self.list_indents)
436 self.list_indents = []
439 def replace(self, match):
440 for type, hit in match.groupdict().items():
442 return getattr(self, '_' + type + '_repl')(hit)
444 raise "Can't handle match " + `match`
446 def print_html(self):
447 print '<div class="wiki"><p>'
449 # For each line, we scan through looking for magic
450 # strings, outputting verbatim any intervening text
451 # TODO: highlight search words (look at referrer)
452 scan_re = re.compile(
455 + r"(?P<b>\*\*|'''|//|''|##|``|__|\^\^|,,)"
456 + r"|(?P<tit>\={2,6})"
458 + r"|(?P<rule>^-{3,})"
459 + r"|(?P<hi>\b(FIXME|TODO|DONE)\b)"
462 + r"|(?P<macro>\<\<([^\s\|\>]+)(?:\s*\|\s*([^\>]+)|)\>\>)"
463 + r"|(?P<hurl>\[\[([^\s\|]+)(?:\s*\|\s*([^\]]+)|)\]\])"
466 + r"|(?P<html><(/|)(br|hr|div|form|iframe|input|span))"
470 + r"|(?P<img>\b[a-zA-Z0-9_/-]+\.(png|gif|jpg|jpeg|bmp))" # LEGACY
471 + r"|(?P<word>\b(?:[A-Z][a-z]+){2,}\b)" # LEGACY
472 + r"|(?P<url>(http|https|ftp|mailto)\:[^\s'\"]+\S)" # LEGACY
473 + r"|(?P<email>[-\w._+]+\@[\w.-]+)" # LEGACY
476 + r"|(?P<li>^\s+[\*#] +)"
477 + r"|(?P<pre>\{\{\{|\s*\}\}\})"
478 + r"|(?P<inl>\{\{([^\s\|]+)(?:\s*\|\s*([^\]]+)|)\}\})" #TODO
481 + r"|(?P<tr>^\s*\|\|(=|)\s*)"
482 + r"|(?P<tre>\s*\|\|(=|)\s*$)"
483 + r"|(?P<td>\s*\|\|(=|)\s*)"
487 + r"(?P<pre>\s*\}\}\})"
490 blank_re = re.compile(r"^\s*$")
491 indent_re = re.compile(r"^\s*")
492 tr_re = re.compile(r"^\s*\|\|")
493 eol_re = re.compile(r"\r?\n")
494 for self.line in eol_re.split(self.raw.expandtabs()):
497 if self.line.startswith('#'):
499 self.in_header = False
502 print re.sub(pre_re, self.replace, self.line)
504 if self.in_table and not tr_re.match(self.line):
505 self.in_table = False
506 print '</tbody></table><p>'
508 if blank_re.match(self.line):
511 indent = indent_re.match(self.line)
512 print self._indent_to(len(indent.group(0)))
513 print re.sub(scan_re, self.replace, self.line)
515 if self.in_pre: print '</pre>'
516 if self.in_table: print '</tbody></table><p>'
521 def __init__(self, page_name):
522 self.page_name = page_name
524 self.msg_type = 'error'
526 def split_title(self):
527 # look for the end of words and the start of a new word and insert a space there
528 return re.sub('([a-z])([A-Z])', r'\1 \2', self.page_name)
531 return path.join(data_dir, self.page_name)
533 def _tmp_filename(self):
534 return path.join(data_dir, ('#' + self.page_name.replace('/','_') + '.' + `os.getpid()` + '#'))
538 os.stat(self._filename())
541 if er.errno == errno.ENOENT:
546 word = self.page_name
548 return link_tag(word, word, 'wikilink')
550 return link_tag(word, nonexist_pfx + word, 'nonexistent')
552 def get_raw_body(self):
554 return open(self._filename(), 'rb').read()
556 if er.errno == errno.ENOENT:
557 return '' # just doesn't exist, use default
558 if er.errno == errno.EISDIR:
559 return self.format_dir()
562 def format_dir(self):
564 for file in page_list(self._filename(), file_re):
565 if img_re.match(file):
566 out += ' * {{' + self.page_name + '/' + file + '}}\n'
568 out += ' * [[' + self.page_name + '/' + file + ']]\n'
571 if 'attrs' in self.__dict__:
575 file = open(self._filename(), 'rt')
576 attr_re = re.compile(r"^#(\S*)(.*)$")
578 m = attr_re.match(line)
581 self.attrs[m.group(1)] = m.group(2).strip()
582 #print "bernie: attrs[" + m.group(1) + "] = " + m.group(2) + "<br>\n"
584 if er.errno != errno.ENOENT:
588 def get_attr(self, name, default):
589 return self.get_attrs().get(name, default)
591 def can(self, action, default=True):
594 #acl SomeUser:read,write All:read
595 acl = self.get_attr("acl", None)
596 for rule in acl.split():
597 (user, perms) = rule.split(':')
598 if user == remote_user() or user == "All":
599 return action in perms.split(',')
601 except Exception, er:
603 self.msg_text = 'Illegal acl line: ' + acl
607 return self.can("write", True)
610 return self.can("read", True)
612 def send_naked(self):
614 WikiFormatter(self.get_raw_body()).print_html()
616 send_guru("Read access denied by ACLs", "notice")
621 page_name = self.page_name
625 css_url = self.get_attr("css", "").split() + css_url
627 send_title(page_name, self.split_title(), msg_text=self.msg_text, msg_type=self.msg_type)
629 send_footer(page_name, self._last_modified())
631 def _last_modified(self):
633 from time import localtime, strftime
634 modtime = localtime(os.stat(self._filename())[stat.ST_MTIME])
636 if er.errno != errno.ENOENT:
639 return strftime(datetime_fmt, modtime)
641 def send_editor(self, preview=None):
642 send_title(None, 'Edit ' + self.split_title(), msg_text=self.msg_text, msg_type=self.msg_type)
643 if not self.can_write():
644 send_guru("Write access denied by ACLs", "error")
649 file = form['file'].value
651 print ('<p><b>Editing ' + self.page_name
652 + ' for ' + cgi.escape(remote_user())
653 + ' from ' + cgi.escape(get_hostname(remote_host()))
655 print '<div class="editor"><form method="post" enctype="multipart/form-data" action="%s">' % relative_url(self.page_name)
656 print '<input type="hidden" name="edit" value="%s">' % (self.page_name)
657 print '<textarea wrap="off" spellcheck="true" id="editor" name="savetext" rows="17" cols="100">%s</textarea>' % (preview or self.get_raw_body())
658 print 'Or upload a file: <input type="file" name="file" value="%s" />' % file
661 <input type="submit" name="save" value="Save" />
662 <input type="submit" name="preview" value="Preview" />
663 <input type="reset" value="Reset" />
664 <input type="submit" name="cancel" value="Cancel" />
667 print "<p>" + Page('EditingTips').link_to() + "</p>"
669 print "<div class='preview'>"
670 WikiFormatter(preview).print_html()
672 send_footer(self.page_name)
674 def send_raw(self, mimetype='text/plain'):
676 body = self.get_raw_body()
677 emit_header(mimetype)
680 send_title(None, msg_text='Read access denied by ACLs', msg_type='notice')
682 def _write_file(self, data):
683 tmp_filename = self._tmp_filename()
684 open(tmp_filename, 'wb').write(data)
685 name = self._filename()
687 # Bad Bill! POSIX rename ought to replace. :-(
691 if er.errno != errno.ENOENT: raise er
692 os.rename(tmp_filename, name)
694 def save(self, newdata):
695 if not self.can_write():
696 self.msg_text = 'Write access denied by ACLs'
697 self.msg_type = 'error'
700 self._write_file(newdata)
703 # FIXME: what's the std way to perform shell quoting in python?
704 cmd = ( post_edit_hook
705 + " '" + data_dir + '/' + self.page_name
706 + "' '" + remote_user()
707 + "' '" + remote_host() + "'"
713 self.msg_text += "Post-editing hook returned %d.\n" % rc
714 self.msg_text += 'Command was: ' + cmd + '\n'
716 self.msg_text += 'Output follows:\n' + output
718 self.msg_text = 'Thank you for your contribution. Your attention to detail is appreciated.'
719 self.msg_type = 'success'
721 # Main ---------------------------------------------------------------
723 execfile("geekigeeki.conf.py")
724 form = cgi.FieldStorage()
727 handler = globals().get('handle_' + cmd)
729 handler(form[cmd].value)
732 path_info = environ.get('PATH_INFO', '')
733 if len(path_info) and path_info[0] == '/':
734 query = path_info[1:] or 'FrontPage'
736 query = environ.get('QUERY_STRING', '') or 'FrontPage'
738 if file_re.match(query):
739 if word_re.match(query):
742 from mimetypes import MimeTypes
743 type, encoding = MimeTypes().guess_type(query)
744 type = type or 'text/plain'
745 Page(query).send_raw(mimetype=type)
747 print "Status: 404 Not Found"
748 send_title(None, msg_text='Can\'t work out query: ' + query)
752 msg_text = traceback.format_exc()
754 send_guru(msg_text, "error")
756 send_title(None, msg_text=msg_text)