#! /usr/bin/python """Quick-quick implementation of WikiWikiWeb in Python """ # PikiPiki copyright (C) 1999, 2000 Martin Pool # MornMorn fork by Steinar Knutsen, 2002, based on PikiPiki version 1.62 # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2 of the License, or # (at your option) any later version. # This program is distributed in the hope that it will be useful, but # WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU # General Public License for more details. # You should have received a copy of the GNU General Public License # along with this program; if not, write to the Free Software # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 # USA # TODO: # Remove all inline HTML # Sanitized HTML generator # HTML-generator generating better HTML __version__ = '$Revision: 0.6 $'[11:-2]; import cgi, sys, string, os, re, errno, time, stat, md5, binascii, exceptions,\ urllib from cgi import log from os import path, environ from socket import gethostbyaddr from time import localtime, strftime from cStringIO import StringIO from mornconfig import * sys.stderr = open('/tmp/morn_wiki_err', 'a') class PageChanged(exceptions.Exception): def __init__(self, new_raw_page, new_checksum): self.new_raw_page = new_raw_page self.new_checksum = new_checksum def emit_header(): sys.stdout.write("Content-type: text/html\n\n") # Regular expression defining a WikiWord (but this definition # is also assumed in other places. word_re_str = r"([A-ZÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞ][a-zßàáâãäåæçèéêëìíîïðñòóôõöøùúûüýþÿ0-9]+){2,}" word_anchored_re = re.compile('^' + word_re_str + '$') command_re_str = "(search|edit|fullsearch|titlesearch)\=(.*)" # Editlog ----------------------------------------------------------- # Functions to keep track of when people have changed pages, so we can # do the recent changes page and so on. # The editlog is stored with one record per line, as tab-separated # words: page_name, host, time # TODO: Check values written in are reasonable def editlog_add(page_name, host): editlog = open(editlog_name, 'a+') try: editlog.seek(0, 2) # to end editlog.write(string.join((page_name, host, `time.time()`), "\t") + "\n") finally: editlog.close() def editlog_raw_lines(): editlog = open(editlog_name, 'r') try: return editlog.readlines() finally: editlog.close() # Formatting stuff -------------------------------------------------- def get_scriptname(): return environ.get('SCRIPT_NAME', '') def send_title(text, link=None, msg=None): sys.stdout.write("%s\n" % text) if css_url: sys.stdout.write('\n' % css_url) sys.stdout.write("") sys.stdout.write('

') if link: sys.stdout.write('%s' % (link, text)) else: sys.stdout.write(text) sys.stdout.write('

\n') if msg: sys.stdout.write("

\n") sys.stdout.write(msg) sys.stdout.write("

\n") sys.stdout.write('\n

\n') def link_tag(params, text=None, ss_class=None): if text is None: text = params # default if ss_class: classattr = 'class="%s" ' % ss_class else: classattr = '' return '%s' % (classattr, get_scriptname(), params, text) def quote_html(raw_text): quoted = string.replace(raw_text, '&', '&') quoted = string.replace(quoted, '<', '<') quoted = string.replace(quoted, '>', '>') return quoted # Search --------------------------------------------------- def do_fullsearch(needle): send_title('Full text search for "%s"' % (needle)) needle_re = re.compile(needle, re.IGNORECASE) hits = [] all_pages = page_list() for page_name in all_pages: body = Page(page_name).get_raw_body() count = len(needle_re.findall(body)) if count: hits.append((count, page_name)) # The default comparison for tuples compares elements in order, # so this sorts by number of hits hits.sort() hits.reverse() sys.stdout.write("

' + Page(page_name).link_to()) sys.stdout.write(' . . . . ' + `count`) sys.stdout.write([' match\n', ' matches\n'][count <> 1]) sys.stdout.write("

\n") print_search_stats(len(hits), len(all_pages)) print_footer("", 0) def do_titlesearch(needle): # TODO: check needle is legal -- but probably we can just accept any # RE send_title("Title search for \"" + needle + '"') needle_re = re.compile(needle, re.IGNORECASE) all_pages = page_list() hits = filter(needle_re.search, all_pages) sys.stdout.write("

%s\n' % Page(filename).link_to()) sys.stdout.write("

\n") print_search_stats(len(hits), len(all_pages)) print_footer("", 0) def print_search_stats(hits, searched): sys.stdout.write("

%d hits " % hits) sys.stdout.write(" out of %d pages searched." % searched) def do_edit(pagename): Page(pagename).send_editor() def do_preview(dummypagename): global form pagename = form["savepage"].value send_title('Preview ' + urllib.unquote(pagename), None, None) raw_body = form['savetext'].value PageFormatter(raw_body).print_html() sys.stdout.write('

' % (get_scriptname())) sys.stdout.write('' % pagename) md5hash = form['checksum'].value sys.stdout.write("""""" % quote_html(raw_body)) sys.stdout.write('' % md5hash) sys.stdout.write("
") sys.stdout.write(Page('SilentEdit').link_to()) sys.stdout.write("""
""") sys.stdout.write("

" + Page('EditingTips').link_to()) print_footer(pagename, 0) def do_savepage(pagename): global form pg = Page(urllib.unquote(pagename)) checksum = binascii.unhexlify(form['checksum'].value) try: if form.has_key("silentedit"): pg.save_text(form['savetext'].value, checksum, 1) else: pg.save_text(form['savetext'].value, checksum) except PageChanged, exception_instance: msg = "This page has changed while you were working." \ " Please merge your changes with the changes with the current" \ " version. Your version is repeated below, the current version" \ " is shown in the editor area.\n

\n%s\n

" % \ quote_html(form['savetext'].value) pg.send_editor(msg) return msg = "Thankyou for your changes. Your attention to" \ " detail is appreciated." pg.send_page(msg) def make_index_key(): s = '

' links = map(lambda ch: '%s' % (ch, ch), string.lowercase) s = s + string.join(links, ' | ') s = s + '

' return s def page_list(): pagelist = map(urllib.unquote, os.listdir(text_dir)) return filter(word_anchored_re.match, pagelist) def print_footer(name, editable=1, mod_string=None, advisory_lock=None): editable = 0 sys.stdout.write('\n

') sys.stdout.write("%s %s %s\n" % (link_tag('NeverWhere', "NeverWhere"), link_tag('RecentChanges', "RecentChanges"), link_tag('TitleIndex', "TitleIndex"))) if editable: sys.stdout.write("
%s of this page" % link_tag('edit='+urllib.quote(name), 'EditText')) if mod_string: sys.stdout.write(" (last modified %s) " % mod_string) if advisory_lock: sys.stdout.write('\n
\n') sys.stdout.write(advisory_lock) sys.stdout.write("
%s by browsing, searching, or an index" % link_tag('FindPage', 'FindPage')) # ---------------------------------------------------------- # Macros def _macro_TitleSearch(): return _macro_search("titlesearch") def _macro_FullSearch(): return _macro_search("fullsearch") def _macro_search(type): if form.has_key('value'): default = form["value"].value else: default = '' return """

\n""" % (type, default) def _macro_GoTo(): return """\n""" # isindex is deprecated, but it gives the right result here def _macro_WordIndex(): s = make_index_key() pages = list(page_list()) map = {} word_re = re.compile('[A-ZÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞ][a-zßàáâãäåæçèéêëìíîïðñòóôõöøùúûüýþÿ0-9]+') for name in pages: for word in word_re.findall(name): try: map[word].append(name) except KeyError: map[word] = [name] all_words = map.keys() all_words.sort() last_letter = None for word in all_words: letter = string.lower(word[0]) if letter <> last_letter: s = s + '

%s

\n' % (letter, letter) last_letter = letter s = s + '%s

' + Page(name).link_to() + '\n' s = s + '\n

\n' return s def _macro_TitleIndex(): s = make_index_key() pages = list(page_list()) pages.sort() current_letter = None for name in pages: letter = string.lower(name[0]) if letter <> current_letter: s = s + '

%s

' % (letter, letter) current_letter = letter else: s = s + '
' s = s + Page(name).link_to() return s def _macro_RecentChanges(): lines = editlog_raw_lines() lines.reverse() ratchet_day = None done_words = {} buf = StringIO() for line in lines: page_name, addr, ed_time = string.split(line, '\t') # year, month, day, DoW time_tuple = localtime(float(ed_time)) day = tuple(time_tuple[0:3]) if day <> ratchet_day: buf.write('

%s

' % strftime(date_fmt, time_tuple)) ratchet_day = day if done_words.has_key(page_name): continue done_words[page_name] = 1 buf.write(Page(page_name).link_to()) if show_hosts: buf.write(' . . . . ') try: buf.write(gethostbyaddr(addr)[0]) except: buf.write(addr) if changed_time_fmt: buf.write(time.strftime(changed_time_fmt, time_tuple)) buf.write('
\n') return buf.getvalue() # ---------------------------------------------------------- class PageFormatter: """Object that turns Wiki markup into HTML. All formatting commands can be parsed one line at a time, though some state is carried over between lines. """ def __init__(self, raw): self.raw = raw self.is_h = self.is_em = self.is_b = 0 self.list_indents = [] self.in_pre = 0 def _emph_repl(self, word): if len(word) == 3: self.is_b = not self.is_b return ['', ''][self.is_b] else: self.is_em = not self.is_em return ['', ''][self.is_em] def _head_repl(self, word): self.is_h = not self.is_h lvl = len(string.strip(word)) return ['' % lvl, '' % lvl][self.is_h] def _rule_repl(self, word): s = self._undent() if len(word) <= 4: s = s + "\n
\n" else: s = s + "\n
\n" % (len(word) - 2 ) return s def _word_repl(self, word): return Page(word).link_to() def _img_repl(self, word): return '' % (word, word) def _url_repl(self, word): return '%s' % (word, word) def _email_repl(self, word): return '%s' % (word, word) def _ent_repl(self, s): return {'&': '&', '<': '<', '>': '>'}[s] def _li_repl(self, match): return '
' def _pre_repl(self, word): if word == '{{{' and not self.in_pre: self.in_pre = 1 return '
' elif self.in_pre: self.in_pre = 0 return '
' else: return '' def _macro_repl(self, word): macro_name = word[2:-2] # TODO: Somehow get the default value into the search field return apply(globals()['_macro_' + macro_name], ()) def _indent_level(self): return len(self.list_indents) and self.list_indents[-1] def _indent_to(self, new_level): s = '' while self._indent_level() > new_level: del(self.list_indents[-1]) s = s + '\n' while self._indent_level() < new_level: self.list_indents.append(new_level) s = s + '
\n' return s def _undent(self): res = '
' * len(self.list_indents) self.list_indents = [] return res def replace(self, match): for type, hit in match.groupdict().items(): if hit: if self.in_pre and type not in ("pre", "ent", "word", "url"): return hit else: return apply(getattr(self, '_' + type + '_repl'), (hit,)) else: raise "Can't handle match " + `match` def print_html(self): # For each line, we scan through looking for magic # strings, outputting verbatim any intervening text scan_re = re.compile( r"(?:(?P'{2,3})" + r"|(?P^\s*={1,6}\s+|\s+={1,6}\s*$)" + r"|(?P[<>&])" + r"|(?P(?:[A-ZÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞ][a-zßàáâãäåæçèéêëìíîïðñòóôõöøùúûüýþÿ0-9]+){2,})" + r"|(?P-{4,})" + r"|(?P(http|ftp)\:[^\s'\"]+\S\.(jpeg|jpg|gif|png))" + r"|(?P(http|ftp|nntp|news|mailto)\:[^\s'\"]+\S)" + r"|(?P[-\w._+]+\@[\w.-]+)" + r"|(?P
^\s+\*)" + r"|(?P
(\{\{\{|\}\}\}))" + r"|(?P\[\[(TitleSearch|FullSearch|WordIndex" + r"|TitleIndex|RecentChanges|GoTo)\]\])" + r")") blank_re = re.compile("^\s*$") bullet_re = re.compile("^\s+\*") indent_re = re.compile("^\s*") eol_re = re.compile(r'\r?\n') raw = string.expandtabs(self.raw) for line in eol_re.split(raw): if not self.in_pre: # XXX: Should we check these conditions in this order? if blank_re.match(line): sys.stdout.write('
\n') continue indent = indent_re.match(line) sys.stdout.write(self._indent_to(len(indent.group(0)))) sys.stdout.write(re.sub(scan_re, self.replace, line)) sys.stdout.write("\n") if self.in_pre: sys.stdout.write('
\n') sys.stdout.write(self._undent()) # ---------------------------------------------------------- class Page: def __init__(self, page_name): self.page_name = page_name def _text_filename(self): return path.join(text_dir, urllib.quote(self.page_name)) def _lock_filename(self): return path.join(text_dir, "!" + urllib.quote(self.page_name)) def _tmp_filename(self): return path.join(text_dir, ('#' + urllib.quote(self.page_name) + '.' + `os.getpid()` + '#')) def exists(self): try: os.stat(self._text_filename()) return 1 except OSError, er: if er.errno == errno.ENOENT: return 0 else: raise er def lock_exists(self): try: os.stat(self._lock_filename()) return 1 except OSError, er: if er.errno == errno.ENOENT: return 0 else: raise er def link_to(self): word = self.page_name if self.exists(): return link_tag(urllib.quote(word), word) else: if nonexist_qm: return link_tag(urllib.quote(word), '?', 'nonexistent') + word else: return link_tag(urllib.quote(word), word, 'nonexistent') def get_raw_body(self): try: return open(self._text_filename(), 'r').read() except IOError, er: if er.errno == errno.ENOENT: # just doesn't exist, use default return 'Describe %s here.' % self.page_name else: raise er def send_page(self, msg=None): link = get_scriptname() + '?fullsearch=' + urllib.quote(self.page_name) send_title(self.page_name, link, msg) PageFormatter(self.get_raw_body()).print_html() print_footer(self.page_name, 1, self._last_modified(), self._advisory_lock()) def _last_modified(self): if not self.exists(): return None modtime = localtime(os.stat(self._text_filename())[stat.ST_MTIME]) return strftime(datetime_fmt, modtime) def _advisory_lock(self): if not self.lock_exists(): return None checkout = localtime(os.stat(self._lock_filename())[stat.ST_MTIME]) who = open(self._lock_filename()).read(-1) try: who = gethostbyaddr(who)[0] except: pass return "(%s editing started from %s)" % (strftime(datetime_fmt, checkout), who) def activate_advisory_lock(self): remote_name = environ.get('REMOTE_ADDR', '') tmp = open(self._lock_filename(), "wb") try: tmp.write(remote_name) finally: tmp.close() def send_editor(self, msg = None): self.activate_advisory_lock() send_title('Edit ' + self.page_name, None, msg) sys.stdout.write('
' % (get_scriptname())) sys.stdout.write('' % urllib.quote((self.page_name))) raw_body = self.get_raw_body() md5hash = binascii.hexlify(md5.new(raw_body).digest()) sys.stdout.write("""%s""" % quote_html(raw_body)) sys.stdout.write('' % md5hash) sys.stdout.write("
") sys.stdout.write(Page('SilentEdit').link_to()) sys.stdout.write("""
""") sys.stdout.write("
" + Page('EditingTips').link_to()) print_footer(self.page_name, 0) def _write_file(self, text): tmp_filename = self._tmp_filename() open(tmp_filename, 'w').write(text) text = self._text_filename() os.rename(tmp_filename, text) try: os.remove(self._lock_filename()) except: pass def save_text(self, newtext, checksum, silent_edit=None): current_raw = self.get_raw_body() current_checksum = md5.new(current_raw).digest() if current_checksum == checksum: # self._write_file(newtext) if not silent_edit: remote_name = environ.get('REMOTE_ADDR', '') editlog_add(self.page_name, remote_name) else: raise PageChanged(current_raw, current_checksum) emit_header() try: form = cgi.FieldStorage() handlers = { 'fullsearch': do_fullsearch, 'titlesearch': do_titlesearch, 'edit': do_edit, 'savepage': do_savepage, 'preview': do_preview } # I know, this is plain sick, but hopefully nobody forced you into # reading this source. The point of the following is that I want to # control the sequence for the testing of commands. This was introduced # when I added preview. The only important thing right now is testing for # preview before savepage. handler_keys = ( 'preview', 'savepage', 'edit', 'titlesearch', 'fullsearch' ) for cmd in handler_keys: if form.has_key(cmd): apply(handlers[cmd], (form[cmd].value,)) break else: path_info = environ.get('PATH_INFO', '') if form.has_key('goto'): query = form['goto'].value elif len(path_info) and path_info[0] == '/': query = path_info[1:] or 'NeverWhere' else: query = environ.get('QUERY_STRING', '') or 'NeverWhere' query = urllib.unquote(query) word_match = re.match(word_re_str, query) if word_match: word = word_match.group(0) Page(word).send_page() else: sys.stdout.write("
Can't work out query \"
" + query + "
\"") except: cgi.print_exception() sys.stdout.flush()