X-Git-Url: https://codewiz.org/gitweb?p=geekigeeki.git;a=blobdiff_plain;f=geekigeeki.py;h=9d443fc6997d5cdf0c39ee774d3c91b3ade82804;hp=b00838ae923d45ac5320579096b7d6f1cebb1b8d;hb=151df764ca017e25015fab92838fc00af426179f;hpb=2bb00b8d6473a767cba8616e18d587f98c773354 diff --git a/geekigeeki.py b/geekigeeki.py index b00838a..9d443fc 100755 --- a/geekigeeki.py +++ b/geekigeeki.py @@ -3,7 +3,7 @@ # # Copyright (C) 1999, 2000 Martin Pool # Copyright (C) 2002 Gerardo Poggiali -# Copyright (C) 2007, 2008, 2009 Bernie Innocenti +# Copyright (C) 2007, 2008, 2009, 2010, 2011 Bernie Innocenti # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as @@ -21,15 +21,14 @@ title_done = False import cgi, sys, os, re, errno, stat, glob image_ext = 'png|gif|jpg|jpeg|bmp|ico' -video_ext = "ogg|ogv|oga" # Not supported by Firefox 3.5: mkv|mpg|mpeg|mp4|avi|asf|flv|wmv|qt +video_ext = "ogg|ogv|oga|webm" # Not supported by Firefox 3.5: mkv|mpg|mpeg|mp4|avi|asf|flv|wmv|qt image_re = re.compile(r".*\.(" + image_ext + "|" + video_ext + ")$", re.IGNORECASE) video_re = re.compile(r".*\.(" + video_ext + ")$", re.IGNORECASE) # FIXME: we accept stuff like foo/../bar and we shouldn't -file_re = re.compile(r"([A-Za-z0-9_\-][A-Za-z0-9_\.\-/]*)$") +file_re = re.compile(r"([A-Za-z0-9_\-][A-Za-z0-9_\.\-/ ]*)$") url_re = re.compile(r"[a-z]{3,8}://[^\s'\"]+\S$") ext_re = re.compile(r"\.([^\./]+)$") -# CGI stuff --------------------------------------------------------- def config_get(key, default=None): return globals().get(key, default) @@ -106,11 +105,15 @@ def url_args(kvargs): return '?' + '&'.join(argv) return '' -# Formatting stuff -------------------------------------------------- def emit_header(mtime=None, mime_type="text/html"): if mtime: + # Prevent caching when the wiki engine gets updated + mtime = max(mtime, os.stat(__file__).st_mtime) print("Last-Modified: " + strftime("%a, %d %b %Y %H:%M:%S GMT", gmtime(mtime))) - print("Content-type: " + mime_type + "; charset=utf-8\n") + else: + print("Cache-Control: must-revalidate, max-age=0") + print("Content-type: " + mime_type + "; charset=utf-8") + print('') def send_guru(msg_text, msg_type): if not msg_text: return @@ -123,83 +126,15 @@ def send_guru(msg_text, msg_type): print('' \ % relative_url('sys/GuruMeditation.js')) -def send_title(name, text="Limbo", msg_text=None, msg_type='error', writable=False, mtime=None, navbar="on"): - global title_done - if title_done: return - - # HEAD - emit_header(mtime) - print('\n') - print("%s: %s" % (config_get('site_name', "Unconfigured Wiki"), text)) - print(' ') - if not name: - print(' ') - - for http_equiv, content in config_get('meta_urls', {}): - print(' ' % (http_equiv, relative_url(content))) - - for link in config_get('link_urls', {}): - rel, href = link - print(' ' % (rel, relative_url(href))) - - editable = name and writable and is_privileged() - if editable: - print(' ' \ - % relative_url('?a=edit&q=' + name, privileged=True)) - - history = config_get('history_url') - if history is not None: - print(' ' \ - % relative_url(history + '?a=rss')) - - print('') - - # BODY - if editable: - print('') - else: - print('') - - title_done = True - send_guru(msg_text, msg_type) - - if navbar != "on": - return - - # NAVBAR - print('') - -def send_httperror(status="403 Not Found", query=""): +def send_httperror(status="404 Not Found", query="", trace=False): print("Status: %s" % status) - send_title(None, msg_text=("%s: on query '%s'" % (status, query))) - send_footer() + msg_text = "%s: on query '%s'" % (status, query) + if trace: + import traceback + msg_text += '\n\n' + traceback.format_exc() + page = Page() + page.send_title(msg_text=msg_text) + page.send_footer() def link_tag(dest, text=None, privileged=False, **kvargs): if text is None: @@ -217,9 +152,9 @@ def link_tag(dest, text=None, privileged=False, **kvargs): text = config_get('nonexist_pfx', '') + text link_class = 'nonexistent' - # Prevent crawlers from following links potentially added by spammers or to generated pages + # Prevent crawlers from following links potentially added by spammers and to autogenerated pages nofollow = '' - if link_class == 'external' or link_class == 'navlink': + if link_class in ('external', 'navlink', 'nonexistent'): nofollow = 'rel="nofollow" ' return '%s' % (link_class, nofollow, relative_url(dest, privileged=privileged), text) @@ -241,20 +176,18 @@ def link_inline(name, descr=None, kvargs={}): def link_inline_glob(pattern, descr=None, kvargs={}): if not url_re.match(pattern) and bool(set(pattern) & set('?*[')): s = '' - for name in glob.glob(pattern): + for name in sorted(glob.glob(pattern), reverse=bool(int(kvargs.get('reverse', '0'))) ): s += link_inline(name, descr, kvargs) return s else: return link_inline(pattern, descr, kvargs) -# Search --------------------------------------------------- - -def print_search_stats(hits, searched): - print("

%d hits out of %d pages searched.

" % (hits, searched)) +def search_stats(hits, searched): + return "%d hits out of %d pages searched.\n" % (hits, searched) def handle_fullsearch(query, form): needle = form['q'].value - send_title(None, 'Full text search for "' + needle + '"') + Page().send_title(text='Full text search for "' + needle + '"') needle_re = re.compile(needle, re.IGNORECASE) hits = [] @@ -265,59 +198,50 @@ def handle_fullsearch(query, form): if count: hits.append((count, page_name)) - # The default comparison for tuples compares elements in order, - # so this sorts by number of hits + # The default comparison for tuples compares elements in order, so this sorts by number of hits hits.sort() hits.reverse() - print("
    ") + out = '' for (count, page_name) in hits: - print('

  • ' + link_tag(page_name)) - print(' . . . . ' + `count`) - print(['match', 'matches'][count != 1]) - print('

    • ') - print("
") + out += ' * [[' + page_name + ']] . . . ' + str(count) + ' ' + ['match', 'matches'][count != 1] + '\n' - print_search_stats(len(hits), len(all_pages)) + out += search_stats(len(hits), len(all_pages)) + WikiFormatter(out).print_html() def handle_titlesearch(query, form): needle = form['q'].value - send_title(None, 'Title search for "' + needle + '"') + Page().send_title(text='Title search for "' + needle + '"') needle_re = re.compile(needle, re.IGNORECASE) all_pages = page_list() hits = list(filter(needle_re.search, all_pages)) - print("
    ") + out = '' for filename in hits: - print('

  • ' + link_tag(filename) + "

    • ") - print("
") + out += ' * [[' + filename + ']]\n' - print_search_stats(len(hits), len(all_pages)) + out += search_stats(len(hits), len(all_pages)) + WikiFormatter(out).print_html() def handle_raw(pagename, form): - if not file_re.match(pagename): - send_httperror("403 Forbidden", pagename) - return - Page(pagename).send_raw() -def handle_edit(pagename, form): - if not file_re.match(pagename): - send_httperror("403 Forbidden", pagename) - return +def handle_atom(pagename, form): + Page(pagename).send_atom() - pg = Page(form['q'].value) +def handle_edit(pagename, form): + pg = Page(pagename) if 'save' in form: if form['file'].value: pg.save(form['file'].file.read(), form['changelog'].value) else: pg.save(form['savetext'].value.replace('\r\n', '\n'), form['changelog'].value) - pg.format() + pg.send() elif 'cancel' in form: pg.msg_text = 'Editing canceled' pg.msg_type = 'notice' - pg.format() + pg.send() else: # preview or edit text = None if 'preview' in form: @@ -325,16 +249,13 @@ def handle_edit(pagename, form): pg.send_editor(text) def handle_get(pagename, form): - if file_re.match(pagename): - # FIMXE: this is all bullshit, MimeTypes bases its guess on the extension! - from mimetypes import MimeTypes - mimetype, encoding = MimeTypes().guess_type(pagename) - if mimetype: - Page(pagename).send_raw(mimetype=mimetype, args=form) - else: - Page(pagename).format() - else: - send_httperror("403 Forbidden", pagename) + if not ext_re.search(pagename): # FIXME: no extension guesses a wiki page + Page(pagename).send() + else: + # FIMXE: this is all bullshit, MimeTypes bases its guess on the extension! + from mimetypes import MimeTypes + mimetype, encoding = MimeTypes().guess_type(pagename) + Page(pagename).send_raw(mimetype=mimetype, args=form) # Used by sys/macros/WordIndex and sys/macros/TitleIndex def make_index_key(): @@ -347,17 +268,6 @@ def page_list(dirname=None, search_re=None): search_re = re.compile(r"^\b((([A-Z][a-z0-9]+){2,}/)*([A-Z][a-z0-9]+){2,})\b$") return sorted(filter(search_re.match, os.listdir(dirname or '.'))) -def send_footer(mtime=None, footer="sys/footer"): - if config_get('debug_cgi', False): - cgi.print_arguments() - cgi.print_form(form) - cgi.print_environ() - if footer != "off": - link_inline(footer, kvargs = { - 'LAST_MODIFIED': strftime(config_get('datetime_fmt', '%a %d %b %Y %I:%M %p'), localtime(mtime)) - }) - print("") - def _macro_ELAPSED_TIME(*args, **kvargs): return "%03f" % (clock() - start_time) @@ -365,11 +275,7 @@ def _macro_VERSION(*args, **kvargs): return __version__ class WikiFormatter: - """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. - """ + """Object that turns Wiki markup into HTML.""" def __init__(self, raw, kvargs=None): self.raw = raw self.kvargs = kvargs or {} @@ -400,13 +306,13 @@ class WikiFormatter: return '—' def _tit_repl(self, word): + link = permalink(self.line) if self.h_level: - result = '

\n' % self.h_level + result = '¶

\n' % (link, self.h_level) self.h_level = 0 else: self.h_level = len(word) - 1 - link = permalink(self.line) - result = '\n

¶ ' % (self.h_level, link, link) + result = '\n

' % (self.h_level, link) return result def _br_repl(self, word): @@ -417,14 +323,19 @@ class WikiFormatter: def _macro_repl(self, word): try: - args, kvargs = parse_args(word) + args, macro_kvargs = parse_args(word) + # Is this a parameter given to the current page? if args[0] in self.kvargs: return self.kvargs[args[0]] + # Is this an internal macro? macro = globals().get('_macro_' + args[0]) if not macro: + # Can we load (and cache) an external macro? exec(open("sys/macros/" + args[0] + ".py").read(), globals()) macro = globals().get('_macro_' + args[0]) - return macro(*args, **kvargs) + # Invoke macro passing both macro args augmented by page args + macro_kvargs.update(self.kvargs) + return macro(*args, **macro_kvargs) except Exception, e: msg = cgi.escape(word) + ": " + cgi.escape(str(e)) if not self.in_html: @@ -560,12 +471,12 @@ class WikiFormatter: | (?P --) # Links - | (?P \<\<([^\s\|\>]+)(?:\s*\|\s*([^\>]+)|)\>\>) - | (?P \[\[([^\s\|]+)(?:\s*\|\s*([^\]]+)|)\]\]) + | (?P \<\<[^\>]+\>\>) + | (?P \[\[[^\]]+\]\]) # Inline HTML - | (?P <(br|hr|div|span|form|iframe|input|textarea|a|img|h[1-5])\b ) - | (?P ( /\s*> | ) ) + | (?P <(br|hr|small|div|span|form|iframe|input|textarea|a|img|h[1-5])\b ) + | (?P ( /\s*> | ) ) | (?P [<>&] ) # Auto links (LEGACY) @@ -577,14 +488,12 @@ class WikiFormatter: # Lists, divs, spans and inline objects | (?P
  • ^\s+[\*\#]\s+) | (?P
       \{\{\{|\s*\}\}\})
-            | (?P   \{\{([^\s\|]+)(?:\s*\|\s*([^\]]+)|)\}\})
+            | (?P   \{\{[^\}]+\}\})
 
             # Tables
             | (?P    ^\s*\|\|(=|)\s*)
             | (?P   \s*\|\|(=|)\s*$)
             | (?P    \s*\|\|(=|)\s*)
-
-            # TODO: highlight search words (look at referrer)
           )""", re.VERBOSE)
         pre_re = re.compile("""(?:
               (?P
    \s*\}\}\})
@@ -623,11 +532,18 @@ class WikiFormatter:
         print(self._indent_to(0))
         print('
    ')
 
+class HttpException(Exception):
+    def __init__(self, error, query):
+        self.error = error
+        self.query = query
+
 class Page:
-    def __init__(self, page_name):
+    def __init__(self, page_name="Limbo"):
         self.page_name = page_name.rstrip('/');
         self.msg_text = ''
         self.msg_type = 'error'
+        if not file_re.match(self.page_name):
+            raise HttpException("403 Forbidden", self.page_name)
 
     def split_title(self):
         # look for the end of words and the start of a new word and insert a space there
@@ -652,14 +568,12 @@ class Page:
             return True
         return False
 
-    def get_raw_body(self, default=None):
+    def get_raw_body(self):
         try:
             return open(self._filename(), 'rb').read()
         except IOError, err:
             if err.errno == errno.ENOENT:
-                if default is None:
-                    default = '//[[?a=edit&q=%s|Describe %s]]//' % (self.page_name, self.page_name)
-                return default
+                return ''
             if err.errno == errno.EISDIR:
                 return self.format_dir()
             raise err
@@ -724,33 +638,128 @@ class Page:
     def can_read(self):
         return self.can("read", True)
 
+    def send_title(self, name=None, text="Limbo", msg_text=None, msg_type='error'):
+        global title_done
+        if title_done: return
+
+        # HEAD
+        emit_header(name and self._mtime())
+        print('\n')
+        print("%s: %s" % (config_get('site_name', "Unconfigured Wiki"), text))
+        print(' ')
+        if not name:
+            print(' ')
+
+        for http_equiv, content in config_get('meta_urls', {}):
+            print(' ' % (http_equiv, relative_url(content)))
+
+        for link in config_get('link_urls', {}):
+            rel, href = link
+            print(' ' % (rel, relative_url(href)))
+
+        editable = name and self.can_write() and is_privileged()
+        if editable:
+            print(' ' \
+                % relative_url(name + '?a=edit', privileged=True))
+
+        history = config_get('history_url')
+        if history is not None:
+            print(' ' \
+                % relative_url(history + '?a=rss'))
+
+        print('')
+
+        # BODY
+        if editable:
+            print('')
+        else:
+            print('')
+
+        title_done = True
+        send_guru(msg_text, msg_type)
+
+        if self.pragma("navbar", "on") != "on":
+            return
+
+        # NAVBAR
+        print('')
+
+    def send_footer(self):
+        if config_get('debug_cgi', False):
+            cgi.print_arguments()
+            cgi.print_form(form)
+            cgi.print_environ()
+        footer = self.pragma("footer", "sys/footer")
+        if footer != "off":
+            link_inline(footer, kvargs = {
+                'LAST_MODIFIED': strftime(config_get('datetime_fmt', '%Y-%m-%dT%I:%M:%S%p'), localtime(self._mtime()))
+            })
+        print('')
+
     def send_naked(self, kvargs=None):
         if self.can_read():
-            WikiFormatter(self.get_raw_body(), kvargs).print_html()
+            body = self.get_raw_body()
+            if not body:
+                body = "//[[%s?a=edit|Describe %s]]//" % (self.page_name, self.page_name)
+            WikiFormatter(body, kvargs).print_html()
         else:
-            send_guru("Read access denied by ACLs", "notice")
+            send_guru('Read access denied by ACLs', 'notice')
 
-    def format(self):
+    def send(self):
         #css foo.css
         value = self.pragma("css", None)
         if value:
             global link_urls
             link_urls += [ [ "stylesheet", value ] ]
 
-        send_title(self.page_name, self.split_title(),
-            msg_text=self.msg_text, msg_type=self.msg_type, writable=self.can_write(), mtime=self._mtime(),
-            navbar=self.pragma("navbar", "on"))
+        self.send_title(name=self.page_name, text=self.split_title(), msg_text=self.msg_text, msg_type=self.msg_type)
         self.send_naked()
-        send_footer(mtime=self._mtime(), footer=self.pragma("footer", "sys/footer"))
+        self.send_footer()
+
+    def send_atom(self):
+        emit_header(self._mtime(), 'application/atom+xml')
+        self.in_html = True
+        link_inline("sys/atom_header", kvargs = {
+            'LAST_MODIFIED': strftime(config_get('datetime_fmt', '%a, %d %b %Y %I:%M:%S %p'), localtime(self._mtime()))
+        })
+        self.in_html = False
+        self.send_naked()
+        self.in_html = True
+        link_inline("sys/atom_footer")
+        self.in_html = False
 
     def send_editor(self, preview=None):
-        send_title(None, 'Edit ' + self.split_title(), msg_text=self.msg_text, msg_type=self.msg_type)
+        self.send_title(text='Edit ' + self.split_title(), msg_text=self.msg_text, msg_type=self.msg_type)
         if not self.can_write():
             send_guru("Write access denied by ACLs", "error")
             return
 
         if preview is None:
-            preview = self.get_raw_body(default='')
+            preview = self.get_raw_body()
 
         link_inline("sys/EditPage", kvargs = {
             'EDIT_BODY': cgi.escape(preview),
@@ -761,11 +770,11 @@ class Page:
             print("
    ")
             WikiFormatter(preview).print_html()
             print("
    ")
-        send_footer()
+        self.send_footer()
 
     def send_raw(self, mimetype='text/plain', args=[]):
         if not self.can_read():
-            send_title(None, msg_text='Read access denied by ACLs', msg_type='notice', mtime=self._mtime())
+            self.send_title(msg_text='Read access denied by ACLs', msg_type='notice')
             return
 
         emit_header(self._mtime(), mimetype)
@@ -831,13 +840,9 @@ try:
     else:
         send_httperror("403 Forbidden", query_string())
 
+except HttpException, e:
+    send_httperror(e.error, query=e.query)
 except Exception:
-    import traceback
-    msg_text = traceback.format_exc()
-    if title_done:
-        send_guru(msg_text, "error")
-    else:
-        send_title(None, msg_text=msg_text)
-    send_footer()
+    send_httperror("500 Internal Server Error", query=query_string(), trace=True)
 
 sys.stdout.flush()