Moved html parsing from phd_pp.py to reindex_blog.py.

Make all URLs absolute for xml feeds.

git-svn-id: file:///home/phd/archive/SVN/phdru.name/scripts@112 7bb0bf08-9e0d-0410-b083-99cee3bf18b8

diff --git a/phd_pp.py b/phd_pp.py
index 95e90db415b1901106c061d90b3111c11793cfab..f9c9a67dca9712272ba8141bf9e7e9d57ea43690 100644 (file)
--- a/phd_pp.py
+++ b/phd_pp.py
@@ -1,7 +1,5 @@
 import os, re, time, urllib
-from HTMLParser import HTMLParseError
 from Cheetah.Template import Template
-from m_lib.net.www.html import HTMLParser as _HTMLParser
 
 
 url_re = r"(((https?|ftp|gopher|telnet)://|(mailto|file|news|about|ed2k|irc|sip|magnet):)[^' \t<>\"]+|(www|web|w3)[A-Za-z0-9_-]*\.[A-Za-z0-9._-]+\.[^' \t<>\"]+)[A-Za-z0-9/]"
@@ -38,6 +36,7 @@ class phd_pp(Template):
          return "%s, %s" % (start_year, this_year)
       return "%s-%s" % (start_year, this_year)
 
+
    def body(self):
       if hasattr(self, "body_html"):
          body = self.body_html()
@@ -45,7 +44,6 @@ class phd_pp(Template):
          body = self.text2html()
       if hasattr(self, "body_rst"):
          body = self.rst2html()
-      self.Body = body
       return body
 
    def text2html(self):
@@ -85,20 +83,6 @@ class phd_pp(Template):
       parts = [part for part in (title, subtitle, body) if part]
       return "\n\n".join(parts)
 
-   def get_first_p(self):
-      parser = HTMLParser()
-
-      try:
-         parser.feed(self.body())
-      except (HTMLParseError, HTMLHeadDone):
-         pass
-
-      try:
-         parser.close()
-      except (HTMLParseError, HTMLHeadDone):
-         pass
-
-      return parser.first_p
 
    def img_thumbnail_800_1024(self, img_name):
       return """\
@@ -127,18 +111,3 @@ class phd_pp(Template):
 
 def quote_string(s, to_encoding="utf-8", ext_safe=''):
    return urllib.quote(unicode(s, "koi8-r").encode(to_encoding), '/' + ext_safe)
-
-
-class HTMLHeadDone(Exception): pass
-
-class HTMLParser(_HTMLParser):
-   def __init__(self, charset=None):
-      _HTMLParser.__init__(self)
-      self.first_p = None
-
-   def start_p(self, attrs):
-      self.accumulator = '<p>'
-
-   def end_p(self):
-      self.first_p = self.accumulator + '</p>'
-      raise HTMLHeadDone()

diff --git a/reindex_blog.py b/reindex_blog.py
index fd82c1839111ae59ddcd7a7e86eace33d7b9a715..beba7c12823ae151a8f5de6619c9776758b9d713 100755 (executable)
--- a/reindex_blog.py
+++ b/reindex_blog.py
@@ -39,9 +39,8 @@ else:
 blog = {}
 years = {}
 
-# excerpts nd bodies are dictionaries mapping file => excerpt/body
+# bodies is a dictionary mapping file => body
 
-excerpts = {}
 bodies = {}
 
 # Walk the directory recursively
@@ -56,7 +55,7 @@ for dirpath, dirs, files in os.walk(blog_root):
       template = Template(file=fullpath)
       title_parts = template.Title.split()
       title = ' '.join(title_parts[6:])
-      lead = getattr(template, "Lead", None)
+      lead = template.Lead
 
       tags = template.Tag
       if isinstance(tags, basestring):
@@ -84,8 +83,7 @@ for dirpath, dirs, files in os.walk(blog_root):
 
          file = file[:-len("tmpl")] + "html"
          key = (year, month, day, file)
-         excerpts[key] = template.get_first_p()
-         bodies[key] = template.Body
+         bodies[key] = template.body()
 
 # Need to save the blog?
 if blog <> old_blog:
@@ -220,15 +218,11 @@ def write_template(level, year, month, day, titles, tags=None):
          else:
             new_text.append('\n<h2>%s %s</h2>' % (day, months_names_ru[int(month)]))
          save_date = year, month, day
-      if lead:
-         lead = lead + ' '
-      else:
-         lead = ''
       new_text.append('''
 <p class="head">
    %s<a href="%s">%s</a>.
 </p>
-''' % (lead, href, title))
+''' % (lead+' ' if lead else '', href, title))
 
    if level == 0:
       new_text.append("""
@@ -407,12 +401,8 @@ for i, (count, tag, links) in enumerate(all_tags):
 
    count = 0
    for year, month, day, filename, title, lead in reversed(links):
-      if lead:
-         lead = lead + ' '
-      else:
-         lead = ''
       link = "../%s/%s/%s/%s" % (year, month, day, filename)
-      item_text = """<li><a href="%s">%s/%s/%s: %s%s</a></li>""" % (link, year, month, day, lead, title)
+      item_text = """<li><a href="%s">%s/%s/%s: %s%s</a></li>""" % (link, year, month, day, lead+' ' if lead else '', title)
 
       count += 1
       if count <= 5:
@@ -437,6 +427,76 @@ $phd_pp_ru.respond(self)
 write_if_changed(os.path.join(blog_root, "tags", "index.tmpl"), ''.join(new_text))
 
 
+from HTMLParser import HTMLParseError
+import cgi
+from urlparse import urljoin
+from m_lib.net.www.html import HTMLParser as _HTMLParser
+
+class HTMLDone(Exception): pass
+
+
+class FirstPHTMLParser(_HTMLParser):
+   def __init__(self):
+      _HTMLParser.__init__(self)
+      self.first_p = None
+
+   def start_p(self, attrs):
+      self.accumulator = '<p>'
+
+   def end_p(self):
+      self.first_p = self.accumulator + '</p>'
+      raise HTMLDone()
+
+def get_first_p(body):
+   parser = FirstPHTMLParser()
+
+   try:
+      parser.feed(body)
+   except (HTMLParseError, HTMLDone):
+      pass
+
+   try:
+      parser.close()
+   except (HTMLParseError, HTMLDone):
+      pass
+
+   return parser.first_p
+
+
+class AbsURLHTMLParser(_HTMLParser):
+   def __init__(self, base):
+      _HTMLParser.__init__(self)
+      self.base = base
+
+   def start_a(self, attrs):
+      self.accumulator += '<a'
+      for attrname, value in attrs:
+         value = cgi.escape(value, True)
+         if attrname == 'href':
+            self.accumulator += ' href="%s"' % urljoin(self.base, value)
+         else:
+            self.accumulator += ' %s="%s"' % (attrname, value)
+      self.accumulator += '>'
+
+   def end_a(self):
+      self.accumulator += '</a>'
+
+def absolute_urls(body, base):
+   parser = AbsURLHTMLParser(base)
+
+   try:
+      parser.feed(body)
+   except HTMLParseError:
+      pass
+
+   try:
+      parser.close()
+   except HTMLParseError:
+      pass
+
+   return parser.accumulator
+
+
 from atom_10 import atom_10
 from rss_20 import rss_20
 from news import NewsItem
@@ -449,20 +509,18 @@ else:
 items = []
 for item in tuple(reversed(all_titles_tags))[:10]:
    year, month, day, file, title, lead, tags = item
-   if lead:
-      lead = lead + ' '
-   else:
-      lead = ''
+   url_path = "%s/%s/%s/%s" % (year, month, day, file)
    item = NewsItem(
       "%s-%s-%s" % (year, month, day),
-      "%s%s" % (lead, title),
-      "%s/%s/%s/%s" % (year, month, day, file)
-   )
+      "%s%s" % (lead+' ' if lead else '', title),
+      url_path)
    items.append(item)
    item.baseURL = baseURL
    item.categoryList = tags
-   item.excerpt = excerpts[(year, month, day, file)]
-   item.body = bodies[(year, month, day, file)]
+   body = bodies[(year, month, day, file)]
+   body = absolute_urls(body, baseURL + url_path)
+   item.body = body
+   item.excerpt = get_first_p(body)
 
 namespace = {
    "title": "Oleg Broytman's blog",