Reparse the HTML if the charset was changed.

[bookmarks_db.git] / Robots / parse_html_beautifulsoup.py

diff --git a/Robots/parse_html_beautifulsoup.py b/Robots/parse_html_beautifulsoup.py
index c45af2b0221a61f441cf5165fc2a25f6fe978bb5..93a4d6fd9a2ba27ba0c1de2fdcdb3fe9971461c0 100644 (file)
--- a/Robots/parse_html_beautifulsoup.py
+++ b/Robots/parse_html_beautifulsoup.py
@@ -49,44 +49,62 @@ class BadDeclParser(BeautifulSoup):
          return j
 
 
-def parse_html(filename, charset=None):
+def _parse_html(filename, charset):
    infile = open(filename, 'r')
    try:
-      root = BadDeclParser(infile, fromEncoding=charset)
+      return BadDeclParser(infile, fromEncoding=charset)
    except TypeError:
       return None
    finally:
       infile.close()
 
+def parse_html(filename, charset=None, log=None):
+   root = _parse_html(filename, charset)
+
+   _charset = root.originalEncoding
+   if _charset in ("ISO-8859-2", "windows-1252", "MacCyrillic"): # Replace default
+      _charset = DEFAULT_CHARSET
+      root = _parse_html(filename, _charset)
+
    try:
-      head = root.html.head
+      html = root.html
    except AttributeError:
-      return None
+      if log: log("No HTML in root")
+      html = root
 
-   if head is None:
-      head = root.html # Some sites put TITLE in HTML without HEAD
+   if html is None:
+      html = root
 
-   _charset = root.originalEncoding
-   if _charset == "windows-1252": # Replace default
-      _charset = DEFAULT_CHARSET
+   try:
+      head = html.head
+   except AttributeError:
+      if log: log("No HEAD in HTML")
+      head = html
+
+   if head is None:
+      head = html # Some sites put TITLE in HTML without HEAD
 
    try:
       title = head.title.string.encode(_charset)
    except AttributeError:
       title = '' # HEAD but no TITLE
 
-   if (not title) and (head is not root.html):
+   if (not title) and (head is not html):
       # Some sites put TITLE in HTML outside of HEAD
 
       try:
-         title = root.html.title.string.encode(_charset)
+         title = html.title.string.encode(_charset)
       except AttributeError:
          title = '' # no TITLE in HTML too
 
    meta = head.find(_find_contenttype, recursive=False)
    if meta:
-      __charset = meta.get("content").lower().split('charset=')[1].split(';')[0]
-      meta_charset = _charset == __charset
+      try:
+         __charset = meta.get("content").lower().split('charset=')[1].split(';')[0]
+      except IndexError: # No charset in the META Content-Type
+         meta_charset = False
+      else:
+         meta_charset = _charset == __charset
    else:
       meta_charset = False