Fix(Robot): Stop splitting and un-splitting URLs

[bookmarks_db.git] / parse_html / bkmk_ph_beautifulsoup4.py

diff --git a/parse_html/bkmk_ph_beautifulsoup4.py b/parse_html/bkmk_ph_beautifulsoup4.py
index aad3f8fb9092bdba2daec856a93f4309f9714795..060f0786d473c17de36ad073c05632c06f30c35d 100644 (file)
--- a/parse_html/bkmk_ph_beautifulsoup4.py
+++ b/parse_html/bkmk_ph_beautifulsoup4.py
@@ -5,25 +5,42 @@ This file is a part of Bookmarks database and Internet robot.
 """
 
 __author__ = "Oleg Broytman <phd@phdru.name>"
-__copyright__ = "Copyright (C) 2017 PhiloSoft Design"
+__copyright__ = "Copyright (C) 2017-2023 PhiloSoft Design"
 __license__ = "GNU GPL"
 
 __all__ = ['parse_html']
 
 
-import re
+import warnings
+
 from bs4 import BeautifulSoup
+
 from .bkmk_ph_util import HTMLParser
+from compat import string_type
+
+warnings.filterwarnings(
+    'ignore', 'No parser was explicitly specified')
+warnings.filterwarnings(
+    'ignore',
+    "It looks like you're parsing an XML document using an HTML parser.")
+
+universal_charset = "utf-8"
+DEFAULT_CHARSET = "cp1251"  # Stupid default for Russian Cyrillic
 
-DEFAULT_CHARSET = "cp1251" # Stupid default for Russian Cyrillic
 
 def _parse_html(html_text, charset):
     try:
-        return BeautifulSoup(html_text, from_encoding=charset)
+        if isinstance(html_text, bytes):
+            return BeautifulSoup(html_text, from_encoding=charset)
+        else:
+            return BeautifulSoup(html_text)
     except TypeError:
         return None
 
+
 def parse_html(html_text, charset=None, log=None):
+    if not html_text:
+        return None
     root = _parse_html(html_text, charset)
     if root is None:
         return None
@@ -35,7 +52,7 @@ def parse_html(html_text, charset=None, log=None):
 
     head = html.head
     if head is None:
-        head = html # Some sites put TITLE in HTML without HEAD
+        head = html  # Some sites put TITLE in HTML without HEAD
 
     title = head.title
     if (title is None) and (html is not head):
@@ -52,9 +69,12 @@ def parse_html(html_text, charset=None, log=None):
         else:
             parts = []
             for part in title:
-                if not isinstance(part, basestring):
-                    part = unicode(part)
-                parts.append(part.strip())
+                #if not isinstance(part, string_type):
+                #    part = part.decode()
+                if part.strip:
+                    parts.append(part.strip())
+                else:
+                    parts.append(' ')  # Skip tags, they're usually `<br>`
             title = ''.join(parts)
 
     meta = head.find(_find_contenttype, recursive=False)
@@ -62,10 +82,11 @@ def parse_html(html_text, charset=None, log=None):
         try:
             meta_content = meta.get("content")
             if meta_content:
-                __charset = meta_content.lower().split('charset=')[1].split(';')[0]
+                __charset = meta_content.lower().split('charset=')[1].\
+                    split(';')[0]
             else:
                 __charset = False
-        except IndexError: # No charset in the META Content-Type
+        except IndexError:  # No charset in the META Content-Type
             meta_charset = False
         else:
             meta_charset = _charset = __charset
@@ -79,8 +100,12 @@ def parse_html(html_text, charset=None, log=None):
             if meta_content:
                 meta_charset = _charset = meta_content.lower()
 
-    if title and (_charset or meta_charset):
-        title = title.encode(_charset or meta_charset)
+    #if title and (_charset or meta_charset):
+    #    try:
+    #        title = title.encode(_charset or meta_charset)
+    #    except LookupError:
+    #        title = title.encode(universal_charset)
+    #        _charset = universal_charset
 
     meta = head.find(_find_refresh, recursive=False)
     if meta:
@@ -98,17 +123,23 @@ def parse_html(html_text, charset=None, log=None):
         return None
     return HTMLParser(_charset, meta_charset, title, refresh, icon)
 
+
 def _find_contenttype(Tag):
     return (Tag.name == "meta") and \
        (Tag.get_attribute_list("http-equiv", '')[0].lower() == "content-type")
 
+
 def _find_charset(Tag):
     return (Tag.name == "meta") and Tag.get("charset", '')
 
+
 def _find_refresh(Tag):
     return (Tag.name == "meta") and \
        (Tag.get_attribute_list("http-equiv", '')[0].lower() == "refresh")
 
+
 def _find_icon(Tag):
-    return (Tag.name == "link") and \
-       (Tag.get_attribute_list("rel", '')[0].lower() in ('icon', 'shortcut icon'))
+    if Tag.name != "link":
+        return False
+    rel = ' '.join(Tag.get_attribute_list("rel", ''))
+    return rel in ('icon', 'shortcut icon')