#15114: the strict mode of HTMLParser and the HTMLParseError exception are deprecated now that the parser is able to parse invalid markup.

ezio-melotti · ezio-melotti · commit 3861d8b27127 · 2012-06-23T15:27:51.000+02:00
diff --git a/Doc/library/html.parser.rst b/Doc/library/html.parser.rst
@@ -16,13 +16,14 @@
 This module defines a class :class:`HTMLParser` which serves as the basis for
 parsing text files formatted in HTML (HyperText Mark-up Language) and XHTML.
 
-.. class:: HTMLParser(strict=True)
+.. class:: HTMLParser(strict=False)
 
-   Create a parser instance.  If *strict* is ``True`` (the default), invalid
-   HTML results in :exc:`~html.parser.HTMLParseError` exceptions [#]_.  If
-   *strict* is ``False``, the parser uses heuristics to make a best guess at
-   the intention of any invalid HTML it encounters, similar to the way most
-   browsers do.  Using ``strict=False`` is advised.
+   Create a parser instance.  If *strict* is ``False`` (the default), the parser
+   will accept and parse invalid markup.  If *strict* is ``True`` the parser
+   will raise an :exc:`~html.parser.HTMLParseError` exception instead [#]_ when
+   it's not able to parse the markup.
+   The use of ``strict=True`` is discouraged and the *strict* argument is
+   deprecated.
 
    An :class:`.HTMLParser` instance is fed HTML data and calls handler methods
    when start tags, end tags, text, comments, and other markup elements are
@@ -34,6 +35,10 @@ parsing text files formatted in HTML (HyperText Mark-up Language) and XHTML.
 
    .. versionchanged:: 3.2 *strict* keyword added
 
+   .. deprecated-removed:: 3.3 3.5
+      The *strict* argument and the strict mode have been deprecated.
+      The parser is now able to accept and parse invalid markup too.
+
 An exception is defined as well:
 
 
@@ -46,6 +51,10 @@ An exception is defined as well:
    detected, and :attr:`offset` is the number of characters into the line at
    which the construct starts.
 
+   .. deprecated-removed:: 3.3 3.5
+      This exception has been deprecated because it's never raised by the parser
+      (when the default non-strict mode is used).
+
 
 Example HTML Parser Application
 -------------------------------
diff --git a/Lib/html/parser.py b/Lib/html/parser.py
@@ -10,6 +10,7 @@
 
 import _markupbase
 import re
+import warnings
 
 # Regular expressions used for parsing
 
@@ -113,14 +114,16 @@ class HTMLParser(_markupbase.ParserBase):
 
     CDATA_CONTENT_ELEMENTS = ("script", "style")
 
-    def __init__(self, strict=True):
+    def __init__(self, strict=False):
         """Initialize and reset this instance.
 
-        If strict is set to True (the default), errors are raised when invalid
-        HTML is encountered.  If set to False, an attempt is instead made to
-        continue parsing, making "best guesses" about the intended meaning, in
-        a fashion similar to what browsers typically do.
+        If strict is set to False (the default) the parser will parse invalid
+        markup, otherwise it will raise an error.  Note that the strict mode
+        is deprecated.
         """
+        if strict:
+            warnings.warn("The strict mode is deprecated.",
+                          DeprecationWarning, stacklevel=2)
         self.strict = strict
         self.reset()
 
@@ -271,8 +274,8 @@ def goahead(self, end):
     # See also parse_declaration in _markupbase
     def parse_html_declaration(self, i):
         rawdata = self.rawdata
-        if rawdata[i:i+2] != '<!':
-            self.error('unexpected call to parse_html_declaration()')
+        assert rawdata[i:i+2] == '<!', ('unexpected call to '
+                                        'parse_html_declaration()')
         if rawdata[i:i+4] == '<!--':
             # this case is actually already handled in goahead()
             return self.parse_comment(i)
@@ -292,8 +295,8 @@ def parse_html_declaration(self, i):
     # see http://www.w3.org/TR/html5/tokenization.html#bogus-comment-state
     def parse_bogus_comment(self, i, report=1):
         rawdata = self.rawdata
-        if rawdata[i:i+2] not in ('<!', '</'):
-            self.error('unexpected call to parse_comment()')
+        assert rawdata[i:i+2] in ('<!', '</'), ('unexpected call to '
+                                                'parse_comment()')
         pos = rawdata.find('>', i+2)
         if pos == -1:
             return -1
diff --git a/Lib/test/test_htmlparser.py b/Lib/test/test_htmlparser.py
@@ -102,7 +102,8 @@ def parse(source=source):
 class HTMLParserStrictTestCase(TestCaseBase):
 
     def get_collector(self):
-        return EventCollector(strict=True)
+        with support.check_warnings(("", DeprecationWarning), quite=False):
+            return EventCollector(strict=True)
 
     def test_processing_instruction_only(self):
         self._run_check("<?processing instruction>", [
@@ -594,7 +595,8 @@ def test_broken_condcoms(self):
 class AttributesStrictTestCase(TestCaseBase):
 
     def get_collector(self):
-        return EventCollector(strict=True)
+        with support.check_warnings(("", DeprecationWarning), quite=False):
+            return EventCollector(strict=True)
 
     def test_attr_syntax(self):
         output = [
diff --git a/Misc/NEWS b/Misc/NEWS
@@ -43,6 +43,9 @@ Core and Builtins
 Library
 -------
 
+- Issue #15114: the strict mode of HTMLParser and the HTMLParseError exception
+  are deprecated now that the parser is able to parse invalid markup.
+
 - Issue #3665: \u and \U escapes are now supported in unicode regular
   expressions.  Patch by Serhiy Storchaka.
 
@@ -78,7 +81,7 @@ Library
 - Issue #9527: datetime.astimezone() method will now supply a class
   timezone instance corresponding to the system local timezone when
   called with no arguments.
-  
+
 - Issue #14653: email.utils.mktime_tz() no longer relies on system
   mktime() when timezone offest is supplied.
 
