diff options
Diffstat (limited to 'Lib/html')
| -rw-r--r-- | Lib/html/__init__.py | 13 | ||||
| -rw-r--r-- | Lib/html/parser.py | 16 |
2 files changed, 18 insertions, 11 deletions
diff --git a/Lib/html/__init__.py b/Lib/html/__init__.py index 02652ef..2ad167f 100644 --- a/Lib/html/__init__.py +++ b/Lib/html/__init__.py @@ -2,11 +2,6 @@ General functions for HTML manipulation. """ - -_escape_map = {ord('&'): '&', ord('<'): '<', ord('>'): '>'} -_escape_map_full = {ord('&'): '&', ord('<'): '<', ord('>'): '>', - ord('"'): '"', ord('\''): '''} - # NB: this is a candidate for a bytes/string polymorphic interface def escape(s, quote=True): @@ -16,6 +11,10 @@ def escape(s, quote=True): characters, both double quote (") and single quote (') characters are also translated. """ + s = s.replace("&", "&") # Must be done first! + s = s.replace("<", "<") + s = s.replace(">", ">") if quote: - return s.translate(_escape_map_full) - return s.translate(_escape_map) + s = s.replace('"', """) + s = s.replace('\'', "'") + return s diff --git a/Lib/html/parser.py b/Lib/html/parser.py index 2d3bef3..22498db 100644 --- a/Lib/html/parser.py +++ b/Lib/html/parser.py @@ -12,6 +12,8 @@ import _markupbase import re import warnings +__all__ = ['HTMLParser'] + # Regular expressions used for parsing interesting_normal = re.compile('[&<]') @@ -92,6 +94,8 @@ class HTMLParseError(Exception): return result +_strict_sentinel = object() + class HTMLParser(_markupbase.ParserBase): """Find tags and other markup and call handler functions. @@ -114,16 +118,18 @@ class HTMLParser(_markupbase.ParserBase): CDATA_CONTENT_ELEMENTS = ("script", "style") - def __init__(self, strict=False): + def __init__(self, strict=_strict_sentinel): """Initialize and reset this instance. 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. + and argument are deprecated. """ - if strict: - warnings.warn("The strict mode is deprecated.", + if strict is not _strict_sentinel: + warnings.warn("The strict argument and mode are deprecated.", DeprecationWarning, stacklevel=2) + else: + strict = False # default self.strict = strict self.reset() @@ -149,6 +155,8 @@ class HTMLParser(_markupbase.ParserBase): self.goahead(1) def error(self, message): + warnings.warn("The 'error' method is deprecated.", + DeprecationWarning, stacklevel=2) raise HTMLParseError(message, self.getpos()) __starttag_text = None |