Add typing support

Author: palfrey

.github/workflows/feedparser.yml

@@ -67,13 +67,13 @@ jobs:
         if: (startsWith(runner.os, 'Linux') || startsWith(runner.os, 'macOS'))
         run: |
           source venv/bin/activate
-          tox -e py,py-chardet
+          tox -e py,py-chardet,mypy
 
       - name: Run tox
         if: startsWith(runner.os, 'Windows')
         run: |
           .\venv\scripts\activate.ps1
-          tox -e py,py-chardet
+          tox -e py,py-chardet,mypy
 
   docs:
     name: "Build documentation"

.gitignore

@@ -11,3 +11,4 @@ build/
 .doit.*
 *.html
 poetry.lock
+.vscode/

CONTRIBUTORS.rst

@@ -20,6 +20,7 @@ bug report!
 * `Tom Most <https://github.com/twm/>`_
 * `Nik Nyby <http://nikolas.us.to/>`_
 * `Ade Oshineye <http://blog.oshineye.com/>`_
+* `Tom Parker-Shemilt <https://tevps.net>`_
 * `Martin Pool <http://sourcefrog.net/>`_
 * `Sam Ruby <http://intertwingly.net/>`_
 * `Bernd Schlapsi <https://github.com/brot>`_

feedparser/api.py

@@ -138,17 +138,11 @@ def _open_resource(url_file_stream_or_string, etag, modified, agent, referrer, h
     return url_file_stream_or_string
 
 
-LooseFeedParser = type(
-    'LooseFeedParser',
-    (LooseXMLParser, XMLParserMixin, BaseHTMLProcessor),
-    {},
-)
+class LooseFeedParser(LooseXMLParser, XMLParserMixin, BaseHTMLProcessor):
+    pass
 
-StrictFeedParser = type(
-    'StrictFeedParser',
-    (StrictXMLParser, XMLParserMixin, xml.sax.handler.ContentHandler),
-    {},
-)
+class StrictFeedParser(StrictXMLParser, XMLParserMixin, xml.sax.handler.ContentHandler):
+    pass
 
 
 def parse(
@@ -214,10 +208,10 @@ def parse(
         agent = feedparser.USER_AGENT
     if sanitize_html is None:
         import feedparser
-        sanitize_html = feedparser.SANITIZE_HTML
+        sanitize_html = bool(feedparser.SANITIZE_HTML)
     if resolve_relative_uris is None:
         import feedparser
-        resolve_relative_uris = feedparser.RESOLVE_RELATIVE_URIS
+        resolve_relative_uris = bool(feedparser.RESOLVE_RELATIVE_URIS)
 
     result = FeedParserDict(
         bozo=False,
@@ -258,7 +252,8 @@ def parse(
         baselang = baselang.decode('utf-8', 'ignore')
 
     if not _XML_AVAILABLE:
-        use_strict_parser = 0
+        use_strict_parser = False
+    feed_parser: Union[JSONParser, StrictFeedParser, LooseFeedParser]
     if use_json_parser:
         result['version'] = None
         feed_parser = JSONParser(baseuri, baselang, 'utf-8')
@@ -288,7 +283,7 @@ def parse(
         except xml.sax.SAXException as e:
             result['bozo'] = 1
             result['bozo_exception'] = feed_parser.exc or e
-            use_strict_parser = 0
+            use_strict_parser = False
 
     # The loose XML parser will be tried if the JSON parser was not used,
     # and if the strict XML parser was not used (or if if it failed).
@@ -301,5 +296,8 @@ def parse(
     result['feed'] = feed_parser.feeddata
     result['entries'] = feed_parser.entries
     result['version'] = result['version'] or feed_parser.version
-    result['namespaces'] = feed_parser.namespaces_in_use
+    if isinstance(feed_parser, JSONParser):
+        result['namespaces'] = {}
+    else:    
+        result['namespaces'] = feed_parser.namespaces_in_use
     return result

feedparser/datetimes/__init__.py

@@ -25,6 +25,8 @@
 # ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 # POSSIBILITY OF SUCH DAMAGE.
 
+from time import struct_time
+from typing import Callable, List, Optional
 from .asctime import _parse_date_asctime
 from .greek import _parse_date_greek
 from .hungarian import _parse_date_hungarian
@@ -34,7 +36,7 @@
 from .rfc822 import _parse_date_rfc822
 from .w3dtf import _parse_date_w3dtf
 
-_date_handlers = []
+_date_handlers: List[Callable[[str], Optional[struct_time]]] = []
 
 
 def registerDateHandler(func):

feedparser/datetimes/iso8601.py

@@ -68,15 +68,7 @@
     + r'(\.(?P<fracsecond>\d+))?'
     + r'(?P<tz>[+-](?P<tzhour>\d{2})(:(?P<tzmin>\d{2}))?|Z)?)?'
     for tmpl in _iso8601_tmpl]
-try:
-    del tmpl
-except NameError:
-    pass
 _iso8601_matches = [re.compile(regex).match for regex in _iso8601_re]
-try:
-    del regex
-except NameError:
-    pass
 
 
 def _parse_date_iso8601(date_string):

feedparser/datetimes/perforce.py

@@ -25,7 +25,7 @@
 # ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 # POSSIBILITY OF SUCH DAMAGE.
 
-import email._parseaddr
+import email.utils
 import re
 import time
 
@@ -41,6 +41,6 @@ def _parse_date_perforce(date_string):
     dow, year, month, day, hour, minute, second, tz = m.groups()
     months = ['Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']
     new_date_string = "%s, %s %s %s %s:%s:%s %s" % (dow, day, months[int(month) - 1], year, hour, minute, second, tz)
-    tm = email._parseaddr.parsedate_tz(new_date_string)
+    tm = email.utils.parsedate_tz(new_date_string)
     if tm:
-        return time.gmtime(email._parseaddr.mktime_tz(tm))
+        return time.gmtime(email.utils.mktime_tz(tm))

feedparser/encodings.py

@@ -32,11 +32,10 @@
 
 try:
     try:
-        import cchardet as chardet
+        import chardet as chardet
     except ImportError:
         import chardet
 except ImportError:
-    chardet = None
     lazy_chardet_encoding = None
 else:
     def lazy_chardet_encoding(data):

feedparser/mixin.py

@@ -30,6 +30,7 @@
 import copy
 import html.entities
 import re
+from typing import Dict
 import xml.sax.saxutils
 
 from .html import _cp1252
@@ -118,7 +119,7 @@ class XMLParserMixin(
         'http://www.w3.org/XML/1998/namespace':                  'xml',
         'http://podlove.org/simple-chapters':                    'psc',
     }
-    _matchnamespaces = {}
+    _matchnamespaces: Dict[str, str] = {}
 
     can_be_relative_uri = {
         'comments',
@@ -170,6 +171,8 @@ def __init__(self):
         self.entries = []  # list of entry-level data
         self.version = ''  # feed type/version, see SUPPORTED_VERSIONS
         self.namespaces_in_use = {}  # dictionary of namespaces defined by the feed
+        self.resolve_relative_uris = False
+        self.sanitize_html = False
 
         # the following are used internally to track state;
         # this is really out of control and should be refactored