Added support to parse dates from page body

newspaper/extractors.py

@@ -17,7 +17,8 @@
 import re
 from collections import defaultdict
 
-from dateutil.parser import parse as date_parser
+from dateutil import parser as date_parser
+from qddate import DateParser
 from tldextract import tldextract
 from urllib.parse import urljoin, urlparse, urlunparse
 
@@ -230,6 +231,16 @@ def parse_date_str(date_str):
                 if datetime_obj:
                     return datetime_obj
 
+        # Uses qddate DateParser to find dates in any strings less than 150 chars long
+        dp = DateParser(generate=True)
+        nodes = doc.xpath('//*[string-length(text())<150]')
+        for node in nodes:
+            if node.text is None:
+                continue
+            datetime_obj = dp.parse(node.text)
+            if datetime_obj is not None:
+                return datetime_obj
+
         return None
 
     def get_title(self, doc):

requirements.txt

@@ -10,3 +10,4 @@ tldextract>=2.0.1
 feedfinder2>=0.0.4
 jieba3k>=0.35.1
 python-dateutil>=2.5.3
+qddate>=0.1.0