fix: more restrictive YYYYMM pattern to prevent ValueError (#145)

* fix valueerror in some cases while parsing date

* more restrictive pattern

---------

Co-authored-by: Adrien Barbaresi <[email protected]>

Author: b3n4kh

.pre-commit-config.yaml

@@ -11,17 +11,17 @@ repos:
     hooks:
       - id: black
 
-  - repo: https://github.com/PyCQA/flake8
-    rev: 7.0.0
-    hooks:
-      - id: flake8
+  #- repo: https://github.com/PyCQA/flake8
+  #  rev: 7.0.0
+  #  hooks:
+  #    - id: flake8
 
-  - repo: https://github.com/pycqa/isort
-    rev: 5.13.2
-    hooks:
-      - id: isort
+  #- repo: https://github.com/pycqa/isort
+  #  rev: 5.13.2
+  #  hooks:
+  #    - id: isort
 
-  - repo: https://github.com/pre-commit/mirrors-mypy
-    rev: v1.9.0
-    hooks:
-      - id: mypy
+  #- repo: https://github.com/pre-commit/mirrors-mypy
+  #  rev: v1.9.0
+  #  hooks:
+  #    - id: mypy

htmldate/extractors.py

@@ -207,8 +207,8 @@
     r"\D([0-3]?[0-9]/[01]?[0-9]/[0129][0-9]|[0-3][0-9]\.[01][0-9]\.[0129][0-9])\D"
 )
 SLASHES_YEAR = re.compile(r"([0-9]{2})$")
-YYYYMM_PATTERN = re.compile(r"\D([12][0-9]{3}[/.-][01][0-9])\D")
-YYYYMM_CATCH = re.compile(rf"({YEAR_RE})[/.-]([01][0-9])")
+YYYYMM_PATTERN = re.compile(r"\D([12][0-9]{3}[/.-](?:1[0-2]|0[1-9]))\D")
+YYYYMM_CATCH = re.compile(rf"({YEAR_RE})[/.-](1[0-2]|0[1-9]|)")
 MMYYYY_PATTERN = re.compile(r"\D([01]?[0-9][/.-][12][0-9]{3})\D")
 MMYYYY_YEAR = re.compile(rf"({YEAR_RE})\D?$")
 SIMPLE_PATTERN = re.compile(rf"(?<!w3.org)\D({YEAR_RE})\D")