helper: split_markdown_front_matter

helper for istio and mozilla importers

Signed-off-by: Hritik Vijay <hritikxx8@gmail.com>

Author: Hritik14

vulnerabilities/helpers.py

@@ -19,9 +19,9 @@
 #  for any legal advice.
 #  VulnerableCode is a free software from nexB Inc. and others.
 #  Visit https://github.com/nexB/vulnerablecode/ for support and download.
-
 import json
 import re
+from typing import Tuple
 
 import requests
 import toml
@@ -106,3 +106,30 @@ def contains_alpha(string):
     """
 
     return any([c.isalpha() for c in string])
+
+
+def split_markdown_front_matter(text: str) -> Tuple[str, str]:
+    r"""
+    Split text into markdown front matter and the markdown body
+    Returns ("", text) for text with non existing front matter
+
+    >>> text='''---
+    ... title: DUMMY-SECURITY-2019-001
+    ... description: Incorrect access control.
+    ... cves: [CVE-2042-1337]
+    ... ---
+    ... # Markdown starts here
+    ... '''
+    >>> split_markdown_front_matter(text)
+    ('title: DUMMY-SECURITY-2019-001\ndescription: Incorrect access control.\ncves: [CVE-2042-1337]\n', '\n# Markdown starts here\n')
+    """
+
+    front_matter = ""
+    body = text
+    text = text.replace("\r\n", "\n")
+    linezero,_, text = text.partition("---\n")
+
+    if not linezero: # nothing before first ---
+        front_matter,_, body = text.partition("---")
+
+    return front_matter, body