Add support to fetch purl

Signed-off-by: Tushar Goel <tushar.goel.dav@gmail.com>

requirements.txt

@@ -41,8 +41,8 @@ MarkupSafe==2.0.1
 more-itertools==8.13.0
 normality==2.3.3
 packagedcode-msitools==0.101.210706
-packageurl-python==0.9.9
-packaging==21.3
+packageurl-python==0.17.4
+packaging==24.0
 parameter-expansion-patched==0.3.1
 patch==1.16
 pdfminer-six==20220506

src/fetchcode/__init__.py

@@ -21,6 +21,7 @@
 from urllib.parse import urlparse
 
 import requests
+from packageurl.contrib.purl2url import get_download_url
 
 
 class Response:
@@ -89,19 +90,34 @@ def fetch_ftp(url, location):
  return resp
 
 
+def fetch_purl(purl, location=None):
+ """
+ Return a `Response` object built from fetching the content at a PURL based `purl` URL string
+ saving the content in a file at `location`
+ """
+ from fetchcode.download_urls import download_url as get_download_url_from_fetchcode
+
+ for resolver in (get_download_url, get_download_url_from_fetchcode):
+ url = resolver(purl)
+ if url:
+ return fetch(url=url)
+ return
+
+
 def fetch(url):
  """
  Return a `Response` object built from fetching the content at the `url` URL string and
  store content at a temporary file.
  """
-
- temp = tempfile.NamedTemporaryFile(delete=False)
- location = temp.name
-
  url_parts = urlparse(url)
  scheme = url_parts.scheme
+ location = None
+
+ if scheme != "purl":
+ temp = tempfile.NamedTemporaryFile(delete=False)
+ location = temp.name
 
- fetchers = {"ftp": fetch_ftp, "http": fetch_http, "https": fetch_http}
+ fetchers = {"ftp": fetch_ftp, "http": fetch_http, "https": fetch_http, "pkg": fetch_purl}
 
  if scheme in fetchers:
  return fetchers.get(scheme)(url, location)