Adding --platform-aware to url-parse command

Fix #600
medialab · Jul 24, 2023 · 23d8cf2 · 23d8cf2
1 parent f8907dc
commit 23d8cf2
Show file tree

Hide file tree

Showing 2 changed files with 11 additions and 1 deletion.
diff --git a/minet/cli/url_parse/__init__.py b/minet/cli/url_parse/__init__.py
@@ -224,5 +224,10 @@ def __call__(self, parser, cli_args, values, option_string=None):
             "help": "Whether to strip the hostname suffix when fingerprinting the url.",
             "action": "store_true",
         },
+        {
+            "flag": "--plaftorm-aware",
+            "help": "Whether url parsing should know about some specififc platform such as Facebook, YouTube etc. into account when normalizing urls. Note that this is different than activating --facebook or --youtube.",
+            "action": "store_true",
+        },
     ],
 )
diff --git a/minet/cli/url_parse/url_parse.py b/minet/cli/url_parse/url_parse.py
@@ -92,8 +92,13 @@ def extract_standard_addendum(cli_args, url):
             strip_irrelevant_subdomains=cli_args.strip_irrelevant_subdomains,
             strip_protocol=cli_args.strip_protocol,
             strip_trailing_slash=cli_args.strip_trailing_slash,
+            platform_aware=cli_args.platform_aware,
+        ),
+        fingerprint_url(
+            url,
+            strip_suffix=cli_args.strip_suffix,
+            platform_aware=cli_args.platform_aware,
         ),
-        fingerprint_url(url, strip_suffix=cli_args.strip_suffix),
         infer_redirection(url),
         get_domain_name(url),
         get_hostname(url),