fix(code-hosting): recognize SSH repository hosts with userinfo

Use parsed hostnames instead of raw netloc values when matching configured code-hosting domains. This keeps ssh://git@host and explicit-port repository URLs on the supported code-hosting path while preserving git clone handling for SSH repository sources.

Add focused regression coverage for SSH URL userinfo and explicit-port repository URL helpers.

Signed-off-by: Asish Kumar <officialasishkumar@gmail.com>

Author: officialasishkumar

openviking/parse/accessors/git_accessor.py

@@ -20,6 +20,7 @@
 
 from openviking.utils import is_github_url, is_gitlab_url, parse_code_hosting_url
 from openviking.utils.code_hosting_utils import (
+    _domain_matches,
     is_code_hosting_url,
     is_git_repo_url,
     validate_git_ssh_uri,
@@ -275,10 +276,7 @@ def _normalize_repo_url(self, url: str) -> str:
                 base_parts = path_parts[: git_index + 1]
 
             config = get_openviking_config()
-            if (
-                parsed.netloc in config.code.github_domains + config.code.gitlab_domains
-                and len(path_parts) >= 2
-            ):
+            if _domain_matches(parsed, config.code.github_domains + config.code.gitlab_domains):
                 base_parts = path_parts[:2]
             base_path = "/" + "/".join(base_parts)
             return parsed._replace(path=base_path, query="", fragment="").geturl()

openviking/parse/parsers/code/code.py

@@ -38,6 +38,7 @@
 )
 from openviking.parse.parsers.upload_utils import upload_directory
 from openviking.utils import is_github_url, parse_code_hosting_url
+from openviking.utils.code_hosting_utils import _domain_matches
 from openviking_cli.utils.config import get_openviking_config
 from openviking_cli.utils.logger import get_logger
 
@@ -291,10 +292,7 @@ def _normalize_repo_url(self, url: str) -> str:
                 base_parts = path_parts[: git_index + 1]
 
             config = get_openviking_config()
-            if (
-                parsed.netloc in config.code.github_domains + config.code.gitlab_domains
-                and len(path_parts) >= 2
-            ):
+            if _domain_matches(parsed, config.code.github_domains + config.code.gitlab_domains):
                 base_parts = path_parts[:2]
             base_path = "/" + "/".join(base_parts)
             return parsed._replace(path=base_path, query="", fragment="").geturl()

openviking/utils/code_hosting_utils.py

@@ -8,11 +8,37 @@
 """
 
 from typing import Optional
-from urllib.parse import urlparse
+from urllib.parse import ParseResult, urlparse
 
 from openviking_cli.utils.config import get_openviking_config
 
 
+def _domain_matches(parsed: ParseResult, domains: list[str]) -> bool:
+    """Return True when parsed URL host matches configured domains.
+
+    ``urlparse().netloc`` includes optional userinfo and port values. Repository
+    clone URLs commonly use forms like ``ssh://git@github.com/org/repo.git``,
+    where the netloc is ``git@github.com`` but the actual host is
+    ``github.com``.
+    """
+    hostname = parsed.hostname
+    if not hostname:
+        return False
+
+    normalized_domains = {domain.lower() for domain in domains}
+    host = hostname.lower()
+    candidates = {host}
+
+    try:
+        port = parsed.port
+    except ValueError:
+        port = None
+    if port is not None:
+        candidates.add(f"{host}:{port}")
+
+    return any(candidate in normalized_domains for candidate in candidates)
+
+
 def _extract_host(url: str) -> str:
     """Extract normalized host for supported git/code-hosting URL forms."""
     if url.startswith("git@"):
@@ -44,7 +70,6 @@ def parse_code_hosting_url(url: str) -> Optional[str]:
             + config.code.code_hosting_domains
         )
     )
-    host = _extract_host(url)
 
     # Handle git@ SSH URLs: git@host:org/repo.git
     if url.startswith("git@"):
@@ -73,7 +98,7 @@ def parse_code_hosting_url(url: str) -> Optional[str]:
 
     # For GitHub/GitLab URLs with org/repo structure
     if (
-        host in config.code.github_domains + config.code.gitlab_domains
+        _domain_matches(parsed, config.code.github_domains + config.code.gitlab_domains)
         and len(path_parts) >= 2
     ):
         # Take first two parts: org/repo
@@ -140,7 +165,7 @@ def is_code_hosting_url(url: str) -> bool:
         host_part = url[4:].split(":", 1)[0]
         return host_part in all_domains
 
-    return _extract_host(url) in all_domains
+    return _domain_matches(urlparse(url), all_domains)
 
 
 def validate_git_ssh_uri(url: str) -> None:
@@ -186,7 +211,7 @@ def is_git_repo_url(url: str) -> bool:
             )
         )
         parsed = urlparse(url)
-        if _extract_host(url) not in all_domains:
+        if not _domain_matches(parsed, all_domains):
             return False
         path_parts = [p for p in parsed.path.split("/") if p]
         # Strip .git suffix from last part for counting

tests/test_code_hosting_utils.py

@@ -43,6 +43,8 @@ def _mock_config():
 _spec.loader.exec_module(_module)
 
 parse_code_hosting_url = _module.parse_code_hosting_url
+is_github_url = _module.is_github_url
+is_gitlab_url = _module.is_gitlab_url
 is_code_hosting_url = _module.is_code_hosting_url
 is_git_repo_url = _module.is_git_repo_url
 validate_git_ssh_uri = _module.validate_git_ssh_uri
@@ -81,12 +83,16 @@ def test_parse_code_hosting_url_https_dotgit():
     assert parse_code_hosting_url("https://github.com/org/repo.git") == "org/repo"
 
 
-def test_parse_code_hosting_url_ssh_with_userinfo():
+def test_parse_code_hosting_url_ssh_url_with_userinfo():
     assert parse_code_hosting_url("ssh://git@github.com/org/repo.git") == "org/repo"
 
 
-def test_parse_code_hosting_url_https_with_explicit_port():
-    assert parse_code_hosting_url("https://github.com:443/org/repo.git") == "org/repo"
+def test_parse_code_hosting_url_gitlab_ssh_url_with_userinfo():
+    assert parse_code_hosting_url("ssh://git@gitlab.com/group/repo.git") == "group/repo"
+
+
+def test_parse_code_hosting_url_https_with_port():
+    assert parse_code_hosting_url("https://github.com:443/org/repo") == "org/repo"
 
 
 # --- validate_git_ssh_uri ---
@@ -126,14 +132,25 @@ def test_is_code_hosting_url_https():
     assert is_code_hosting_url("https://github.com/org/repo") is True
 
 
-def test_is_code_hosting_url_ssh_with_userinfo():
+def test_is_code_hosting_url_ssh_url_with_userinfo():
     assert is_code_hosting_url("ssh://git@github.com/org/repo.git") is True
 
 
-def test_is_code_hosting_url_https_with_explicit_port():
+def test_is_code_hosting_url_https_with_port():
     assert is_code_hosting_url("https://github.com:443/org/repo") is True
 
 
+# --- is_github_url / is_gitlab_url ---
+
+
+def test_is_github_url_ssh_url_with_userinfo():
+    assert is_github_url("ssh://git@github.com/org/repo.git") is True
+
+
+def test_is_gitlab_url_ssh_url_with_userinfo():
+    assert is_gitlab_url("ssh://git@gitlab.com/group/repo.git") is True
+
+
 # --- is_git_repo_url ---
 
 
@@ -145,11 +162,11 @@ def test_is_git_repo_url_https_repo():
     assert is_git_repo_url("https://github.com/org/repo") is True
 
 
-def test_is_git_repo_url_ssh_with_userinfo():
+def test_is_git_repo_url_ssh_url_with_userinfo():
     assert is_git_repo_url("ssh://git@github.com/org/repo.git") is True
 
 
-def test_is_git_repo_url_https_with_explicit_port():
+def test_is_git_repo_url_https_with_port():
     assert is_git_repo_url("https://github.com:443/org/repo") is True
 
 

tests/unit/test_accessors_git.py

@@ -3,10 +3,29 @@
 """Unit tests for GitAccessor."""
 
 from pathlib import Path
+from types import SimpleNamespace
+from unittest.mock import patch
 
 import pytest
 
 from openviking.parse.accessors import GitAccessor
+from openviking.utils import code_hosting_utils
+
+
+def _mock_config():
+    return SimpleNamespace(
+        code=SimpleNamespace(
+            github_domains=["github.com", "www.github.com"],
+            gitlab_domains=["gitlab.com", "www.gitlab.com"],
+            code_hosting_domains=["github.com", "gitlab.com"],
+        )
+    )
+
+
+@pytest.fixture(autouse=True)
+def _patch_config():
+    with patch.object(code_hosting_utils, "get_openviking_config", side_effect=_mock_config):
+        yield
 
 
 class TestGitAccessor:
@@ -60,6 +79,13 @@ def test_can_handle_git_protocol_url(self, accessor: GitAccessor) -> None:
         """GitAccessor should handle git:// URLs."""
         assert accessor.can_handle("git://github.com/volcengine/OpenViking.git") is True
 
+    def test_normalize_repo_url_ssh_with_userinfo_and_ref(self, accessor: GitAccessor) -> None:
+        """GitAccessor should normalize ssh URLs with userinfo using the shared host matcher."""
+        assert (
+            accessor._normalize_repo_url("ssh://git@github.com:443/volcengine/OpenViking/tree/main")
+            == "ssh://git@github.com:443/volcengine/OpenViking"
+        )
+
     @pytest.mark.parametrize(
         "source",
         [