Fix get protocol and path (#4409)

* Fixed _parse_filepath Signed-off-by: Elena Khaustova <[email protected]> * Updated unit tests Signed-off-by: Elena Khaustova <[email protected]> * Updated release notes Signed-off-by: Elena Khaustova <[email protected]> --------- Signed-off-by: Elena Khaustova <[email protected]>
kedro-org · Jan 14, 2025 · 25bb58b · 25bb58b
1 parent f72d5cd
commit 25bb58b
Show file tree

Hide file tree

Showing 3 changed files with 24 additions and 0 deletions.
diff --git a/RELEASE.md b/RELEASE.md
@@ -11,6 +11,7 @@
 * Added `node` import to the pipeline template.
 * Update error message when executing kedro run without pipeline.
 * Safeguard hooks when user incorrectly registers a hook class in settings.py.
+* Fixed parsing paths with query and fragment.
 
 ## Breaking changes to the API
 ## Documentation changes

diff --git a/kedro/io/core.py b/kedro/io/core.py
@@ -901,6 +901,11 @@ def _parse_filepath(filepath: str) -> dict[str, str]:
         if windows_path:
             path = ":".join(windows_path.groups())
 
+    if parsed_path.query:
+        path = f"{path}?{parsed_path.query}"
+    if parsed_path.fragment:
+        path = f"{path}#{parsed_path.fragment}"
+
     options = {"protocol": protocol, "path": path}
 
     if parsed_path.netloc and protocol in CLOUD_PROTOCOLS:

diff --git a/tests/io/test_core.py b/tests/io/test_core.py
@@ -271,6 +271,24 @@ def test_get_filepath_str(self):
             ("file:///C:\\Projects\\file.txt", ("file", "C:\\Projects\\file.txt")),
             ("https://example.com/file.txt", ("https", "example.com/file.txt")),
             ("http://example.com/file.txt", ("http", "example.com/file.txt")),
+            (
+                "https://example.com/search?query=books&category=fiction#reviews",
+                ("https", "example.com/search?query=books&category=fiction#reviews"),
+            ),
+            (
+                "https://example.com/search#reviews",
+                ("https", "example.com/search#reviews"),
+            ),
+            (
+                "http://example.com/search?query=books&category=fiction",
+                ("http", "example.com/search?query=books&category=fiction"),
+            ),
+            (
+                "s3://some/example?query=query#filename",
+                ("s3", "some/example?query=query#filename"),
+            ),
+            ("s3://some/example#filename", ("s3", "some/example#filename")),
+            ("s3://some/example?query=query", ("s3", "some/example?query=query")),
         ],
     )
     def test_get_protocol_and_path(self, filepath, expected_result):