Make load_document_by_path applicable to a URI with http and https schemes (#117)

* Rename `load_document_by_path` to `load_document_by_uri`

Author: tom-tan

README.rst

@@ -69,13 +69,13 @@ Using the CWL Parsers
    from ruamel import yaml
    import sys
 
-   from cwl_utils.parser import load_document_by_path, save
+   from cwl_utils.parser import load_document_by_uri, save
 
    # File Input - This is the only thing you will need to adjust or take in as an input to your function:
    cwl_file = Path("testdata/md5sum.cwl")  # or a plain string works as well
 
    # Import CWL Object
-   cwl_obj = load_document_by_path(cwl_file)
+   cwl_obj = load_document_by_uri(cwl_file)
 
    # View CWL Object
    print("List of object attributes:\n{}".format("\n".join(map(str, dir(cwl_obj)))))

cwl_utils/parser/__init__.py

@@ -65,21 +65,27 @@ def cwl_version(yaml: Any) -> Any:
     return yaml["cwlVersion"]
 
 
-def load_document_by_path(
+def load_document_by_uri(
     path: Union[str, Path],
     loadingOptions: Optional[LoadingOptions] = None,
 ) -> Any:
-    """Load a CWL object from a path."""
+    """Load a CWL object from a URI or a path."""
     if isinstance(path, str):
-        if path.startswith("file:/"):
-            path = unquote_plus(urlparse(path).path)
-        real_path = Path(path)
+        uri = urlparse(path)
+        if not uri.scheme or uri.scheme == "file":
+            real_path = Path(unquote_plus(uri.path)).resolve().as_uri()
+        else:
+            real_path = path
     else:
-        real_path = path
-    baseuri = real_path.resolve().as_uri()
-    with open(path) as handle:
-        doc = yaml_no_ts().load(handle)
-    return load_document_by_yaml(doc, baseuri, loadingOptions)
+        real_path = path.resolve().as_uri()
+
+    baseuri = str(real_path)
+
+    if loadingOptions is None:
+        loadingOptions = cwl_v1_2.LoadingOptions(fileuri=baseuri)
+
+    doc = loadingOptions.fetcher.fetch_text(real_path)
+    return load_document_by_string(doc, baseuri, loadingOptions)
 
 
 def load_document(

tests/load_cwl_by_path.py

@@ -4,13 +4,13 @@
 from ruamel import yaml
 import sys
 
-from cwl_utils.parser import load_document_by_path, save
+from cwl_utils.parser import load_document_by_uri, save
 
 # File Input - This is the only thing you will need to adjust or take in as an input to your function:
 cwl_file = Path("testdata/md5sum.cwl")  # or a plain string works as well
 
 # Import CWL Object
-cwl_obj = load_document_by_path(cwl_file)
+cwl_obj = load_document_by_uri(cwl_file)
 
 # View CWL Object
 print("List of object attributes:\n{}".format("\n".join(map(str, dir(cwl_obj)))))

tests/test_parser.py

@@ -1,10 +1,11 @@
 """Test the load and save functions for CWL."""
-from cwl_utils.parser import cwl_version, load_document, save
+from cwl_utils.parser import cwl_version, load_document, load_document_by_uri, save
 from pathlib import Path
 from ruamel import yaml
 
 HERE = Path(__file__).resolve().parent
 TEST_v1_0_CWL = HERE / "../testdata/md5sum.cwl"
+TEST_v1_0_CWL_REMOTE = "https://raw.githubusercontent.com/common-workflow-language/cwl-utils/main/testdata/md5sum.cwl"
 TEST_v1_2_CWL = HERE / "../testdata/workflow_input_format_expr_v1_2.cwl"
 
 
@@ -25,6 +26,22 @@ def test_load_document() -> None:
     assert cwl_obj.inputs[0].id.endswith("input_file")
 
 
+def test_load_document_with_local_uri() -> None:
+    """Test load_document for a CommandLineTool in a local URI."""
+    uri = Path(TEST_v1_0_CWL).resolve().as_uri()
+    assert uri.startswith("file://")
+    cwl_obj = load_document_by_uri(uri)
+    assert cwl_obj.cwlVersion == "v1.0"
+    assert cwl_obj.inputs[0].id.endswith("input_file")
+
+
+def test_load_document_with_remote_uri() -> None:
+    """Test load_document for a CommandLineTool in a remote URI."""
+    cwl_obj = load_document_by_uri(TEST_v1_0_CWL_REMOTE)
+    assert cwl_obj.cwlVersion == "v1.0"
+    assert cwl_obj.inputs[0].id.endswith("input_file")
+
+
 def test_save() -> None:
     """Test save for a list of Process objects with different cwlVersions."""
     with open(TEST_v1_0_CWL, "r") as cwl_h: