Allow for filepaths to include (oras-project#161)

* Allow for filepaths to include

When playing with omlmd I tried to push a file which contained a ":" and
oras choked on it. I belive that you should check if the file exists and
only split off the last : looking for options, not the first colon.

Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Signed-off-by: tarilabs <matteo.mortari@gmail.com>

CHANGELOG.md

@@ -14,6 +14,7 @@ and **Merged pull requests**. Critical items to know are:
 The versions coincide with releases on pip. Only major versions will be released as tags on Github.
 
 ## [0.0.x](https://github.com/oras-project/oras-py/tree/main) (0.0.x)
+ - allow for filepaths to include `:` (0.2.22)
  - release request (0.2.21)
  - add missing basic auth data for request token function in token auth backend (0.2.2)
  - re-enable chunked upload (0.2.1)

oras/tests/test_provider.py

@@ -63,6 +63,65 @@ def test_annotated_registry_push(tmp_path, registry, credentials, target):
         )
 
 
+@pytest.mark.with_auth(False)
+def test_file_contains_column(tmp_path, registry, credentials, target):
+    """
+    Test for file containing column symbol
+    """
+    client = oras.client.OrasClient(hostname=registry, insecure=True)
+    artifact = os.path.join(here, "artifact.txt")
+    assert os.path.exists(artifact)
+
+    # file containing `:`
+    try:
+        contains_column = here / "some:file"
+        with open(contains_column, "w") as f:
+            f.write("hello world some:file")
+
+        res = client.push(files=[contains_column], target=target)
+        assert res.status_code in [200, 201]
+
+        files = client.pull(target, outdir=tmp_path / "download")
+        download = str(tmp_path / "download/some:file")
+        assert download in files
+        assert oras.utils.get_file_hash(
+            str(contains_column)
+        ) == oras.utils.get_file_hash(download)
+    finally:
+        contains_column.unlink()
+
+    # file containing `:` as prefix, pushed with type
+    try:
+        contains_column = here / ":somefile"
+        with open(contains_column, "w") as f:
+            f.write("hello world :somefile")
+
+        res = client.push(files=[f"{contains_column}:text/plain"], target=target)
+        assert res.status_code in [200, 201]
+
+        files = client.pull(target, outdir=tmp_path / "download")
+        download = str(tmp_path / "download/:somefile")
+        assert download in files
+        assert oras.utils.get_file_hash(
+            str(contains_column)
+        ) == oras.utils.get_file_hash(download)
+    finally:
+        contains_column.unlink()
+
+    # error: file does not exist
+    with pytest.raises(FileNotFoundError):
+        client.push(files=[".doesnotexist"], target=target)
+
+    with pytest.raises(FileNotFoundError):
+        client.push(files=[":doesnotexist"], target=target)
+
+    with pytest.raises(FileNotFoundError, match=r".*does:not:exists .*"):
+        client.push(files=["does:not:exists:text/plain"], target=target)
+
+    with pytest.raises(FileNotFoundError, match=r".*does:not:exists .*"):
+        client.push(files=["does:not:exists:text/plain+ext"], target=target)
+
+
 @pytest.mark.with_auth(False)
 def test_chunked_push(tmp_path, registry, credentials, target):
     """
@@ -130,12 +189,6 @@ def test_parse_manifest(registry):
     assert ref == "path/to/config"
     assert content_type == "application/vnd.oci.image.config.v1+json"
 
-    testref = "path/to/config:application/vnd.oci.image.config.v1+json:extra"
-    remote = oras.provider.Registry(hostname=registry, insecure=True)
-    ref, content_type = remote._parse_manifest_ref(testref)
-    assert ref == "path/to/config"
-    assert content_type == "application/vnd.oci.image.config.v1+json:extra"
-
     testref = "/dev/null:application/vnd.oci.image.manifest.v1+json"
     ref, content_type = remote._parse_manifest_ref(testref)
     assert ref == "/dev/null"

oras/utils/fileio.py

@@ -346,7 +346,8 @@ def split_path_and_content(ref: str) -> PathAndOptionalContent:
     : return: A Tuple of the path in the reference, and the content-type if one found,
               otherwise None.
     """
-    if ":" not in ref:
+
+    if os.path.exists(ref) or ":" not in ref:
         return PathAndOptionalContent(ref, None)
 
     if pathlib.Path(ref).drive:
@@ -370,5 +371,5 @@ def split_path_and_content(ref: str) -> PathAndOptionalContent:
             )
         return PathAndOptionalContent(ref, None)
     else:
-        path_content_list = ref.split(":", 1)
+        path_content_list = ref.rsplit(":", 1)
         return PathAndOptionalContent(path_content_list[0], path_content_list[1])

oras/version.py

@@ -2,7 +2,7 @@
 __copyright__ = "Copyright The ORAS Authors."
 __license__ = "Apache-2.0"
 
-__version__ = "0.2.21"
+__version__ = "0.2.22"
 AUTHOR = "Vanessa Sochat"
 EMAIL = "vsoch@users.noreply.github.com"
 NAME = "oras"