apacheGH-43097: [C++] Implement PathFromUri support for Azure file …

…system (apache#43098) ### Rationale for this change See apache#43097. ### What changes are included in this PR? Implements `AzureFS::PathFromUri` using existing URI parsing and path extraction inside the `AzureOptions`. ### Are these changes tested? Yes, added a unit test. ### Are there any user-facing changes? No, but calling `PathFromUri` will now work instead of throwing due to no implementation provided. * GitHub Issue: apache#43097 Authored-by: Oliver Layer <[email protected]> Signed-off-by: Sutou Kouhei <[email protected]>
vibhatha · Aug 14, 2024 · 88e8140 · 88e8140
1 parent fc80d7d
commit 88e8140
Show file tree

Hide file tree

Showing 3 changed files with 38 additions and 0 deletions.
diff --git a/cpp/src/arrow/filesystem/azurefs.cc b/cpp/src/arrow/filesystem/azurefs.cc
@@ -3199,4 +3199,31 @@ Result<std::shared_ptr<io::OutputStream>> AzureFileSystem::OpenAppendStream(
   return impl_->OpenAppendStream(location, metadata, false, this);
 }
 
+Result<std::string> AzureFileSystem::PathFromUri(const std::string& uri_string) const {
+  /// We can not use `internal::PathFromUriHelper` here because for Azure we have to
+  /// support different URI schemes where the authority is handled differently.
+  /// Example (both should yield the same path `container/some/path`):
+  ///   - (1) abfss://storageacc.blob.core.windows.net/container/some/path
+  ///   - (2) abfss://acc:pw@container/some/path
+  /// The authority handling is different with these two URIs. (1) requires no prepending
+  /// of the authority to the path, while (2) requires to preprend the authority to the
+  /// path.
+  std::string path;
+  Uri uri;
+  RETURN_NOT_OK(uri.Parse(uri_string));
+  RETURN_NOT_OK(AzureOptions::FromUri(uri, &path));
+
+  std::vector<std::string> supported_schemes = {"abfs", "abfss"};
+  const auto scheme = uri.scheme();
+  if (std::find(supported_schemes.begin(), supported_schemes.end(), scheme) ==
+      supported_schemes.end()) {
+    std::string expected_schemes =
+        ::arrow::internal::JoinStrings(supported_schemes, ", ");
+    return Status::Invalid("The filesystem expected a URI with one of the schemes (",
+                           expected_schemes, ") but received ", uri_string);
+  }
+
+  return path;
+}
+
 }  // namespace arrow::fs
diff --git a/cpp/src/arrow/filesystem/azurefs.h b/cpp/src/arrow/filesystem/azurefs.h
@@ -367,6 +367,8 @@ class ARROW_EXPORT AzureFileSystem : public FileSystem {
   Result<std::shared_ptr<io::OutputStream>> OpenAppendStream(
       const std::string& path,
       const std::shared_ptr<const KeyValueMetadata>& metadata) override;
+
+  Result<std::string> PathFromUri(const std::string& uri_string) const override;
 };
 
 }  // namespace arrow::fs
diff --git a/cpp/src/arrow/filesystem/azurefs_test.cc b/cpp/src/arrow/filesystem/azurefs_test.cc
@@ -2958,5 +2958,14 @@ TEST_F(TestAzuriteFileSystem, OpenInputFileClosed) {
   ASSERT_RAISES(Invalid, stream->ReadAt(1, 1));
   ASSERT_RAISES(Invalid, stream->Seek(2));
 }
+
+TEST_F(TestAzuriteFileSystem, PathFromUri) {
+  ASSERT_EQ(
+      "container/some/path",
+      fs()->PathFromUri("abfss://storageacc.blob.core.windows.net/container/some/path"));
+  ASSERT_EQ("container/some/path",
+            fs()->PathFromUri("abfss://acc:pw@container/some/path"));
+  ASSERT_RAISES(Invalid, fs()->PathFromUri("http://acc:pw@container/some/path"));
+}
 }  // namespace fs
 }  // namespace arrow