(beam) branch master updated: Fix for issue 34772 to include user provided providers (#35033)

[shunping]

This is an automated email from the ASF dual-hosted git repository.

shunping pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/beam.git


The following commit(s) were added to refs/heads/master by this push:
     new b69f23f6818 Fix for issue 34772 to include user provided providers 
(#35033)
b69f23f6818 is described below

commit b69f23f6818787cb2f0a9f02608228a07405c0c8
Author: Bhargavkonidena <bhargavkonid...@users.noreply.github.com>
AuthorDate: Mon Jun 2 18:14:26 2025 +0530

    Fix for issue 34772 to include user provided providers (#35033)
    
    * Fix for issue 34772 to include user provided providers
    
    * Fix CI checks
    
    * Incorporate review comments
---
 sdks/python/apache_beam/yaml/generate_yaml_docs.py | 13 ++++++++++++-
 1 file changed, 12 insertions(+), 1 deletion(-)

diff --git a/sdks/python/apache_beam/yaml/generate_yaml_docs.py 
b/sdks/python/apache_beam/yaml/generate_yaml_docs.py
index 592177dc2bc..ae1a04661e6 100644
--- a/sdks/python/apache_beam/yaml/generate_yaml_docs.py
+++ b/sdks/python/apache_beam/yaml/generate_yaml_docs.py
@@ -32,6 +32,7 @@ from apache_beam.utils.python_callable import 
PythonCallableWithSource
 from apache_beam.version import __version__ as beam_version
 from apache_beam.yaml import json_utils
 from apache_beam.yaml import yaml_provider
+from apache_beam.yaml import yaml_utils
 from apache_beam.yaml.yaml_errors import ErrorHandlingConfig
 
 
@@ -263,7 +264,12 @@ def create_index(include, exclude, options):
   with subprocess_server.SubprocessServer.cache_subprocesses():
     json_config_schemas = []
     markdown_out = io.StringIO()
-    providers = yaml_provider.standard_providers()
+    if options.provider_source_file:
+      providers = yaml_provider.merge_providers(
+          yaml_provider.load_providers(
+              yaml_utils.locate_data_file(options.provider_source_file)))
+    else:
+      providers = yaml_provider.standard_providers()
     for transform_base, transforms in itertools.groupby(
         sorted(providers.keys(), key=io_grouping_key),
         key=lambda s: s.split('-')[0]):
@@ -557,6 +563,11 @@ def main():
   parser.add_argument('--schema_file')
   parser.add_argument('--include', default='.*')
   parser.add_argument('--exclude', default='')
+  parser.add_argument(
+      "--provider_source_file",
+      help="Path to a YAML file containing custom providers. "
+      "If not provided, uses standard Beam providers.",
+  )
   options = parser.parse_args()
   include = re.compile(options.include).match
   exclude = (