feat(asr): Add profanity_filter in python-clients (#12)

* feat(asr): Add profanity_filter in python-clients

* update patch version for testing

* update pre_release version since 0.0.5 already exists in test.pypi.org

* set profanity_filter=True to make it discoverable

* update package version for testing

* Enable profanity_filter in other transcribe example files

* Update release version

* Older release doesn't work

* Pass profanity_filter value to RecognitionConfig

* remove PRE_RELEASE from VERSION

* revert patch no

Author: shikshapatel

riva/client/argparse_utils.py

@@ -5,7 +5,7 @@
 
 
 def add_asr_config_argparse_parameters(
-    parser: argparse.ArgumentParser, max_alternatives: bool = False, word_time_offsets: bool = False
+    parser: argparse.ArgumentParser, max_alternatives: bool = False, profanity_filter: bool = False, word_time_offsets: bool = False
 ) -> argparse.ArgumentParser:
     if word_time_offsets:
         parser.add_argument(
@@ -18,6 +18,13 @@ def add_asr_config_argparse_parameters(
             type=int,
             help="Maximum number of alternative transcripts to return (up to limit configured on server).",
         )
+    if profanity_filter:
+        parser.add_argument(
+        "--profanity-filter",
+        default=False,
+        action='store_true',
+        help="Flag that controls the profanity filtering in the generated transcripts",
+    )
     parser.add_argument(
         "--automatic-punctuation",
         default=False,

scripts/asr/riva_streaming_asr_client.py

@@ -38,7 +38,7 @@ def parse_args() -> argparse.Namespace:
         "--file-streaming-chunk", type=int, default=1600, help="Number of frames in one chunk sent to server."
     )
     parser = add_connection_argparse_parameters(parser)
-    parser = add_asr_config_argparse_parameters(parser, max_alternatives=True, word_time_offsets=True)
+    parser = add_asr_config_argparse_parameters(parser, max_alternatives=True, profanity_filter=True, word_time_offsets=True)
     args = parser.parse_args()
     if args.max_alternatives < 1:
         parser.error("`--max-alternatives` must be greater than or equal to 1")
@@ -57,6 +57,7 @@ def streaming_transcription_worker(
                 encoding=riva.client.AudioEncoding.LINEAR_PCM,
                 language_code=args.language_code,
                 max_alternatives=args.max_alternatives,
+                profanity_filter=args.profanity_filter,
                 enable_automatic_punctuation=args.automatic_punctuation,
                 verbatim_transcripts=not args.no_verbatim_transcripts,
                 enable_word_time_offsets=args.word_time_offsets,

scripts/asr/transcribe_file.py

@@ -49,7 +49,7 @@ def parse_args() -> argparse.Namespace:
         "--print-confidence", action="store_true", help="Whether to print stability and confidence of transcript."
     )
     parser = add_connection_argparse_parameters(parser)
-    parser = add_asr_config_argparse_parameters(parser)
+    parser = add_asr_config_argparse_parameters(parser, profanity_filter=True)
     args = parser.parse_args()
     if not args.list_devices and args.input_file is None:
         parser.error(
@@ -73,6 +73,7 @@ def main() -> None:
             encoding=riva.client.AudioEncoding.LINEAR_PCM,
             language_code=args.language_code,
             max_alternatives=1,
+            profanity_filter=args.profanity_filter,
             enable_automatic_punctuation=args.automatic_punctuation,
             verbatim_transcripts=not args.no_verbatim_transcripts,
         ),

scripts/asr/transcribe_file_offline.py

@@ -18,7 +18,7 @@ def parse_args() -> argparse.Namespace:
     )
     parser.add_argument("--input-file", required=True, type=Path, help="A path to a local file to transcribe.")
     parser = add_connection_argparse_parameters(parser)
-    parser = add_asr_config_argparse_parameters(parser)
+    parser = add_asr_config_argparse_parameters(parser, profanity_filter=True)
     args = parser.parse_args()
     args.input_file = args.input_file.expanduser()
     return args
@@ -32,6 +32,7 @@ def main() -> None:
         encoding=riva.client.AudioEncoding.LINEAR_PCM,
         language_code=args.language_code,
         max_alternatives=1,
+        profanity_filter=args.profanity_filter,
         enable_automatic_punctuation=args.automatic_punctuation,
         verbatim_transcripts=not args.no_verbatim_transcripts,
     )

scripts/asr/transcribe_mic.py

@@ -18,7 +18,7 @@ def parse_args() -> argparse.Namespace:
     )
     parser.add_argument("--input-device", type=int, default=default_device_index, help="An input audio device to use.")
     parser.add_argument("--list-devices", action="store_true", help="List input audio device indices.")
-    parser = add_asr_config_argparse_parameters(parser)
+    parser = add_asr_config_argparse_parameters(parser, profanity_filter=True)
     parser = add_connection_argparse_parameters(parser)
     parser.add_argument(
         "--sample-rate-hz",
@@ -48,6 +48,7 @@ def main() -> None:
             encoding=riva.client.AudioEncoding.LINEAR_PCM,
             language_code=args.language_code,
             max_alternatives=1,
+            profanity_filter=args.profanity_filter,
             enable_automatic_punctuation=args.automatic_punctuation,
             verbatim_transcripts=not args.no_verbatim_transcripts,
             sample_rate_hertz=args.sample_rate_hz,