Merge branch 'master' into hotfix/multichannel_version_increment

Author: James Walker

.github/workflows/tests.yml

@@ -22,7 +22,7 @@ jobs:
         python-version: ${{ matrix.python-version }}
 
     - name: Install dependencies
-      run: pip3 install -r requirements.txt -r requirements-dev.txt
+      run: pip3 install -r requirements.txt -r requirements-dev.txt -r requirements-metrics.txt
 
     - name: Lint
       run: make lint

.gitignore

@@ -87,6 +87,7 @@ target/
 # pyenv
 .python-version
 venv
+.venv
 
 # pipenv
 #   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.

CHANGELOG.md

@@ -4,11 +4,41 @@ All notable changes to this project will be documented in this file.
 
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+## [4.0.1] - 2025-04-30
+
+### Added 
+
+- Support RT Multichannel and channel DZ
+
+## [4.0.0] - 2025-05-20
+
+### Changed
+
+BREAKING CHANGE: Metrics functionality now requires explicit installation
+
+Previously, all metrics dependencies (pyannote, pandas, jiwer, etc.) were
+installed by default. This change moves them to an optional '[metrics]' extra
+to reduce the default installation footprint.
+
+- Move metrics dependencies to requirements-metrics.txt
+- Configure extras_require in setup.py for optional installation
+- Add graceful error handling in CLI when dependencies are missing
+
+## [3.0.6] - 2025-05-20
+
+### Added
+
+- Support end-of-utterance messages (DEL-24982)
+
+## [3.0.5] - 2025-05-15
+
+- cli: fix some config options not being set when defined in a config file: `topic_detection_config` and `speaker_diarization_config`
 
 ## [3.0.4] - 2025-04-16
 
 - Support for new parameters `prefer_current_speaker` and `speaker_sensitivity` in Speaker Diarization
 
+
 ## [3.0.3] - 2025-03-03
 
 ### Added

Makefile

@@ -9,6 +9,11 @@ lint:
 	black --check --diff $(SOURCES)
 	ruff $(SOURCES)
 
+.PHONY: lint-fix
+lint-fix:
+	black $(SOURCES)
+	ruff --fix $(SOURCES)
+
 .PHONY: format
 format:
 	black $(SOURCES)

README.md

@@ -9,6 +9,11 @@ To install from PyPI:
 ```bash
 pip install speechmatics-python
 ```
+
+To use the sm-metrics tool for diarization features and speaker identification metrics, install with the optional dependencies:
+```bash
+pip install speechmatics-python[metrics]
+```
 To install from source:
 ```bash
 git clone https://github.com/speechmatics/speechmatics-python

asr_metrics/cli.py

@@ -1,33 +1,74 @@
 """Entrypoint for SM metrics"""
+
 import argparse
+import sys
+
+try:
+    import asr_metrics.wer.__main__ as wer_metrics
+
+    WER_AVAILABLE = True
+except ImportError:
+    WER_AVAILABLE = False
+
+try:
+    import asr_metrics.diarization.sm_diarization_metrics.cookbook as diarization_metrics
 
-import asr_metrics.diarization.sm_diarization_metrics.cookbook as diarization_metrics
-import asr_metrics.wer.__main__ as wer_metrics
+    DIARIZATION_AVAILABLE = True
+except ImportError:
+    DIARIZATION_AVAILABLE = False
 
 
 def main():
-    parser = argparse.ArgumentParser(description="Your CLI description")
+    parser = argparse.ArgumentParser(
+        description="Speechmatics metrics tool for WER and diarization"
+    )
 
     # Create subparsers
     subparsers = parser.add_subparsers(
         dest="mode", help="Metrics mode. Choose from 'wer' or 'diarization'"
     )
-    subparsers.required = True  # Make sure a subparser id always provided
+    subparsers.required = True  # Make sure a subparser is always provided
 
-    wer_parser = subparsers.add_parser("wer", help="Entrypoint for WER metrics")
-    wer_metrics.get_wer_args(wer_parser)
+    if WER_AVAILABLE:
+        wer_parser = subparsers.add_parser("wer", help="Entrypoint for WER metrics")
+        wer_metrics.get_wer_args(wer_parser)
+    else:
+        wer_parser = subparsers.add_parser(
+            "wer", help="Entrypoint for WER metrics (requires additional dependencies)"
+        )
 
-    diarization_parser = subparsers.add_parser(
-        "diarization", help="Entrypoint for diarization metrics"
-    )
-    diarization_metrics.get_diarization_args(diarization_parser)
+    if DIARIZATION_AVAILABLE:
+        diarization_parser = subparsers.add_parser(
+            "diarization", help="Entrypoint for diarization metrics"
+        )
+        diarization_metrics.get_diarization_args(diarization_parser)
+    else:
+        diarization_parser = subparsers.add_parser(
+            "diarization",
+            help="Entrypoint for diarization metrics (requires pyannote dependencies)",
+        )
+        diarization_parser.add_argument(
+            "--help-install",
+            action="store_true",
+            help="Show instructions for installing diarization dependencies",
+        )
 
     args = parser.parse_args()
 
     if args.mode == "wer":
-        wer_metrics.main(args)
+        if WER_AVAILABLE:
+            wer_metrics.main(args)
+        else:
+            print("Error: WER metrics require additional dependencies.")
+            print("Please install them with: pip install speechmatics-python[metrics]")
+            sys.exit(1)
     elif args.mode == "diarization":
-        diarization_metrics.main(args)
+        if DIARIZATION_AVAILABLE:
+            diarization_metrics.main(args)
+        else:
+            print("Error: Diarization metrics require additional dependencies.")
+            print("Please install them with: pip install speechmatics-python[metrics]")
+            sys.exit(1)
     else:
         print("Unsupported mode. Please use 'wer' or 'diarization'")
 

asr_metrics/wer/__main__.py

@@ -2,6 +2,7 @@
 Simple script to run WER analysis using Whisper normalisers
 Prints results to terminal
 """
+
 import difflib
 import json
 from pathlib import Path

asr_metrics/wer/normalizers/basic.py

@@ -50,15 +50,19 @@ def remove_symbols_and_diacritics(self, s: str, keep=""):
         and drop any diacritics (category 'Mn' and some manual mappings)
         """
         return "".join(
-            c
-            if c in keep
-            else self.additional_diacritics[c]
-            if c in self.additional_diacritics
-            else ""
-            if unicodedata.category(c) == "Mn"
-            else " "
-            if unicodedata.category(c)[0] in "MSP"
-            else c
+            (
+                c
+                if c in keep
+                else (
+                    self.additional_diacritics[c]
+                    if c in self.additional_diacritics
+                    else (
+                        ""
+                        if unicodedata.category(c) == "Mn"
+                        else " " if unicodedata.category(c)[0] in "MSP" else c
+                    )
+                )
+            )
             for c in unicodedata.normalize("NFKD", s)
         )
 

examples/notification_flow/batch_transcription_client.py

@@ -1,6 +1,7 @@
 """
 Client module which calls the Speechmatics API
 """
+
 import sqlite3
 
 from speechmatics.batch_client import BatchClient

examples/notification_flow/cronjob.py

@@ -3,6 +3,7 @@
 
 Async fallback checking in case of webhook failure is a common pattern in production systems
 """
+
 import sqlite3
 
 from speechmatics.batch_client import BatchClient