apunkt/dograh
1
0
Fork 0
mirror of https://github.com/dograh-hq/dograh.git synced 2026-07-10 11:12:13 +02:00
Code Issues Projects Releases Packages Wiki Activity

Feat/enhanced timestamped transcript (#501)

Browse source 
* feat: add additional timestamps in call transcript optionally

* fi: timestamp precision to millisec instead of micro

* fix: address enhanced transcript review issues

* fix: non vad user turn timestamp
This commit is contained in:
Sabiha Khan 2026-07-07 14:47:43 +05:30 • committed by GitHub
parent d9b9a1efc8
commit ac01f7775e
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
12 changed files with 441 additions and 15 deletions
Download patch file Download diff file Expand all files Collapse all files

23
api/utils/transcript.py
View file

@ -3,11 +3,26 @@ from typing import List
from pipecat.utils.enums import RealtimeFeedbackType
def generate_transcript_text(events: List[dict]) -> str:
def _format_timestamp_range(payload: dict, event: dict, include_end_timestamps: bool) -> str:
start_timestamp = payload.get("timestamp") or event.get("timestamp", "")
if not include_end_timestamps:
return start_timestamp
end_timestamp = payload.get("end_timestamp")
if end_timestamp:
return f"{start_timestamp} -> {end_timestamp}" if start_timestamp else end_timestamp
return start_timestamp
def generate_transcript_text(
events: List[dict], *, include_end_timestamps: bool = False
) -> str:
"""Generate transcript text from realtime feedback events.
Filters for rtf-user-transcription (final) and rtf-bot-text events,
formats them as '[timestamp] user/assistant: text\\n'.
formats them as '[timestamp] user/assistant: text\\n'. When
include_end_timestamps is True, formats as
'[start_timestamp -> end_timestamp] user/assistant: text\\n'.
"""
lines: List[str] = []
for event in events:
@ -18,11 +33,11 @@ def generate_transcript_text(events: List[dict]) -> str:
event_type == RealtimeFeedbackType.USER_TRANSCRIPTION.value
and payload.get("final") is True
):
timestamp = payload.get("timestamp") or event.get("timestamp", "")
timestamp = _format_timestamp_range(payload, event, include_end_timestamps)
prefix = f"[{timestamp}] " if timestamp else ""
lines.append(f"{prefix}user: {payload.get('text', '')}\n")
elif event_type == RealtimeFeedbackType.BOT_TEXT.value:
timestamp = payload.get("timestamp") or event.get("timestamp", "")
timestamp = _format_timestamp_range(payload, event, include_end_timestamps)
prefix = f"[{timestamp}] " if timestamp else ""
lines.append(f"{prefix}assistant: {payload.get('text', '')}\n")