Dymas/abogen
1
0
Fork 0
mirror of https://github.com/denizsafak/abogen.git synced 2026-09-20 11:40:57 +02:00
Code Issues Packages Projects Releases Wiki Activity

refactor: SynthParams dataclass for synthesize_text

Browse Source 
- Frozen dataclass in domain/conversion_engine.py with common params
- synthesize_text now takes params=SynthParams + unique kwargs
- Executor, PyQt legacy, WebUI legacy, and tests updated
- Adding new common params now only requires changing the dataclass
This commit is contained in:
Artem Akymenko
2026-07-22 08:21:45 +00:00
parent 93f5a46485
commit c4cebb8822
5 changed files with 130 additions and 110 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/test_conversion_executor.py
+57 -32
View File
@@ -16,6 +16,7 @@ from typing import Any
from abogen.domain.conversion_engine import (
synthesize_text,
SynthParams,
run_tts_segment_loop,
process_and_write_subtitles,
SegmentStats,
@@ -133,18 +134,22 @@ class TestSynthesizeText:
def on_progress(pct, etr):
progress_calls.append((pct, etr))
segments, tokens = synthesize_text(
text="Hello world.",
params = SynthParams(
tts_context=tts_ctx,
backend=backend,
voice="M1",
speed=1.0,
stats=stats,
check_cancel=cancel,
on_progress=on_progress,
audio_sink=sink,
)
segments, tokens = synthesize_text(
text="Hello world.",
params=params,
backend=backend,
voice="M1",
speed=1.0,
)
assert segments >= 1
assert len(sink.written) >= 1
assert len(progress_calls) >= 1
@@ -159,17 +164,21 @@ class TestSynthesizeText:
def on_progress(pct, etr):
progress_calls.append((pct, etr))
segments, tokens = synthesize_text(
text="Hello world.",
params = SynthParams(
tts_context=tts_ctx,
backend=backend,
voice="M1",
speed=1.0,
stats=stats,
check_cancel=cancel,
on_progress=on_progress,
)
segments, tokens = synthesize_text(
text="Hello world.",
params=params,
backend=backend,
voice="M1",
speed=1.0,
)
# Should stop early due to cancellation
assert segments == 0
@@ -184,19 +193,23 @@ class TestSynthesizeText:
def on_progress(pct, etr):
pass
segments, tokens = synthesize_text(
text="Hello world.",
params = SynthParams(
tts_context=tts_ctx,
backend=backend,
voice="M1",
speed=1.0,
stats=stats,
check_cancel=cancel,
on_progress=on_progress,
chapter_sink=chapter_sink,
audio_sink=merged_sink,
)
segments, tokens = synthesize_text(
text="Hello world.",
params=params,
backend=backend,
voice="M1",
speed=1.0,
chapter_sink=chapter_sink,
)
# Both sinks should receive audio
assert len(chapter_sink.written) >= 1
assert len(merged_sink.written) >= 1
@@ -210,15 +223,19 @@ class TestSynthesizeText:
def on_progress(pct, etr):
pass
segments, tokens = synthesize_text(
text="Hello world.",
params = SynthParams(
tts_context=tts_ctx,
backend=backend,
voice="M1",
speed=1.0,
stats=stats,
check_cancel=cancel,
on_progress=on_progress,
)
segments, tokens = synthesize_text(
text="Hello world.",
params=params,
backend=backend,
voice="M1",
speed=1.0,
split_pattern_override=r"\n+",
)
@@ -323,16 +340,11 @@ class TestFullPipeline:
progress_calls.append((pct, etr))
# Simulate full pipeline: synthesize → subtitles → finalize
segments, tokens = synthesize_text(
text="This is a test sentence. Another sentence here.",
params = SynthParams(
tts_context=tts_ctx,
backend=backend,
voice="M1",
speed=1.0,
stats=stats,
check_cancel=cancel,
on_progress=on_progress,
chapter_sink=chapter_sink,
audio_sink=merged_sink,
subtitle_mode="Sentence",
max_subtitle_words=5,
@@ -340,6 +352,15 @@ class TestFullPipeline:
use_spacy_segmentation=False,
)
segments, tokens = synthesize_text(
text="This is a test sentence. Another sentence here.",
params=params,
backend=backend,
voice="M1",
speed=1.0,
chapter_sink=chapter_sink,
)
# Process accumulated tokens
process_and_write_subtitles(
tokens,
@@ -373,16 +394,20 @@ class TestFullPipeline:
def on_progress(pct, etr):
progress_calls.append((pct, etr))
segments, tokens = synthesize_text(
text="Hello world. " * 100,
params = SynthParams(
tts_context=tts_ctx,
backend=backend,
voice="M1",
speed=1.0,
stats=stats,
check_cancel=cancel_fn,
on_progress=on_progress,
)
segments, tokens = synthesize_text(
text="Hello world. " * 100,
params=params,
backend=backend,
voice="M1",
speed=1.0,
)
# Should have stopped before processing all text
assert segments <= 4