mirror of
https://github.com/yusufipk/dikte.git
synced 2026-09-11 19:06:11 +00:00
Everything the last commit built assumed one agent was installed, which is a poor assumption to bake into a dictation tool. So the provider is a setting, and what it selects is one of three quite different things. Claude Code and Codex are the same shape: a CLI, streaming JSONL, a session id to resume, tools that reach the machine and whatever is connected to it. They share the runner. What differs is spelled out where it differs, which is more than the flag names: Codex has no system prompt to append, so the instruction rides in front of the command with a rule between them; it confines its commands in a sandbox rather than asking about them, so the permission setting is a sandbox mode; and `-s` is not accepted by `exec resume`, so both settings go through `-c` overrides, which are. OpenRouter is the odd one and is meant to be. No tools, no files, no calendar: it can say what the capital of Peru is and not what is in your diary, and the settings box says so rather than letting it be discovered. It also has no session to resume, so the conversation is kept here and resent, capped at 24 messages. A stored conversation names the provider that made it, and is ignored by any other: none of them can pick up another's thread, and a stale id would otherwise fail every command until the timeout cleared it. The interface calls the thing by its name, which in Turkish means the suffix has to agree with it: Claude'a but Codex'e, Claude'u but Codex'i. A name dropped into a sentence through t() cannot be inflected by that sentence, so it arrives inflected, from a small table in i18n. English takes the name as it is and keeps the preposition in the sentence.
186 lines
6.7 KiB
Python
186 lines
6.7 KiB
Python
"""The dictation chain: transcribe → clean up → clipboard → paste.
|
|
|
|
The same chain also carries the other thing a dictation can be. Asked to, it
|
|
hands the transcript to Claude Code instead of pasting it, and pastes back
|
|
whatever came of it: an answer to a question, or a sentence saying what was
|
|
done.
|
|
"""
|
|
|
|
import os
|
|
import shutil
|
|
import sys
|
|
import threading
|
|
import time
|
|
import traceback
|
|
|
|
from PyQt6.QtCore import QObject, pyqtSignal
|
|
|
|
import api
|
|
import assistant
|
|
import audio
|
|
import config as cfg
|
|
import i18n
|
|
import paste
|
|
import vad
|
|
from i18n import t
|
|
|
|
CHUNK_SECONDS = audio.CHUNK_FRAMES / audio.RATE
|
|
|
|
|
|
class Pipeline(QObject):
|
|
stage = pyqtSignal(str) # human-readable progress line
|
|
finished = pyqtSignal(str, str, str) # raw transcript, final text, warning
|
|
failed = pyqtSignal(str)
|
|
cancelled = pyqtSignal()
|
|
|
|
def __init__(self, conf, parent=None):
|
|
super().__init__(parent)
|
|
self.conf = conf
|
|
self._thread = None
|
|
self._stop = threading.Event()
|
|
|
|
@property
|
|
def busy(self):
|
|
return self._thread is not None and self._thread.is_alive()
|
|
|
|
def run(self, wav_path, duration, rms_values=(), ask=False):
|
|
if self.busy:
|
|
return
|
|
self._stop.clear()
|
|
self._thread = threading.Thread(
|
|
target=self._work, args=(wav_path, duration, list(rms_values), ask),
|
|
daemon=True,
|
|
)
|
|
self._thread.start()
|
|
|
|
def cancel(self):
|
|
"""Give up on a job already under way.
|
|
|
|
Only the Claude call can honour this, and it is the only one long enough
|
|
to be worth interrupting: a transcription is over in seconds, a command
|
|
that went looking through the web is not.
|
|
"""
|
|
self._stop.set()
|
|
|
|
def _work(self, wav_path, duration, rms_values, ask):
|
|
conf = self.conf
|
|
started = time.monotonic()
|
|
raw = ""
|
|
|
|
# Room tone only: don't spend an API call, and don't invite a
|
|
# hallucinated sentence back.
|
|
if conf["skip_silent"]:
|
|
stats = vad.analyse(rms_values, CHUNK_SECONDS, conf["speech_margin_db"])
|
|
if vad.is_silent(stats, conf["silence_db"], conf["speech_margin_db"],
|
|
conf["min_voiced_seconds"]):
|
|
self._discard(wav_path)
|
|
self.failed.emit(
|
|
t("No speech detected ({level} dB)", level=round(stats["speech_db"]))
|
|
)
|
|
return
|
|
|
|
try:
|
|
self.stage.emit(t("Transcribing…"))
|
|
target = conf.transcribe_target()
|
|
raw = api.transcribe(
|
|
target,
|
|
wav_path,
|
|
language=conf["language"],
|
|
prompt=conf["transcribe_prompt"],
|
|
)
|
|
|
|
if conf["filter_hallucinations"] and vad.looks_like_hallucination(raw, duration):
|
|
self._discard(wav_path)
|
|
self.failed.emit(t("Discarded a stock phrase: “{text}”", text=raw[:60]))
|
|
return
|
|
|
|
text = raw
|
|
warning = ""
|
|
# Claude reads through “eee” and “hani” without help, so a dictation
|
|
# on its way there is normally sent as it was heard, one API call and
|
|
# a second or two lighter.
|
|
if (conf["assistant_cleanup"] if ask else conf["cleanup_enabled"]):
|
|
self.stage.emit(t("Cleaning up…"))
|
|
try:
|
|
text = api.cleanup(
|
|
raw,
|
|
conf.openrouter_key(),
|
|
conf["cleanup_model"],
|
|
conf.cleanup_prompt(),
|
|
reasoning=conf["cleanup_reasoning"],
|
|
base_url=conf["openrouter_base_url"],
|
|
)
|
|
except api.ApiError as exc:
|
|
# Keep the transcript, but never let the failure pass unseen:
|
|
# a rejected key would otherwise look like working dictation.
|
|
text = raw
|
|
warning = str(exc)
|
|
print(f"dikte: cleanup failed: {exc}", file=sys.stderr)
|
|
|
|
question = ""
|
|
if ask:
|
|
question = text
|
|
self.stage.emit(t("Asking {name}…", name=i18n.name(
|
|
assistant.display_name(conf), "dative")))
|
|
text, denied = assistant.ask(
|
|
question, conf,
|
|
on_stage=self.stage.emit,
|
|
should_stop=self._stop.is_set,
|
|
)
|
|
warning = "\n".join(x for x in (warning, denied) if x)
|
|
|
|
previous = paste.read_clipboard() if conf["restore_clipboard"] else None
|
|
paste.copy(text)
|
|
|
|
if (conf["assistant_paste"] if ask else conf["auto_paste"]):
|
|
self.stage.emit(t("Pasting…"))
|
|
paste.press(conf["paste_shortcut"])
|
|
if previous is not None:
|
|
time.sleep(0.35)
|
|
paste.copy_bytes(previous)
|
|
|
|
cfg.append_history({
|
|
"ts": time.strftime("%Y-%m-%d %H:%M:%S"),
|
|
"duration": round(duration, 1),
|
|
"elapsed": round(time.monotonic() - started, 1),
|
|
"model": target.model,
|
|
"cleanup_model": conf["cleanup_model"] if conf["cleanup_enabled"] else "",
|
|
"cleanup_error": warning,
|
|
"mode": "ask" if ask else "",
|
|
"question": question,
|
|
"assistant_model": conf["assistant_model"] if ask else "",
|
|
"raw": raw,
|
|
"text": text,
|
|
})
|
|
try:
|
|
cfg.trim_history(conf["history_limit"])
|
|
except OSError as exc:
|
|
print(f"dikte: could not trim the history: {exc}", file=sys.stderr)
|
|
self.finished.emit(raw, text, warning)
|
|
|
|
except assistant.Cancelled:
|
|
self.cancelled.emit()
|
|
except (api.ApiError, paste.PasteError, assistant.AssistantError) as exc:
|
|
print(f"dikte: {exc}", file=sys.stderr)
|
|
self.failed.emit(str(exc))
|
|
except Exception as exc: # never fail silently
|
|
traceback.print_exc()
|
|
self.failed.emit(t("Unexpected error: {error}", error=exc))
|
|
finally:
|
|
self._discard(wav_path)
|
|
|
|
def _discard(self, wav_path):
|
|
if not os.path.exists(wav_path):
|
|
return
|
|
if self.conf["keep_audio"]:
|
|
try:
|
|
cfg.RECORDINGS_DIR.mkdir(parents=True, exist_ok=True)
|
|
shutil.move(wav_path, cfg.RECORDINGS_DIR / (time.strftime("%Y%m%d-%H%M%S") + ".wav"))
|
|
return
|
|
except OSError:
|
|
pass
|
|
try:
|
|
os.unlink(wav_path)
|
|
except OSError:
|
|
pass
|