Fixes to get voice working properly, not fully confident it'll work under pressure but it's done well in small tests

This commit is contained in:
2026-01-01 17:00:39 -05:00
parent 81b8085b51
commit 7df2bb6dfc
5 changed files with 39 additions and 15 deletions

View File

@@ -169,11 +169,11 @@ def mix_audio_with_ffmpeg(files: list[TranscriptableFile], media_folder_path: st
output_file_name = Path( output_file_name = Path(
media_folder_path, media_folder_path,
call_id, str(call_id),
"output.mp3" if is_final_pass else "intermediate-" + "".join(random.choices(string.ascii_uppercase + string.digits, k=10)) + ".mp3" "output.mp3" if is_final_pass else "intermediate-" + "".join(random.choices(string.ascii_uppercase + string.digits, k=10)) + ".mp3"
) )
command_list.append(output_file_name) command_list.append(str(output_file_name))
# TODO shell = True isn't great, I don't remember the reason why it has to be this way # TODO shell = True isn't great, I don't remember the reason why it has to be this way
# I *think* it had something to do with me not using ffmpeg's absolute path # I *think* it had something to do with me not using ffmpeg's absolute path

View File

@@ -2,7 +2,7 @@ import json
import os import os
import copy import copy
from pathlib import Path from pathlib import Path
from datetime import datetime from datetime import datetime, timezone
from breadbot_common import SQLite, MySQL, TranscriptableFile, mix_audio_with_ffmpeg from breadbot_common import SQLite, MySQL, TranscriptableFile, mix_audio_with_ffmpeg
from txtai.pipeline import Transcription from txtai.pipeline import Transcription
@@ -36,20 +36,25 @@ transcriber = Transcription("openai/whisper-base")
for call in calls_needing_work[1]: for call in calls_needing_work[1]:
all_files = os.listdir(Path( all_files = os.listdir(Path(
config_json["media_voice_folder"], config_json["media_voice_folder"],
call[0] str(call[0])
)) ))
transcriptable_files = [] transcriptable_files = []
for file in all_files: for file in all_files:
print(file)
file_name_no_extension = file.split('.')[0] file_name_no_extension = file.split('.')[0]
timestamp = int(file_name_no_extension.split('-')[0]) timestamp = int(file_name_no_extension.split('-')[0])
user_snowflake = file_name_no_extension.split('-')[1] user_snowflake = file_name_no_extension.split('-')[1]
file_stamp_as_datetime = datetime.fromtimestamp(timestamp / 1000) file_stamp_as_datetime = datetime.fromtimestamp(timestamp / 1000, timezone.utc)
time_diff = file_stamp_as_datetime - call[1] print(file_stamp_as_datetime)
print(type(call[1]))
print(call[1])
time_diff = file_stamp_as_datetime - datetime.fromisoformat(call[1] + 'Z')
print(time_diff)
transcriptable_files.append(TranscriptableFile( transcriptable_files.append(TranscriptableFile(
file_path = file, file_path = str(Path(config_json["media_voice_folder"], str(call[0]), file)),
real_date = file_stamp_as_datetime, real_date = file_stamp_as_datetime,
milliseconds_from_start = int((time_diff.seconds * 1000) + (time_diff.microseconds / 1000)), milliseconds_from_start = int((time_diff.seconds * 1000) + (time_diff.microseconds / 1000)),
user_snowflake = user_snowflake user_snowflake = user_snowflake
@@ -60,6 +65,11 @@ for call in calls_needing_work[1]:
# TODO Possibly RAM abusive solution to wanting to keep the original list around # TODO Possibly RAM abusive solution to wanting to keep the original list around
ffmpeg_files = copy.deepcopy(transcriptable_files) ffmpeg_files = copy.deepcopy(transcriptable_files)
for file in ffmpeg_files:
print(file.file_path)
print(file.real_date)
print(file.milliseconds_from_start)
# TODO Error handling for all ffmpeg operations # TODO Error handling for all ffmpeg operations
while len(ffmpeg_files) > MAX_FILES_PER_CYCLE: while len(ffmpeg_files) > MAX_FILES_PER_CYCLE:
ffmpeg_files = [ ffmpeg_files = [
@@ -84,9 +94,9 @@ for call in calls_needing_work[1]:
"compare": "=" "compare": "="
}]) }])
for file in os.listdir(Path(config_json["media_voice_folder"], call[0])): for file in os.listdir(Path(config_json["media_voice_folder"], str(call[0]))):
if file.startswith("intermediate"): if file.startswith("intermediate"):
os.remove(Path(config_json["media_voice_folder"], call[0], file)) os.remove(Path(config_json["media_voice_folder"], str(call[0]), file))
for file in transcriptable_files: for file in transcriptable_files:
text = transcriber(file.file_path) text = transcriber(file.file_path)

View File

@@ -22,6 +22,7 @@ import { DBCallTranscriptions } from "./utilties/storage/entities/DBCallTranscri
import { DBCallUsers } from "./utilties/storage/entities/DBCallUsers" import { DBCallUsers } from "./utilties/storage/entities/DBCallUsers"
import { DBMessageRegex } from "./utilties/storage/entities/DBMessageRegex" import { DBMessageRegex } from "./utilties/storage/entities/DBMessageRegex"
import { DBMessageRegexMatches } from "./utilties/storage/entities/DBMessageRegexMatches" import { DBMessageRegexMatches } from "./utilties/storage/entities/DBMessageRegexMatches"
import { setupVoice } from "./utilties/events/voice"
console.log(__dirname + path.sep + "utilities" + path.sep + "storage" + path.sep + "entities" + path.sep + "*.ts") console.log(__dirname + path.sep + "utilities" + path.sep + "storage" + path.sep + "entities" + path.sep + "*.ts")
@@ -68,6 +69,8 @@ client.once(Events.ClientReady, async () => {
const maRepo = dataSource.getRepository(DBMessageAttachments) const maRepo = dataSource.getRepository(DBMessageAttachments)
const regexesRepo = dataSource.getRepository(DBMessageRegex) const regexesRepo = dataSource.getRepository(DBMessageRegex)
const matchesRepo = dataSource.getRepository(DBMessageRegexMatches) const matchesRepo = dataSource.getRepository(DBMessageRegexMatches)
const callRepo = dataSource.getRepository(DBCall)
const callUserRepo = dataSource.getRepository(DBCallUsers)
client.guilds.cache.forEach(async (guild: Guild) => { client.guilds.cache.forEach(async (guild: Guild) => {
const server: DBServer | null = await insertGuild(serverRepo, guild) const server: DBServer | null = await insertGuild(serverRepo, guild)
@@ -116,7 +119,8 @@ client.once(Events.ClientReady, async () => {
setupRoleCapture(client, serverRepo, roleRepo) setupRoleCapture(client, serverRepo, roleRepo)
setupMessageCapture(client, serverRepo, channelRepo, userRepo, messageRepo, mccRepo, maRepo, regexesRepo, matchesRepo) setupMessageCapture(client, serverRepo, channelRepo, userRepo, messageRepo, mccRepo, maRepo, regexesRepo, matchesRepo)
setupVoice(client, callRepo, channelRepo, userRepo, callUserRepo)
console.log("Breadbot is Ready") console.log("Breadbot is Ready")
}) })

View File

@@ -40,7 +40,9 @@ export async function returnOrCreateNewCallID(db: Repository<DBCall>, channel: V
export async function setCallEndTime(db: Repository<DBCall>, channel: VoiceBasedChannel) : Promise<DBCall | null> { export async function setCallEndTime(db: Repository<DBCall>, channel: VoiceBasedChannel) : Promise<DBCall | null> {
const call: DBCall | null = await db.findOne({ const call: DBCall | null = await db.findOne({
"where": { "where": {
channel: channel, channel: {
channel_snowflake: channel.id
},
call_end_time: IsNull() call_end_time: IsNull()
} }
}) })

View File

@@ -25,9 +25,11 @@ export function setupVoice(client: Client, callDB: Repository<DBCall>, channelDB
let existingCallID : number | undefined = await getExistingCallID(callDB, newState.channel) let existingCallID : number | undefined = await getExistingCallID(callDB, newState.channel)
console.log(`Call ID Pre Existing Call Check: ${existingCallID}`)
if (existingCallID === undefined) { if (existingCallID === undefined) {
existingCallID = await returnOrCreateNewCallID(callDB, newState.channel) existingCallID = await returnOrCreateNewCallID(callDB, newState.channel)
console.log(`Call does not exist new callID value: ${existingCallID}`)
mkdirSync(config.MEDIA_VOICE_FOLDER + path.sep + existingCallID, {recursive: true}) mkdirSync(config.MEDIA_VOICE_FOLDER + path.sep + existingCallID, {recursive: true})
// TODO NULL armor here is probably just going to blow up the call to joinVoiceChannel with no error catching // TODO NULL armor here is probably just going to blow up the call to joinVoiceChannel with no error catching
@@ -99,6 +101,13 @@ export function setupVoice(client: Client, callDB: Repository<DBCall>, channelDB
return // If the user is breadbot, ignore and exit return // If the user is breadbot, ignore and exit
} }
const connection = getVoiceConnection(oldState.guild.id)
if(oldState !== null && oldState.member !== null && connection?.receiver.subscriptions.has(oldState.member.id)) {
console.log(`Remove receiver subscription ${connection.receiver.subscriptions.delete(oldState.member.id)}`)
}
const existingCall : number | undefined = await getExistingCallID(callDB, oldState.channel) const existingCall : number | undefined = await getExistingCallID(callDB, oldState.channel)
if (existingCall !== undefined && oldState.member) { if (existingCall !== undefined && oldState.member) {
@@ -107,12 +116,11 @@ export function setupVoice(client: Client, callDB: Repository<DBCall>, channelDB
const usersInCall: number = await numberOfUsersInCall(callUserDB, existingCall) const usersInCall: number = await numberOfUsersInCall(callUserDB, existingCall)
if (usersInCall == 0) { if (usersInCall == 0) {
const connection = getVoiceConnection(oldState.guild.id) connection?.destroy()
connection?.disconnect()
await setCallEndTime(callDB, oldState.channel) await setCallEndTime(callDB, oldState.channel)
} }
} }
} }
}) })
} }