nsinnovations/fusionpbx
2
0
Fork 0
mirror of https://github.com/fusionpbx/fusionpbx.git synced 2025-12-30 00:53:50 +00:00
Code Issues Packages Projects Releases Wiki Activity

Update record_message.lua (#4762)

Browse Source 
IBM Watson supports MP3 transcription and my testing shows it is very similar to WAV in terms of overall quality.
The benefit, of course is it greatly reduces the voicemail file size.
If MP3 is not set for the system then it will use WAV.
If other users prefer having the option of MP3 for the system but WAV for Watson transcription we could add a new variable for Watson transcription and add this new variable as a qualifier.
This commit is contained in:
Andy-Seattle
2019-10-09 11:13:29 -07:00
committed by FusionPBX
parent a5239443d3
commit cc4494eb15
1 changed files with 12 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
resources/install/scripts/app/voicemail/resources/functions/record_message.lua
View File

@@ -214,7 +214,11 @@
local api_key = settings:get('voicemail', 'watson_key', 'text') or '';
local transcription_server = settings:get('voicemail', 'watson_url', 'text') or '';
if (api_key ~= '') then
transcribe_cmd = [[ curl -X POST -u "apikey:]]..api_key..[[" --header "Content-type: audio/wav" --data-binary @]]..file_path..[[ "]]..transcription_server..[[" ]]
if (vm_message_ext == "mp3") then
transcribe_cmd = [[ curl -X POST -u "apikey:]]..api_key..[[" --header "Content-type: audio/mp3" --data-binary @]]..file_path..[[ "]]..transcription_server..[[" ]]
else
transcribe_cmd = [[ curl -X POST -u "apikey:]]..api_key..[[" --header "Content-type: audio/wav" --data-binary @]]..file_path..[[ "]]..transcription_server..[[" ]]
end
local handle = io.popen(transcribe_cmd);
local transcribe_result = handle:read("*a");
handle:close();
@@ -330,7 +334,8 @@
local message_silence_threshold = settings:get('voicemail', 'message_silence_threshold', 'numeric') or 200;
local message_silence_seconds = settings:get('voicemail', 'message_silence_seconds', 'numeric') or 3;
transcribe_enabled = settings:get('voicemail', 'transcribe_enabled', 'boolean') or "false";
local transcribe_provider = settings:get('voicemail', 'transcribe_provider', 'text') or '';
--debug information
if (debug["info"]) then
freeswitch.consoleLog("notice", "[voicemail] transcribe_enabled: " .. transcribe_enabled .. "\n");
@@ -468,8 +473,12 @@
shout_exists = trim(api:execute("module_exists", "mod_shout"));
if (shout_exists == "true" and transcribe_enabled == "false") or (shout_exists == "true" and transcribe_enabled == "true" and voicemail_transcription_enabled ~= "true") then
freeswitch.consoleLog("notice", "using mod_shout for mp3 encoding\n");
--record in mp3 directly
--record in mp3 directly, no transcription
result = session:recordFile(voicemail_dir.."/"..voicemail_id.."/msg_"..uuid..".mp3", message_max_length, message_silence_threshold, message_silence_seconds);
elseif (shout_exists == "true" and transcribe_enabled == "true" and voicemail_transcription_enabled == "true" and transcribe_provider == "watson") then
--record in mp3 directly with mp3 transcription if watson selected
result = session:recordFile(voicemail_dir.."/"..voicemail_id.."/msg_"..uuid..".mp3", message_max_length, message_silence_threshold, message_silence_seconds);
transcription = transcribe(voicemail_dir.."/"..voicemail_id.."/msg_"..uuid..".mp3",settings,start_epoch);
else
--create initial wav recording
result = session:recordFile(voicemail_dir.."/"..voicemail_id.."/msg_"..uuid..".wav", message_max_length, message_silence_threshold, message_silence_seconds);