No replies to this topic

[the bigg]

This is a tp2 designed to allow modders to quickly add voice over to .tra lines (or entire files/mods) that previously lacked them. Specifically, it does the following two things for you:
1) translate generic [character] markers into numbered .wav references, listing the lines to be voiced to a separate file (that is, you can write [STIVAN] as the sound for all lines by Stivan and [Bob] as the sound for all lines by Bob, and it will be translated respectively to [TB#SS001]...[TB#SS999] and [TB#BOB01]..[TB#BOB99].
2) copy any new [sound] declarations from a 'master' language to the other languages the mod is already translated to (independently of whether you added a sound file or it was translated by the former functionality).

Code:

Spoiler

// Customize these definitions

<<<<<<<< names-to-soundsets
Description prefix
STIVAN TB#SS
ROSALINA TB#SRO
CARLINO TB#SCA
>>>>>>>>

OUTER_TEXT_SPRINT ~source~ ~stivan/tra/english~
ACTION_DEFINE_ARRAY ~early~ BEGIN ~setup.tra~ END
ACTION_DEFINE_ARRAY ~langs~ BEGIN ~stivan/tra/italian~ END
OUTER_TEXT_SPRINT ~sound_reference~ ~to_be_merged/sounds_to_record~

// stop altering here

DEFINE_PATCH_MACRO read BEGIN
	REPLACE_EVALUATE "\(@[0-9]* *\)= \(~[^~]*~\) \[\([^]]*\)\]" BEGIN
		TEXT_SPRINT result ~%MATCH1%= %MATCH2% [%MATCH3%]~
		TEXT_SPRINT EVALUATE_BUFFER ~%SOURCE_RES% %MATCH1%~ ~%MATCH3%~
		PATCH_IF VARIABLE_IS_SET $names(EVALUATE_BUFFER ~%MATCH3%~) THEN BEGIN
			i = $last(EVALUATE_BUFFER ~%MATCH3%~)
			i += 1
			SET $last(EVALUATE_BUFFER ~%MATCH3%~) = i
			TEXT_SPRINT str ~%i%~
			SET length = $sizes(EVALUATE_BUFFER ~%MATCH3%~)
			TEXT_SPRINT prefix $names(EVALUATE_BUFFER ~%MATCH3%~)
			WHILE STRING_LENGTH ~%str%~ < length BEGIN
				TEXT_SPRINT str ~0%str%~
			END
			TEXT_SPRINT EVALUATE_BUFFER ~%SOURCE_RES% %MATCH3%~ ~%prefix%%str%~
			TEXT_SPRINT result ~%MATCH1%= %MATCH2% [%prefix%%str%]~
			
			INNER_ACTION BEGIN
				APPEND_OUTER - ~%MATCH3%.tmp~ ~[%prefix%%str%] = %MATCH2%~
			END
		END
	END ~%result%~
END

<<<<<<<< empty
>>>>>>>>
COPY - ~names-to-soundsets~ ~names-to-soundsets~
	READ_2DA_ENTRIES_NOW names2da 2
	FOR (i = 1; i < names2da; ++i) BEGIN
		READ_2DA_ENTRY_FORMER names2da i 0 name
		READ_2DA_ENTRY_FORMER names2da i 1 prefix
		TEXT_SPRINT $names(EVALUATE_BUFFER ~%name%~) ~%prefix%~
		SET $sizes(EVALUATE_BUFFER ~%name%~) = 8 - STRING_LENGTH ~%prefix%~
		SET $last(EVALUATE_BUFFER ~%name%~) = 0
		INNER_ACTION BEGIN
			COPY - ~empty~ ~%name%.tmp~
			
			ACTION_IF ! FILE_EXISTS ~%sound_reference%/%name%.txt~ THEN BEGIN
				COPY ~empty~ ~%sound_reference%/%name%.txt~
				APPEND_OUTER ~%sound_reference%/%name%.txt~ 0
			END ELSE BEGIN
				COPY ~%sound_reference%/%name%.txt~ ~%sound_reference%/%name%.txt~
					READ_2DA_ENTRY 0 0 0 val
					SET $last(EVALUATE_BUFFER ~%name%~) = val
				BUT_ONLY
			END
		END
	END
BUT_ONLY

SILENT

ACTION_PHP_EACH early AS from => to BEGIN
	COPY ~%source%/%to%~ ~%source%/%to%~
		LPM ~read~
	BUT_ONLY
END

COPY ~%source%~ ~%source%~
	LPM ~read~
BUT_ONLY

VERBOSE

ACTION_PHP_EACH langs AS from => to BEGIN
	COPY ~%to%~ ~%to%~
		REPLACE_EVALUATE "\(@[0-9]* *\)= \(~[^~]*~\)%LNL%" BEGIN
			TEXT_SPRINT result ~%MATCH1%= %MATCH2%%LNL%~
			PATCH_IF VARIABLE_IS_SET EVALUATE_BUFFER ~%SOURCE_RES% %MATCH1%~ THEN BEGIN
				TEXT_SPRINT str ~%SOURCE_RES% %MATCH1%~
				TEXT_SPRINT str EVALUATE_BUFFER ~%%str%%~
				TEXT_SPRINT result ~%MATCH1%= %MATCH2% [%str%]%LNL%~
			END
		END "%result%"
	BUT_ONLY
END

COPY - ~names-to-soundsets~ ~names-to-soundsets~
	FOR (i = 1; i < names2da; ++i) BEGIN
		READ_2DA_ENTRY_FORMER names2da i 0 name
		
		INNER_ACTION BEGIN
			COPY ~%sound_reference%/%name%.txt~ ~%sound_reference%/%name%.txt~
				SET_2DA_ENTRY 0 0 0 $last(EVALUATE_BUFFER ~%name%~)
				APPEND_FILE ~%name%.tmp~
				REPLACE_TEXTUALLY ~%LNL%%LNL%*~ ~%LNL%~
		END
	END
BUT_ONLY

To work, this requires you to change some variables at the top of the tp2:
'source' defines the directory containing the TRA files (where markers will be translated, and from where sounds will be copied to other languages).
'early' is an array of files in ~%source%~ that will be processed as first (for example, to put sounds used in the character soundset in front of sounds used in banters).
'names-to-soundsets' defines the character-prefix mapping; that is to say, if you write [STIVAN] as the sound file for any tra reference, it will be translated to [TB#SS001] and [ROSALINA] to [TB#SRO01].
'sound_reference' is a directory where files containing the script for the various sound files to be recorded will be written to.
'langs' is an array of directories where new sound files will be copied to (without overwriting existing references).

Currently, there are a couple of restrictions (that prevent functionality on affected lines, while still working correctly on the rest of the tra file):
1) cannot handle .tra files with dialogF references
2) cannot handle .tra entries written on more than only one line
Both will be amended on request.

I hope this was confusing enough. As usual, I welcome all comments, suggestions and criticism, as long as it is done on a public channel.