-
-
Notifications
You must be signed in to change notification settings - Fork 1
Command Line Usage
This article describes how to use the various command line tools which Luna Purpura provides.
cluprxmmfwtools/clu2pngxpktools/fakesqs2
The clu program reads CLU files, which are color lookup tables. The tool
doesn't do anything except list each of the colors in the table. The program
really ought to let you convert a CLU to a PNG image, but as of this
writing, that's taken care of by the tools/clu2png Ruby script for now.
$ ./build/src/clu/clu ./zz-tricky/Scene2/\!Scene2.CLU 2>/dev/null | head
========================================
<255,255,255>
<128,0,0>
<0,128,0>
<128,128,0>
<0,0,128>
<128,0,128>
<0,128,128>
<128,128,128>
<253,253,253>
Every CLU file specifies exactly 256 colors.
The prx tool reads PRX archives. It currently does NOT understand PRD/PRS
archives. You can use it to list the individual members of an archive or to
extract all members, or a single member.
prx: usage: prx [options ...] [file]
prx -t <file> List members
prx -x [-v] [-o <path>] -n <ID> <file> Extract a single member
prx -x [-v] [-o <path>] -f <TYPE> -r <RID> <file> Extract a single member by RID
prx -x [-v] -a <file> Extract all members
prx -h Show this help message
The -t option indicates you want to list all the members of the PRX
archive.
$ ./build/src/prx/prx -t ./data/tricky/Credits.PRX
>> 3 entries:
1 !Credits CLU 64000 2056
2 !Credits DIB 64000 1082
3 credit SMF 64000 3845087
Each row represents a single member in the archive. The columns are, from left to right: "internal ID," filename, file type, "resource ID," and finally the file size in bytes.
If you want to extract a single member from an archive, you have two options. You may either refer to the desired member by its "internal" ID, or you may use its "resource ID" (or RID).
Extract via internal ID. prx -x -n <ID> <file>. -x indicates you want
to extract, and -n <ID> signifies your intention to extract only a
specific member. The number should be the "internal ID" of the asset, that
is, the number in the first column.
Extract via resource ID (RID). prx -x -f <TYPE> -r <RID> <file>.
Resource IDs are not globally unique across all resources, instead they are
"namespaced" by filetype. Therefore, you must provide a 3-letter TYPE as
well as a numerical resource ID (the value in the third column of the
listing).
By default, the name of the extracted file is based on what is stored inside
the archive. (To avoid writing duplicate files, the default filename also
embeds member's RID.) However, you may extract the member to a location of
your choosing with the -o <path> option. If the specified path is - (a
single dash), then the member will be written to the standard output.
If you want to extract all members from the given archive, pass the -a
option alongside -x.
If you pass the -v ("verbose") option to an extract command, then a
message is printed for each member as they are extracted.
The mmfw reads the various kinds of MMFW archives. MMFW archives usually
have the following file extensions: MMA ("MMFW Sounds"), MMF ("MMFW
Films"), and MMP ("MMFW Pictures"). These are in fact different file
formats (because each one stores different kinds of metadata) but they're
mostly the same and therefore only 1 tool is needed to extract files from
any of them.
mmfw: usage: mmfw [options ...] [file]
mmfw -t <file> List archive members
mmfw -x [-v] -a <file> Extract all members
mmfw -x [-v] -n <id> <file> Extract a single member
The -t option indicates you want to list all members of the archive.
$ ./build/src/mmfw/mmfw -t ./MAKER/Comics.MMF
mmfw: "Films" MMFW archive with 7 members
0 presents.flm 3552
1 ramCredits.flm 646588
2 ramFlash.flm 7238060
3 ramIntro.flm 7793948
4 ramTitle.flm 21813148
5 IbsLogo.flm 22949984
6 PMsplash.flm 24137272
Each row represents a single member in the archive. The columns are, from left to right: the ID, filename, and its internal offset.
If you want to extract a single member from the archive, you will need to
consult its ID (above) and use this number for the -n <ID> option in mmfw -x -n <ID>. Otherwise, use -a in place of -n <ID> to signify your
intention to extract all members into the current directory.
While extracting, you may enable verbose mode with -v, which will write a
status message to stderr as each member is extracted.
The xpk tool tries to decode XPK graphics. Note that a single XPK file can
define multiple images (e.g. individual frames in an animation) but as of
this writing, xpk will only try to decode the first image.
xpk: usage: xpk [-h] [-p] -c clu -o output file
XPK graphics do not store color information directly, they only store
indices into a color lookup table, or CLU. Therefore, you MUST provide the
path to a CLU file in order to decode an XPK. You'll know which CLU to use
because (typically) a single one can be found in each PRX (or PRS) archive.
IDPalette.CLU or _IDGlobal.CLU will be used for anything that's
contained in a student's locker.
By default, xpk will write raw RGBA samples (24 bits per pixel).
GraphicsMagick (and/or
ImageMagick) will allow you to do
interesting things with them, but you'll need to pay attention to the width
and height of the image. You'll need GraphicsMagick's -depth and -size
options like so:
$ ./build/src/xpk/xpk -c tricky/IDGlobal/IDPalette.CLU -o /tmp/MAVglasses_01_H.rgba tricky/IDL/Mavi/MAVglasses_01_H.XPK
1 entry(ies):
XPKEntry offset=36 dimensions=106x73 pos=188x362
$ gm convert -depth 8 -size 106x73 rgba:/tmp/MAVglasses_01_H.rgba [output_options ...] <outfile>
If PNG support is present (CMake option LUNAPURPURA_PNG_SUPPORT, enabled
by default) then xpk gains support for the -p option. When enabled, -p
will generate a PNG image instead of raw RGBA samples.
$ ./build/src/xpk/xpk -c tricky/IDGlobal/IDPalette.CLU -p -o /tmp/MAVglasses_01_H.png tricky/IDL/Mavi/MAVglasses_01_H.XPK
SQS2 compression is unknown to FFmpeg, VLC, vgmstream, and, well, virtually
all audio software ever. We'll need to reverse engineer it. However, it
turns out that the far more popular, well-understood compression scheme
"SDX2" is a fairly close approximation. fakesqs2 is a Ruby script which
"patches" a AIFF-C file with SQS2 compression and writes a new file which
pretends to be encoded with SDX2 compression. This allows the most of the
audio for most of the Rockett titles to be audible, but note that it's
heavily distorted.
usage: fakesqs2 [options ...] file
-i, --in-place overwrite input file
-o, --outfile PATH path to write new file
You can safely use fakesqs2 on AIFF files, and AIFF-C files which do not
use SQS2 compression. A copy of the original file will be generated instead.
By the way, you can install the sox software
and use the soxi tool to quickly determine whether file is SQS2-ified or
not:
$ soxi D1TrailerBox.Aif
Input File : 'D1TrailerBox.Aif'
Channels : 1
Sample Rate : 22050
Precision : 16-bit
Duration : 00:00:00.28 = 6208 samples ~ 21.1156 CDDA sectors
File Size : 12.5k
Bit Rate : 355k
Sample Encoding: 16-bit Signed Integer PCM
$ soxi marquee.Aif
soxi FAIL formats: can't open input file `marquee.Aif': Unsupported AIFC compression type `SQS2'