A Python module for parsing VGM (Video Game Music)
files. .vgm
and .vgz
(Gzip compressed) files are supported.
Currently, only version 1.50 of the VGM specification is supported.
The vgmparse
module can be installed directly from GitHub using pip
:
sudo pip install -e git+https://github.com/cdodd/vgmparse.git#egg=vgmparse
First import the vgmparse
module:
@>>> import vgmparse
Open a file in binary mode and pass the data to the Parser
class:
@>>> file_data = open('go_straight.vgm', 'rb').read()
@>>> vgm_data = vgmparse.Parser(file_data)
The VGM metadata is available as a dict
in the metadata
variable:
@>>> vgm_data.metadata
{'total_samples': 4609970, 'vgm_data_offset': 12, 'vgm_ident': b'Vgm ', 'gd3_offset': 1110192, 'version': 336, 'ym2151_clock': 0, 'eof_offset': 1110478, 'sn76489_clock': 3579545, 'ym2413_clock': 0, 'loop_samples': 3951361, 'sn76489_feedback': 9, 'ym2612_clock': 7670453, 'loop_offset': 50883, 'rate': 60, 'sn76489_shift_register_width': 16}
GD3 data (track title, artist, etc) is available as a dict
in the gd3_data
variable:
@>>> vgm_data.gd3_data.keys()
dict_keys(['title_eng', 'console_jap', 'notes', 'game_eng', 'game_jap', 'artist_eng', 'vgm_creator', 'console_eng', 'artist_jap', 'title_jap', 'date'])
@>>> print(vgm_data.gd3_data['title_eng'].decode())
Go Straight
@>>> print(vgm_data.gd3_data['artist_eng'].decode())
Yuzo Koshiro
As per the GD3 specification the GD3 data uses 2-byte character encoding, however the exact encoding is not specified, leaving it up to the creator of the VGM file. This modules does not attempt to interpret the encoding, leaving the raw 2-byte encoding intact from the VGM file.
The VGM commands are available as a list
of dict
s in the variable
command_list
:
@>>> vgm_data.command_list[:3]
[{'command': b'R', 'data': b'(\x00'}, {'command': b'R', 'data': b'(\x04'}, {'command': b'R', 'data': b'(\x01'}]
If a command does not have have any data then data
will be None
.
Note: command_list
will not include any data block commands as the data
block is parsed out for you.
If a data block is present it is read into the variable data_block
. You can
then seek and read like a file:
@>>> vgm_data.data_block.seek(0)
0
@>>> vgm_data.data_block.read(5)
b'\x82\x88\x8a\x8a\x88'