This script converts cosmos memo files to a more usable JSON format with sorted timestamps.

For documentation on the file format see src/memson.ts.

The ready to use index.js can be found in the dist directory.

Make sure you have the node runtime >=10 installed. No other dependecies are needed.

To parse a memo file to json simply navigate to the projects directory and execute following command:

node dist/index.js [path-to-memo-or-dir] <pretty> <folderize>

The path can either be a single file or a directory containing memo files. File names should be in the following format.

some_name_(bsc|adv|ext).(memo|txt)

Example file names:

awesome_song_adv.memo
another_cool_song.txt

Output file name is the input file name with a .json ending and the difficulty reference removed. So for an input of awesome_song_adv.memo and output file with the name awesome_song.json will be created.

If a json file with the correct name for the same song already exists, then the chart information for another difficulty will be appended.

E.g. awesome_song.json already exists and contains the adv chart. If awesome_song_ext.memo is parsed after that, then the ext chart data will be appended to awesome_song.json.

The option pretty is optional and takes either true or false as input. If set to true it will stringify the JSON in a more readable way but this takes up more file space.

With folderize set to true, the parser will output the file into a subfolder with the same name as the parsed memo.

This script is a bit hacky, so please make sure that you copy the memo like it is displayed at cosmos. Make sure that your file does not contain any leading or trailing whitespaces or newlines.

Example file:

Another Phase
TAG

EXTREME

Level: 9
BPM: 160
Notes: 693

1
口口口口 |－－－－|
口口口口 |－－－－|
口口口口 |－－－－|
口口口口 |－－－－|
...

There should be a blank line between the artist and the difficulty category. And also one between the category and the other metadata. The last blank line separates the metadata from the note data. There shouldn't be any blank lines after that.

In case anyone would ever need a more stable parser then let me know.