this is a fairly simple script to download manga from mangadex and get new issues whenever rerun. the config should be put in config.json and looks like so:
{
"outputFolder": "./output",
"23424": {
"langCode": "gb",
"volumes": {
"1": 1,
"2": 5,
"3": 7,
"4": 9,
"5": 10,
"6": 12
},
},
"607": {
"langCode": "gb",
"ignoredGroups": [
4007
]
}
}outputFolder is where images and cached chapter information will be stored. every other key in the config should be a mangadex manga id. each accepts a langCode to determine which language to get (defaults to english). volumes is optional, used for specifying the first chapter of each volume and referred to when mangadex doesn't have a volume listed. if it's left out, volume-less chapters of the manga will be put in volume 0. ignoredGroups is an optional array of ids for translation groups to skip when downloading chapters (useful if there are two translations of the same thing).
also included is a bash script, cbzgen.sh, which runs manga downloads and then generates (or updates if already-existing) a cbz for each volume.
metadata is generated from mangadex and saved in the comicrack style of a ComicInfo.xml for each volume. currently what's probably the most annoying limitation is that covers aren't downloaded. mangadex only supplies a single cover of dubious authenticity for each manga (currently ignored), and though it also frequently stores per-volume covers, they aren't programattically accessible. you can download them yourself and put them into the chapter folder or volume cbz manually if you like; mangadextrous won't tamper with them.