cobblemon-survival-pack/tools/crosslink.py

#!/usr/bin/env python3
"""Give every packwiz mod both a Modrinth and a CurseForge source.

packwiz only records the platform a mod was added from. Exporting to the other
platform then copies the jar into overrides/, which would bundle a third-party
jar. This script finds the same file on the other platform and verifies it by
SHA-1 before linking it:

  * Modrinth -> CurseForge: CurseForge project found via cfwidget (slug from
    crosslink.json or the Modrinth slug), file matched by filename, then both
    files are downloaded and compared: same SHA-1, or (for jars re-built with
    other timestamps) the same entry names and CRC32s.
  * CurseForge -> Modrinth: exact lookup of the SHA-1 on Modrinth.

Mods with no counterpart are reported; build.py keeps them out of the export
for the platform that lacks them instead of bundling them.

Usage: tools/crosslink.py [pack_dir]      (then run `packwiz refresh`)
"""
import io
import json
import sys
import tomllib
import urllib.error
import urllib.parse
import urllib.request
import zipfile
from pathlib import Path

ROOT = Path(__file__).resolve().parent.parent
OVERRIDES = json.loads((ROOT / "tools" / "crosslink.json").read_text())
UA = {"User-Agent": "cobblemon-survival-pack-tools/0.1 (packwiz crosslink)"}


def get_json(url):
    req = urllib.request.Request(url, headers=UA)
    try:
        with urllib.request.urlopen(req, timeout=60) as r:
            return json.load(r)
    except urllib.error.HTTPError as e:
        if e.code == 404:
            return None
        raise


def download(url):
    req = urllib.request.Request(url, headers=UA)
    with urllib.request.urlopen(req, timeout=300) as r:
        return r.read()


def same_file(a, b):
    """'sha1' if byte-identical, 'content' if the jars hold identical entries.

    The manifest is ignored: authors who upload to each platform from a
    separate build get a different Implementation-Timestamp and nothing else.
    """
    if a == b:
        return "sha1"
    try:
        entries = [sorted((i.filename, i.CRC) for i in zipfile.ZipFile(io.BytesIO(x)).infolist()
                          if i.filename != "META-INF/MANIFEST.MF")
                   for x in (a, b)]
    except zipfile.BadZipFile:
        return None
    return "content" if entries[0] == entries[1] else None


def forgecdn_url(file_id, filename):
    return "https://edge.forgecdn.net/files/{}/{}/{}".format(
        file_id // 1000, file_id % 1000, urllib.parse.quote(filename))


def link_curseforge(path, meta, override):
    mr = meta["update"]["modrinth"]
    version = get_json(f"https://api.modrinth.com/v2/version/{mr['version']}")
    project = get_json(f"https://api.modrinth.com/v2/project/{mr['mod-id']}")
    file = next(f for f in version["files"] if f["filename"] == meta["filename"])
    slug = override.get("curseforge_slug", project["slug"])
    cf = get_json(f"https://api.cfwidget.com/minecraft/mc-mods/{slug}")
    if not cf or "files" not in cf:
        return f"no CurseForge project for slug '{slug}'"
    candidates = [f for f in cf["files"] if f["name"] == meta["filename"]]
    if not candidates:
        return f"CurseForge project {cf['id']} has no file named {meta['filename']}"
    mr_bytes = download(file["url"])
    for cand in candidates:
        match = same_file(mr_bytes, download(forgecdn_url(cand["id"], cand["name"])))
        if not match and override.get("trust_same_filename") and len(candidates) == 1:
            match = f"filename, trusted: {override['trust_same_filename']}"
        if match:
            text = path.read_text().rstrip() + (
                f"\n\n[update.curseforge]\nfile-id = {cand['id']}\nproject-id = {cf['id']}\n")
            path.write_text(text)
            return f"linked CurseForge {cf['id']}/{cand['id']} (same {match})"
    return "CurseForge file found by name but its contents differ"


def link_modrinth(path, meta):
    sha1 = meta["download"]["hash"] if meta["download"]["hash-format"] == "sha1" else None
    if not sha1:
        return "no SHA-1 in metadata"
    version = get_json(f"https://api.modrinth.com/v2/version_file/{sha1}?algorithm=sha1")
    if not version:
        return "not on Modrinth"
    file = next(f for f in version["files"] if f["hashes"]["sha1"] == sha1)
    cf = meta["update"]["curseforge"]
    path.write_text(
        f'name = "{meta["name"]}"\n'
        f'filename = "{meta["filename"]}"\n'
        f'side = "{meta.get("side", "both")}"\n\n'
        f'[download]\nurl = "{file["url"]}"\nhash-format = "sha1"\nhash = "{sha1}"\n\n'
        f'[update]\n[update.curseforge]\nfile-id = {cf["file-id"]}\nproject-id = {cf["project-id"]}\n\n'
        f'[update.modrinth]\nmod-id = "{version["project_id"]}"\nversion = "{version["id"]}"\n')
    return f"linked Modrinth {version['project_id']}/{version['id']}"


def main():
    pack = Path(sys.argv[1]) if len(sys.argv) > 1 else ROOT / "pack"
    missing = []
    for path in sorted((pack / "mods").glob("*.pw.toml")):
        meta = tomllib.loads(path.read_text())
        update = meta.get("update", {})
        override = OVERRIDES.get(path.name.removesuffix(".pw.toml"), {})
        if "modrinth" in update and "curseforge" in update:
            status = "already linked"
        elif override.get("skip"):
            status = f"skipped: {override['skip']}"
        elif "modrinth" in update:
            status = link_curseforge(path, meta, override)
        else:
            status = link_modrinth(path, meta)
        if not status.startswith(("linked", "already")):
            missing.append(path.name)
        print(f"{path.name:45} {status}")
    if missing:
        print("\nSingle-platform mods (kept out of the other platform's export):")
        print("\n".join(f"  {m}" for m in missing))


if __name__ == "__main__":
    main()