#!/usr/bin/env python3
"""Baixa os objetos Git LFS de uma arvore (pelo endpoint batch do LFS, sem o git-lfs instalado).

  python3 -I lfs_fetch.py <arvore> <url do repositorio .git> <cache>

Cada ponteiro (version/oid/size) vira o arquivo real; o sha256 do objeto e conferido contra o oid do ponteiro.
Os objetos ficam no cache, com o oid no nome.
"""
import hashlib
import json
import os
import sys
import urllib.request

tree, base, cache = sys.argv[1:4]
os.makedirs(cache, exist_ok=True)
MAGIC = b'version https://git-lfs.github.com/spec/v1'

ptrs = {}
for dp, dn, fn in os.walk(tree):
    dn[:] = [d for d in dn if d != '.git']
    for f in sorted(fn):
        p = os.path.join(dp, f)
        if os.path.islink(p) or os.path.getsize(p) > 200:
            continue
        with open(p, 'rb') as fh:
            t = fh.read()
        if t.startswith(MAGIC):
            d = dict(l.split(' ', 1) for l in t.decode().strip().split('\n')[1:])
            ptrs.setdefault(d['oid'].split(':')[1], (int(d['size']), []))[1].append(p)

need = [(o, v) for o, v in sorted(ptrs.items()) if not os.path.exists(os.path.join(cache, 'lfs-' + o))]
print(len(ptrs), 'objetos LFS,', len(need), 'a baixar')
for i in range(0, len(need), 50):
    chunk = need[i:i + 50]
    body = json.dumps({'operation': 'download', 'transfers': ['basic'],
                       'objects': [{'oid': o, 'size': v[0]} for o, v in chunk]}).encode()
    req = urllib.request.Request(base + '/info/lfs/objects/batch', data=body, headers={
        'Accept': 'application/vnd.git-lfs+json', 'Content-Type': 'application/vnd.git-lfs+json',
        'User-Agent': 'git-lfs/3.4'})
    r = json.load(urllib.request.urlopen(req))
    for ob in r['objects']:
        a = ob.get('actions', {}).get('download')
        if not a:
            sys.exit('sem download: %s %s' % (ob['oid'], ob.get('error')))
        rq = urllib.request.Request(a['href'], headers=a.get('header', {}))
        data = urllib.request.urlopen(rq).read()
        if hashlib.sha256(data).hexdigest() != ob['oid']:
            sys.exit('sha256 diferente: ' + ob['oid'])
        with open(os.path.join(cache, 'lfs-' + ob['oid'] + '.partial'), 'wb') as fh:
            fh.write(data)
        os.rename(os.path.join(cache, 'lfs-' + ob['oid'] + '.partial'), os.path.join(cache, 'lfs-' + ob['oid']))

for o, (size, paths) in ptrs.items():
    src = os.path.join(cache, 'lfs-' + o)
    data = open(src, 'rb').read()
    if len(data) != size or hashlib.sha256(data).hexdigest() != o:
        sys.exit('cache corrompido: ' + o)
    for p in paths:
        with open(p, 'wb') as fh:
            fh.write(data)
print('ok')