#!/usr/bin/env python3

## Copyright 2005-2015 Linus Torvalds and others
## Copyright 2015, 2019, 2020 Codethink Limited
##
## Based on git's contrib/fast-import/import-zips.py; modified for Lorry.

## zip archive frontend for git-fast-import
##
## For example:
##
##  mkdir project; cd project; git init
##  python import-zips.py *.zip
##  git log --stat import-zips

import calendar
import os.path
import struct
import subprocess
import sys
from zipfile import ZipFile


branch_name = 'master'
branch_ref = 'refs/heads/%s' % branch_name
committer_name = 'Lorry Zip Importer'
committer_email = 'lorry-zip-importer@lorry'


# File header 'extra' field tags
EXT_TAG_UNIX0 = 0x000d   # PKWARE Unix, aka Unix type 0
EXT_TAG_TIME = 0x5455    # Extended Timestamp
EXT_TIME_FLAG_MTIME = 1  # mtime present (and first)
EXT_TAG_UNIX1 = 0x5855   # Info-ZIP Unix type 1


# Iterate over fields within a file header 'extra' block
def zip_extra_fields(extra):
    pos = 0
    while len(extra) >= pos + 4:
        tag, size = struct.unpack('<HH', extra[pos : pos + 4])
        pos += 4
        if len(extra) < pos + size:
            return
        yield tag, extra[pos : pos + size]
        pos += size


# Make our best guess at the mtime of a zip file entry
def zip_info_mtime(info):
    # Look for Unix-format mtime in the 'extra' block
    for tag, data in zip_extra_fields(info.extra):
        format = None
        if tag in [EXT_TAG_UNIX0, EXT_TAG_UNIX1]:
            format = '<4xL'  # AcTime, ModTime
        elif tag == EXT_TAG_TIME:
            # First byte indicates which timestamps follow
            if len(data) >= 1 and data[0] & EXT_TIME_FLAG_MTIME:
                format = '<xL'  # Flags, ModTime
        if format:
            min_len = struct.calcsize(format)
            if len(data) >= min_len:
                return struct.unpack(format, data[:min_len])[0]

    # Timestamps in the main header are in local time, but the time
    # zone offset is unspecified.  We choose to interpret them as UTC.
    return calendar.timegm(info.date_time + (0, 0, 0))


def export(zipfile, fast_import):
    def printlines(list):
        for str in list:
            fast_import.write(str.encode('utf-8') + b"\n")

    commit_time = 0
    next_mark = 1
    common_prefix = None
    mark = dict()

    zip = ZipFile(zipfile, 'r')
    for name in zip.namelist():
        if name.endswith('/'):
            continue
        info = zip.getinfo(name)

        commit_time = max(commit_time, zip_info_mtime(info))
        if common_prefix == None:
            common_prefix = name[:name.rfind('/') + 1]
        else:
            while not name.startswith(common_prefix):
                last_slash = common_prefix[:-1].rfind('/') + 1
                common_prefix = common_prefix[:last_slash]

        mark[name] = ':' + str(next_mark)
        next_mark += 1

        printlines(('blob', 'mark ' + mark[name], \
                    'data ' + str(info.file_size)))
        fast_import.write(zip.read(name) + b"\n")

    committer = committer_name + ' <' + committer_email + '> %d +0000' % \
        commit_time

    zipfile_basename = os.path.basename(zipfile)
    printlines(('commit ' + branch_ref, 'committer ' + committer, \
        'data <<EOM', 'Imported from ' + zipfile_basename + '.', 'EOM', \
        '', 'deleteall'))

    for name in mark.keys():
        printlines(('M 100644 ' + mark[name] + ' ' +
                    name[len(common_prefix):],))

    zipname, _ = os.path.splitext(zipfile_basename)

    printlines(('',  'tag ' + zipname, \
        'from ' + branch_ref, 'tagger ' + committer, \
        'data <<EOM', 'Package ' + zipfile, 'EOM', ''))


def main():
    if len(sys.argv) < 2:
        print('usage:', sys.argv[0], '<zipfile>...')
        sys.exit(1)

    with subprocess.Popen('git fast-import --quiet', shell=True,
                          stdin=subprocess.PIPE) as import_proc:
        for zipfile in sys.argv[1:]:
            export(zipfile, import_proc.stdin)
        import_proc.stdin.close()
        if import_proc.wait() != 0:
            sys.exit(1)


main()