path: root/git-fat

#!/usr/bin/env python

from __future__ import print_function, with_statement

import sys
import hashlib
import tempfile
import os
import subprocess
import shlex
import shutil
import itertools
import threading

BLOCK_SIZE = 4096

def verbose_stderr(*args, **kwargs):
    return print(*args, file=sys.stderr, **kwargs)
def verbose_ignore(*args, **kwargs):
    pass

def mkdir_p(path):
    import errno
    try:
        os.makedirs(path)
    except OSError as exc: # Python >2.5
        if exc.errno == errno.EEXIST and os.path.isdir(path):
            pass
        else: raise

def readblocks(stream):
    bytes = 0
    while True:
        data = stream.read(BLOCK_SIZE)
        bytes += len(data)
        if not data:
            break
        yield data
def cat_iter(initer, outstream):
    for block in initer:
        outstream.write(block)
def cat(instream, outstream):
    return cat_iter(readblocks(instream), outstream)
def gitconfig_get(name, file=None):
    args = ['git', 'config', '--get']
    if file is not None:
        args += ['--file', file]
    args.append(name)
    p = subprocess.Popen(args, stdout=subprocess.PIPE)
    output = p.communicate()[0].strip()
    if p.returncode != 0:
        return None
    else:
        return output
def gitconfig_set(name, value, file=None):
    args = ['git', 'config']
    if file is not None:
        args += ['--file', file]
    args += [name, value]
    p = subprocess.check_call(args)

class GitFat(object):
    DecodeError = RuntimeError
    def __init__(self):
        self.verbose = verbose_stderr if os.environ.get('GIT_FAT_VERBOSE') else verbose_ignore
        self.gitroot = subprocess.check_output('git rev-parse --show-toplevel'.split()).strip()
        self.objdir = os.path.join(self.gitroot, '.git', 'fat', 'objects')
        if os.environ.get('GIT_FAT_VERSION') == '1':
            self.encode = self.encode_v1
        else:
            self.encode = self.encode_v2
        def magiclen(enc):
            return len(enc(hashlib.sha1('dummy').hexdigest(), 5))
        self.magiclen = magiclen(self.encode) # Current version
        self.magiclens = [magiclen(enc) for enc in [self.encode_v1, self.encode_v2]] # All prior versions
    def setup(self):
        mkdir_p(self.objdir)
    def get_rsync(self):
        cfgpath = os.path.join(self.gitroot,'.gitfat')
        remote = gitconfig_get('rsync.remote', file=cfgpath)
        if remote is None:
            raise RuntimeError('No rsync.remote in %s' % cfgpath)
        return remote
    def revparse(self, revname):
        return subprocess.check_output(['git', 'rev-parse', revname]).strip()
    def encode_v1(self, digest, bytes):
        'Produce legacy representation of file to be stored in repository.'
        return '#$# git-fat %s\n' % (digest,)
    def encode_v2(self, digest, bytes):
        'Produce representation of file to be stored in repository. 20 characters can hold 64-bit integers.'
        return '#$# git-fat %s %20d\n' % (digest, bytes)
    def decode(self, string, noraise=False):
        cookie = '#$# git-fat '
        if string.startswith(cookie):
            parts = string[len(cookie):].split()
            digest = parts[0]
            bytes = int(parts[1]) if len(parts) > 1 else None
            return digest, bytes
        elif noraise:
            return None, None
        else:
            raise GitFat.DecodeError('Could not decode %s' % (string))
    def decode_stream(self, stream):
        'Return digest if git-fat cache, otherwise return iterator over entire file contents'
        preamble = stream.read(self.magiclen)
        try:
            return self.decode(preamble)
        except GitFat.DecodeError:
            # Not sure if this is the right behavior
            return itertools.chain([preamble], readblocks(stream)), None
    def decode_file(self, fname):
        # Fast check
        stat = os.stat(fname)
        if stat.st_size != self.magiclen:
            return False, None
        # read file
        digest, bytes = self.decode_stream(open(fname))
        if isinstance(digest, str):
            return digest, bytes
        else:
            return None, bytes
    def decode_clean(self, body):
        '''
        Attempt to decode version in working tree. The tree version could be changed to have a more
        useful message than the machine-readable copy that goes into the repository. If the tree
        version decodes successfully, it indicates that the fat data is not currently available in
        this repository.
        '''
        digest, bytes = self.decode(body, noraise=True)
        return digest
    def cmd_clean(self):
        '''
        The clean filter runs when a file is added to the index. It gets the "smudged" (tree)
        version of the file on stdin and produces the "clean" (repository) version on stdout.
        '''
        self.setup()
        h = hashlib.new('sha1')
        bytes = 0
        fd, tmpname = tempfile.mkstemp(dir=self.objdir)
        try:
            ishanging = False
            cached = False                # changes to True when file is cached
            with os.fdopen(fd, 'w') as cache:
                outstream = cache
                blockiter = readblocks(sys.stdin)
                # Check whether this file is hanging
                block = next(blockiter)
                if self.decode_clean(block[0:self.magiclen]):
                    ishanging = True
                    outstream = sys.stdout
                h.update(block)
                bytes += len(block)
                outstream.write(block)
                for block in blockiter:
                    h.update(block)
                    bytes += len(block)
                    outstream.write(block)
                outstream.flush()
            digest = h.hexdigest()
            objfile = os.path.join(self.objdir, digest)
            if not ishanging:
                if os.path.exists(objfile):
                    self.verbose('git-fat filter-clean: cache already exists %s' % objfile)
                    os.remove(tmpname)
                else:
                    os.rename(tmpname, objfile)
                    self.verbose('git-fat filter-clean: caching to %s' % objfile)
                cached = True
                sys.stdout.write(self.encode(digest, bytes))
        finally:
            if not cached:
                os.remove(tmpname)

    def cmd_smudge(self):
        self.setup()
        result, bytes = self.decode_stream(sys.stdin)
        if isinstance(result, str):       # We got a digest
            objfile = os.path.join(self.objdir, result)
            try:
                cat(open(objfile), sys.stdout)
                self.verbose('git-fat filter-smudge: restoring from %s' % objfile)
            except IOError:                 # file not found
                self.verbose('git-fat filter-smudge: fat object missing %s' % objfile)
                sys.stdout.write(self.encode(result, bytes))   # could leave a better notice about how to recover this file
        else:                             # We have an iterable over the original input.
            self.verbose('git-fat filter-smudge: not a managed file')
            cat_iter(result, sys.stdout)
    def catalog_objects(self):
        return set(os.listdir(self.objdir))
    def referenced_objects(self, rev=None, all=False):
        referenced = set()
        if all:
            rev = '--all'
        elif rev is None:
            rev = self.revparse('HEAD')
        p1 = subprocess.Popen(['git','rev-list','--objects',rev], stdout=subprocess.PIPE)
        p2 = subprocess.Popen(['git','cat-file','--batch-check'], stdin=subprocess.PIPE, stdout=subprocess.PIPE)
        def cut_sha1hash(input, output):
            for line in input:
                output.write(line.split()[0] + '\n')
            output.close()
        cut_thread = threading.Thread(target=cut_sha1hash, args=(p1.stdout, p2.stdin))
        cut_thread.start()
        for line in p2.stdout:
            objhash, objtype, size = line.split()
            if objtype == 'blob' and int(size) in self.magiclens:
                fathash = self.decode(subprocess.check_output(['git', 'cat-file', '-p', objhash]))[0]
                referenced.add(fathash)
        cut_thread.join()
        p1.wait()
        p2.wait()
        return referenced
    def orphan_files(self):
        'generator for all orphan placeholders in the working tree'
        for fname in subprocess.check_output(['git', 'ls-files']).splitlines():
            digest = self.decode_file(fname)[0]
            if digest:
                yield (digest, fname)
    def cmd_status(self, args):
        self.setup()
        catalog = self.catalog_objects()
        refargs = dict()
        if '--all' in args:
            refargs['all'] = True
        referenced = self.referenced_objects(**refargs)
        garbage = catalog - referenced
        orphans = referenced - catalog
        if '--all' in args:
            for obj in referenced:
                print(obj)
        if orphans:
            print('Orphan objects:')
            for orph in orphans:
                print('    ' + orph)
        if garbage:
            print('Garbage objects:')
            for g in garbage:
                print('    ' + g)
    def is_dirty(self):
        return subprocess.call(['git', 'diff-index', '--quiet', 'HEAD']) == 0
    def cmd_push(self, args):
        'Push anything that I have stored and referenced'
        self.setup()
        # Default to push only those objects referenced by current HEAD
        # (includes history). Finer-grained pushing would be useful.
        pushall = '--all' in args
        files = self.referenced_objects(all=pushall) & self.catalog_objects()
        remote = self.get_rsync()
        self.verbose('Pushing to %s' % (remote))
        cmd = ['rsync', '--progress', '--ignore-existing', '--from0', '--files-from=-', self.objdir + '/', remote + '/']
        p = subprocess.Popen(cmd, stdin=subprocess.PIPE)
        p.communicate(input='\x00'.join(files))
    def checkout(self, show_orphans=False):
        'Update any stale files in the present working tree'
        for digest, fname in self.orphan_files():
            objpath = os.path.join(self.objdir, digest)
            if os.access(objpath, os.R_OK):
                print('Restoring %s -> %s' % (digest, fname))
                # The output of our smudge filter depends on the existence of
                # the file in .git/fat/objects, but git caches the file stat
                # from the previous time the file was smudged, therefore it
                # won't try to re-smudge. I don't know a git command that
                # specifically invalidates that cache, but touching the file
                # also does the trick.
                os.utime(fname, None)
                # This re-smudge is essentially a copy that restores permissions.
                subprocess.check_call(['git', 'checkout-index', '--index', '--force', fname])
            elif show_orphans:
                print('Data unavailable: %s %s' % (digest,fname))
    def cmd_pull(self, args):
        'Pull anything that I have referenced, but not stored'
        self.setup()
        refargs = dict()
        if '--all' in args:
            refargs['all'] = True
        for arg in args:
            if arg.startswith('-') or len(arg) != 40:
                continue
            rev = self.revparse(arg)
            if rev:
                refargs['rev'] = rev
        files = self.referenced_objects(**refargs) - self.catalog_objects()
        remote = self.get_rsync()
        cmd = ['rsync', '--progress', '--ignore-existing', '--from0', '--files-from=-', remote + '/', self.objdir + '/']
        p = subprocess.Popen(cmd, stdin=subprocess.PIPE)
        p.communicate(input='\x00'.join(files))
        self.checkout()
    def cmd_checkout(self, args):
        self.checkout(show_orphans=True)
    def cmd_gc(self):
        garbage = self.catalog_objects() - self.referenced_objects()
        print('Unreferenced objects to remove: %d' % len(garbage))
        for obj in garbage:
            fname = os.path.join(self.objdir, obj)
            print('%10d %s' % (os.stat(fname).st_size, obj))
            os.remove(fname)
    def cmd_init(self):
        self.setup()
        if gitconfig_get('filter.fat.clean') or gitconfig_get('filter.fat.smudge'):
            print('Git fat already configured, check configuration in .git/config')
        else:
            gitconfig_set('filter.fat.clean', 'git-fat filter-clean')
            gitconfig_set('filter.fat.smudge', 'git-fat filter-smudge')
            print('Initialized git fat')

if __name__ == '__main__':
    fat = GitFat()
    cmd = sys.argv[1] if len(sys.argv) > 1 else ''
    if cmd == 'filter-clean':
        fat.cmd_clean()
    elif cmd == 'filter-smudge':
        fat.cmd_smudge()
    elif cmd == 'init':
        fat.cmd_init()
    elif cmd == 'status':
        fat.cmd_status(sys.argv[2:])
    elif cmd == 'push':
        fat.cmd_push(sys.argv[2:])
    elif cmd == 'pull':
        fat.cmd_pull(sys.argv[2:])
    elif cmd == 'gc':
        fat.cmd_gc()
    elif cmd == 'checkout':
        fat.cmd_checkout(sys.argv[2:])
    else:
        print('Usage: git fat [init|status|push|pull|gc|checkout]', file=sys.stderr)