Mercurial > hg
view mercurial/archival.py @ 29550:1c22400db72d
mercurial: implement a source transforming module loader on Python 3
The most painful part of ensuring Python code runs on both Python 2
and 3 is string encoding. Making this difficult is that string
literals in Python 2 are bytes and string literals in Python 3 are
unicode. So, to ensure consistent types are used, you have to
use "from __future__ import unicode_literals" and/or prefix literals
with their type (e.g. b'foo' or u'foo').
Nearly every string in Mercurial is bytes. So, to use the same source
code on both Python 2 and 3 would require prefixing nearly every
string literal with "b" to make it a byte literal. This is ugly and
not something mpm is willing to do at this point in time.
This patch implements a custom module loader on Python 3 that performs
source transformation to convert string literals (unicode in Python 3)
to byte literals. In effect, it changes Python 3's string literals to
behave like Python 2's.
In addition, the module loader recognizes well-known built-in
functions (getattr, setattr, hasattr) and methods (encode and decode)
that barf when bytes are used and prevents these from being rewritten.
This prevents excessive source changes to accommodate this change
(we would have to rewrite every occurrence of these functions passing
string literals otherwise).
The module loader is only used on Python packages belonging to
Mercurial.
The loader works by tokenizing the loaded source and replacing
"string" tokens if necessary. The modified token stream is
untokenized back to source and loaded like normal. This does add some
overhead. However, this all occurs before caching: .pyc files will
cache the transformed version. This means the transformation penalty
is only paid on first load.
As the extensive inline comments explain, the presence of a custom
source transformer invalidates assumptions made by Python's built-in
bytecode caching mechanism. So, we have to wrap bytecode loading and
writing and add an additional header to bytecode files to facilitate
additional cache validation when the source transformations
change in the future.
There are still a few things this code doesn't handle well, namely
support for zip files as module sources and for extensions. Since
Mercurial doesn't officially support Python 3 yet, I'm inclined to
leave these as to-do items: getting a basic module loading mechanism
in place to unblock further Python 3 porting effort is more important
than comprehensive module importing support.
check-py3-compat.py has been updated to ignore frames. This is
necessary because CPython has built-in code to strip frames from the
built-in importer. When our custom code is present, this doesn't work
and the frames get all messed up. The new code is not perfect. It
works for now. But once you start chasing import failures you find
some edge cases where the files aren't being printed properly. This
only burdens people doing future Python 3 porting work so I'm inclined
to punt on the issue: the most important thing is for the source
transforming module loader to land.
There was a bit of churn in test-check-py3-compat.t because we now
trip up on str/unicode/bytes failures as a result of source
transformation. This is unfortunate but what are you going to do.
It's worth noting that other approaches were investigated.
We considered using a custom file encoding whose decode() would
apply source transformations. This was rejected because it would
require each source file to declare its custom Mercurial encoding.
Furthermore, when changing the source transformation we'd need to
version bump the encoding name otherwise the module caching layer
wouldn't know the .pyc file was invalidated. This would mean mass
updating every file when the source transformation changes. Yuck.
We also considered transforming at the AST layer. However, Python's
ast module is quite gnarly and doing AST transforms is quite
complicated, even for trivial rewrites. There are whole Python packages
that exist to make AST transformations usable. AST transforms would
still require import machinery, so the choice was basically to
perform source-level, token-level, or ast-level transforms.
Token-level rewriting delivers the metadata we need to rewrite
intelligently while being relatively easy to understand. So it won.
General consensus seems to be that this approach is the best available
to avoid bulk rewriting of '' to b''. However, we aren't confident
that this approach will never be a future maintenance burden. This
approach does unblock serious Python 3 porting efforts. So we can
re-evaulate once more work is done to support Python 3.
author | Gregory Szorc <gregory.szorc@gmail.com> |
---|---|
date | Mon, 04 Jul 2016 11:18:03 -0700 |
parents | 86db5cb55d46 |
children | 31a6d5e14508 |
line wrap: on
line source
# archival.py - revision archival for mercurial # # Copyright 2006 Vadim Gelfer <vadim.gelfer@gmail.com> # # This software may be used and distributed according to the terms of the # GNU General Public License version 2 or any later version. from __future__ import absolute_import import gzip import os import struct import tarfile import time import zipfile import zlib from .i18n import _ from . import ( cmdutil, encoding, error, match as matchmod, scmutil, util, ) stringio = util.stringio # from unzip source code: _UNX_IFREG = 0x8000 _UNX_IFLNK = 0xa000 def tidyprefix(dest, kind, prefix): '''choose prefix to use for names in archive. make sure prefix is safe for consumers.''' if prefix: prefix = util.normpath(prefix) else: if not isinstance(dest, str): raise ValueError('dest must be string if no prefix') prefix = os.path.basename(dest) lower = prefix.lower() for sfx in exts.get(kind, []): if lower.endswith(sfx): prefix = prefix[:-len(sfx)] break lpfx = os.path.normpath(util.localpath(prefix)) prefix = util.pconvert(lpfx) if not prefix.endswith('/'): prefix += '/' # Drop the leading '.' path component if present, so Windows can read the # zip files (issue4634) if prefix.startswith('./'): prefix = prefix[2:] if prefix.startswith('../') or os.path.isabs(lpfx) or '/../' in prefix: raise error.Abort(_('archive prefix contains illegal components')) return prefix exts = { 'tar': ['.tar'], 'tbz2': ['.tbz2', '.tar.bz2'], 'tgz': ['.tgz', '.tar.gz'], 'zip': ['.zip'], } def guesskind(dest): for kind, extensions in exts.iteritems(): if any(dest.endswith(ext) for ext in extensions): return kind return None def _rootctx(repo): # repo[0] may be hidden for rev in repo: return repo[rev] return repo['null'] def buildmetadata(ctx): '''build content of .hg_archival.txt''' repo = ctx.repo() hex = ctx.hex() if ctx.rev() is None: hex = ctx.p1().hex() if ctx.dirty(): hex += '+' base = 'repo: %s\nnode: %s\nbranch: %s\n' % ( _rootctx(repo).hex(), hex, encoding.fromlocal(ctx.branch())) tags = ''.join('tag: %s\n' % t for t in ctx.tags() if repo.tagtype(t) == 'global') if not tags: repo.ui.pushbuffer() opts = {'template': '{latesttag}\n{latesttagdistance}\n' '{changessincelatesttag}', 'style': '', 'patch': None, 'git': None} cmdutil.show_changeset(repo.ui, repo, opts).show(ctx) ltags, dist, changessince = repo.ui.popbuffer().split('\n') ltags = ltags.split(':') tags = ''.join('latesttag: %s\n' % t for t in ltags) tags += 'latesttagdistance: %s\n' % dist tags += 'changessincelatesttag: %s\n' % changessince return base + tags class tarit(object): '''write archive to tar file or stream. can write uncompressed, or compress with gzip or bzip2.''' class GzipFileWithTime(gzip.GzipFile): def __init__(self, *args, **kw): timestamp = None if 'timestamp' in kw: timestamp = kw.pop('timestamp') if timestamp is None: self.timestamp = time.time() else: self.timestamp = timestamp gzip.GzipFile.__init__(self, *args, **kw) def _write_gzip_header(self): self.fileobj.write('\037\213') # magic header self.fileobj.write('\010') # compression method fname = self.name if fname and fname.endswith('.gz'): fname = fname[:-3] flags = 0 if fname: flags = gzip.FNAME self.fileobj.write(chr(flags)) gzip.write32u(self.fileobj, long(self.timestamp)) self.fileobj.write('\002') self.fileobj.write('\377') if fname: self.fileobj.write(fname + '\000') def __init__(self, dest, mtime, kind=''): self.mtime = mtime self.fileobj = None def taropen(name, mode, fileobj=None): if kind == 'gz': mode = mode[0] if not fileobj: fileobj = open(name, mode + 'b') gzfileobj = self.GzipFileWithTime(name, mode + 'b', zlib.Z_BEST_COMPRESSION, fileobj, timestamp=mtime) self.fileobj = gzfileobj return tarfile.TarFile.taropen(name, mode, gzfileobj) else: return tarfile.open(name, mode + kind, fileobj) if isinstance(dest, str): self.z = taropen(dest, mode='w:') else: # Python 2.5-2.5.1 have a regression that requires a name arg self.z = taropen(name='', mode='w|', fileobj=dest) def addfile(self, name, mode, islink, data): i = tarfile.TarInfo(name) i.mtime = self.mtime i.size = len(data) if islink: i.type = tarfile.SYMTYPE i.mode = 0o777 i.linkname = data data = None i.size = 0 else: i.mode = mode data = stringio(data) self.z.addfile(i, data) def done(self): self.z.close() if self.fileobj: self.fileobj.close() class tellable(object): '''provide tell method for zipfile.ZipFile when writing to http response file object.''' def __init__(self, fp): self.fp = fp self.offset = 0 def __getattr__(self, key): return getattr(self.fp, key) def write(self, s): self.fp.write(s) self.offset += len(s) def tell(self): return self.offset class zipit(object): '''write archive to zip file or stream. can write uncompressed, or compressed with deflate.''' def __init__(self, dest, mtime, compress=True): if not isinstance(dest, str): try: dest.tell() except (AttributeError, IOError): dest = tellable(dest) self.z = zipfile.ZipFile(dest, 'w', compress and zipfile.ZIP_DEFLATED or zipfile.ZIP_STORED) # Python's zipfile module emits deprecation warnings if we try # to store files with a date before 1980. epoch = 315532800 # calendar.timegm((1980, 1, 1, 0, 0, 0, 1, 1, 0)) if mtime < epoch: mtime = epoch self.mtime = mtime self.date_time = time.gmtime(mtime)[:6] def addfile(self, name, mode, islink, data): i = zipfile.ZipInfo(name, self.date_time) i.compress_type = self.z.compression # unzip will not honor unix file modes unless file creator is # set to unix (id 3). i.create_system = 3 ftype = _UNX_IFREG if islink: mode = 0o777 ftype = _UNX_IFLNK i.external_attr = (mode | ftype) << 16L # add "extended-timestamp" extra block, because zip archives # without this will be extracted with unexpected timestamp, # if TZ is not configured as GMT i.extra += struct.pack('<hhBl', 0x5455, # block type: "extended-timestamp" 1 + 4, # size of this block 1, # "modification time is present" int(self.mtime)) # last modification (UTC) self.z.writestr(i, data) def done(self): self.z.close() class fileit(object): '''write archive as files in directory.''' def __init__(self, name, mtime): self.basedir = name self.opener = scmutil.opener(self.basedir) def addfile(self, name, mode, islink, data): if islink: self.opener.symlink(data, name) return f = self.opener(name, "w", atomictemp=True) f.write(data) f.close() destfile = os.path.join(self.basedir, name) os.chmod(destfile, mode) def done(self): pass archivers = { 'files': fileit, 'tar': tarit, 'tbz2': lambda name, mtime: tarit(name, mtime, 'bz2'), 'tgz': lambda name, mtime: tarit(name, mtime, 'gz'), 'uzip': lambda name, mtime: zipit(name, mtime, False), 'zip': zipit, } def archive(repo, dest, node, kind, decode=True, matchfn=None, prefix='', mtime=None, subrepos=False): '''create archive of repo as it was at node. dest can be name of directory, name of archive file, or file object to write archive to. kind is type of archive to create. decode tells whether to put files through decode filters from hgrc. matchfn is function to filter names of files to write to archive. prefix is name of path to put before every archive member.''' if kind == 'files': if prefix: raise error.Abort(_('cannot give prefix when archiving to files')) else: prefix = tidyprefix(dest, kind, prefix) def write(name, mode, islink, getdata): data = getdata() if decode: data = repo.wwritedata(name, data) archiver.addfile(prefix + name, mode, islink, data) if kind not in archivers: raise error.Abort(_("unknown archive type '%s'") % kind) ctx = repo[node] archiver = archivers[kind](dest, mtime or ctx.date()[0]) if repo.ui.configbool("ui", "archivemeta", True): name = '.hg_archival.txt' if not matchfn or matchfn(name): write(name, 0o644, False, lambda: buildmetadata(ctx)) if matchfn: files = [f for f in ctx.manifest().keys() if matchfn(f)] else: files = ctx.manifest().keys() total = len(files) if total: files.sort() repo.ui.progress(_('archiving'), 0, unit=_('files'), total=total) for i, f in enumerate(files): ff = ctx.flags(f) write(f, 'x' in ff and 0o755 or 0o644, 'l' in ff, ctx[f].data) repo.ui.progress(_('archiving'), i + 1, item=f, unit=_('files'), total=total) repo.ui.progress(_('archiving'), None) if subrepos: for subpath in sorted(ctx.substate): sub = ctx.workingsub(subpath) submatch = matchmod.subdirmatcher(subpath, matchfn) total += sub.archive(archiver, prefix, submatch) if total == 0: raise error.Abort(_('no files match the archive pattern')) archiver.done() return total