hgdemandimport: apply lazy module loading to sys.meta_path finders
Python's `sys.meta_path` finders are the primary objects whose job it
is to find a module at import time. When `import` is called, Python
iterates objects in this list and calls `o.find_spec(...)` to find
a `ModuleSpec` (or None if the module couldn't be found by that
finder). If no meta path finder can find a module, import fails.
One of the default meta path finders is `PathFinder`. Its job is to
import modules from the filesystem and is probably the most important
importer. This finder looks at `sys.path` and `sys.path_hooks` to do
its job.
The `ModuleSpec` returned by `MetaPathImporter.find_spec()` has a
`loader` attribute, which defines the concrete module loader to use.
`sys.path_hooks` is a hook point for teaching `PathFinder` to
instantiate custom loader types.
Previously, we injected a custom `sys.path_hook` that told `PathFinder`
to wrap the default loaders with a loader that creates a module object
that is lazy.
This approach worked. But its main limitation was that it only applied
to the `PathFinder` meta path importer. There are other meta path
importers that are registered. And in the case of PyOxidizer loading
modules from memory, `PathFinder` doesn't come into play since
PyOxidizer's own meta path importer was handling all imports.
This commit changes our approach to lazy module loading by proxying
all meta path importers. Specifically, we overload the `find_spec()`
method to swap in a wrapped loader on the `ModuleSpec` before it
is returned. The end result of this is all meta path importers should
be lazy.
As much as I would have loved to utilize .__class__ manipulation to
achieve this, some meta path importers are implemented in C/Rust
in such a way that they cannot be monkeypatched. This is why we
use __getattribute__ to define a proxy.
Also, this change could theoretically open us up to regressions in
meta path importers whose loader is creating module objects which
can't be monkeypatched. But I'm not aware of any of these in the
wild. So I think we'll be safe.
According to hyperfine, this change yields a decent startup time win of
5-6ms:
```
Benchmark #1: ~/.pyenv/versions/3.6.10/bin/python ./hg version
Time (mean ? ?): 86.8 ms ? 0.5 ms [User: 78.0 ms, System: 8.7 ms]
Range (min ? max): 86.0 ms ? 89.1 ms 50 runs
Time (mean ? ?): 81.1 ms ? 2.7 ms [User: 74.5 ms, System: 6.5 ms]
Range (min ? max): 77.8 ms ? 90.5 ms 50 runs
Benchmark #2: ~/.pyenv/versions/3.7.6/bin/python ./hg version
Time (mean ? ?): 78.9 ms ? 0.6 ms [User: 70.2 ms, System: 8.7 ms]
Range (min ? max): 78.1 ms ? 81.2 ms 50 runs
Time (mean ? ?): 73.4 ms ? 0.6 ms [User: 65.3 ms, System: 8.0 ms]
Range (min ? max): 72.4 ms ? 75.7 ms 50 runs
Benchmark #3: ~/.pyenv/versions/3.8.1/bin/python ./hg version
Time (mean ? ?): 78.1 ms ? 0.6 ms [User: 70.2 ms, System: 7.9 ms]
Range (min ? max): 77.4 ms ? 80.9 ms 50 runs
Time (mean ? ?): 72.1 ms ? 0.4 ms [User: 64.4 ms, System: 7.6 ms]
Range (min ? max): 71.4 ms ? 74.1 ms 50 runs
```
Differential Revision: https://phab.mercurial-scm.org/D7954
# statichttprepo.py - simple http repository class for mercurial
#
# This provides read-only repo access to repositories exported via static http
#
# Copyright 2005-2007 Matt Mackall <mpm@selenic.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 errno
from .i18n import _
from . import (
branchmap,
changelog,
error,
localrepo,
manifest,
namespaces,
pathutil,
pycompat,
url,
util,
vfs as vfsmod,
)
urlerr = util.urlerr
urlreq = util.urlreq
class httprangereader(object):
def __init__(self, url, opener):
# we assume opener has HTTPRangeHandler
self.url = url
self.pos = 0
self.opener = opener
self.name = url
def __enter__(self):
return self
def __exit__(self, exc_type, exc_value, traceback):
self.close()
def seek(self, pos):
self.pos = pos
def read(self, bytes=None):
req = urlreq.request(pycompat.strurl(self.url))
end = b''
if bytes:
end = self.pos + bytes - 1
if self.pos or end:
req.add_header('Range', 'bytes=%d-%s' % (self.pos, end))
try:
f = self.opener.open(req)
data = f.read()
code = f.code
except urlerr.httperror as inst:
num = inst.code == 404 and errno.ENOENT or None
raise IOError(num, inst)
except urlerr.urlerror as inst:
raise IOError(None, inst.reason)
if code == 200:
# HTTPRangeHandler does nothing if remote does not support
# Range headers and returns the full entity. Let's slice it.
if bytes:
data = data[self.pos : self.pos + bytes]
else:
data = data[self.pos :]
elif bytes:
data = data[:bytes]
self.pos += len(data)
return data
def readlines(self):
return self.read().splitlines(True)
def __iter__(self):
return iter(self.readlines())
def close(self):
pass
# _RangeError and _HTTPRangeHandler were originally in byterange.py,
# which was itself extracted from urlgrabber. See the last version of
# byterange.py from history if you need more information.
class _RangeError(IOError):
"""Error raised when an unsatisfiable range is requested."""
class _HTTPRangeHandler(urlreq.basehandler):
"""Handler that enables HTTP Range headers.
This was extremely simple. The Range header is a HTTP feature to
begin with so all this class does is tell urllib2 that the
"206 Partial Content" response from the HTTP server is what we
expected.
"""
def http_error_206(self, req, fp, code, msg, hdrs):
# 206 Partial Content Response
r = urlreq.addinfourl(fp, hdrs, req.get_full_url())
r.code = code
r.msg = msg
return r
def http_error_416(self, req, fp, code, msg, hdrs):
# HTTP's Range Not Satisfiable error
raise _RangeError(b'Requested Range Not Satisfiable')
def build_opener(ui, authinfo):
# urllib cannot handle URLs with embedded user or passwd
urlopener = url.opener(ui, authinfo)
urlopener.add_handler(_HTTPRangeHandler())
class statichttpvfs(vfsmod.abstractvfs):
def __init__(self, base):
self.base = base
self.options = {}
def __call__(self, path, mode=b'r', *args, **kw):
if mode not in (b'r', b'rb'):
raise IOError(b'Permission denied')
f = b"/".join((self.base, urlreq.quote(path)))
return httprangereader(f, urlopener)
def join(self, path):
if path:
return pathutil.join(self.base, path)
else:
return self.base
return statichttpvfs
class statichttppeer(localrepo.localpeer):
def local(self):
return None
def canpush(self):
return False
class statichttprepository(
localrepo.localrepository, localrepo.revlogfilestorage
):
supported = localrepo.localrepository._basesupported
def __init__(self, ui, path):
self._url = path
self.ui = ui
self.root = path
u = util.url(path.rstrip(b'/') + b"/.hg")
self.path, authinfo = u.authinfo()
vfsclass = build_opener(ui, authinfo)
self.vfs = vfsclass(self.path)
self.cachevfs = vfsclass(self.vfs.join(b'cache'))
self._phasedefaults = []
self.names = namespaces.namespaces()
self.filtername = None
self._extrafilterid = None
try:
requirements = set(self.vfs.read(b'requires').splitlines())
except IOError as inst:
if inst.errno != errno.ENOENT:
raise
requirements = set()
# check if it is a non-empty old-style repository
try:
fp = self.vfs(b"00changelog.i")
fp.read(1)
fp.close()
except IOError as inst:
if inst.errno != errno.ENOENT:
raise
# we do not care about empty old-style repositories here
msg = _(b"'%s' does not appear to be an hg repository") % path
raise error.RepoError(msg)
supportedrequirements = localrepo.gathersupportedrequirements(ui)
localrepo.ensurerequirementsrecognized(
requirements, supportedrequirements
)
localrepo.ensurerequirementscompatible(ui, requirements)
# setup store
self.store = localrepo.makestore(requirements, self.path, vfsclass)
self.spath = self.store.path
self.svfs = self.store.opener
self.sjoin = self.store.join
self._filecache = {}
self.requirements = requirements
rootmanifest = manifest.manifestrevlog(self.svfs)
self.manifestlog = manifest.manifestlog(
self.svfs, self, rootmanifest, self.narrowmatch()
)
self.changelog = changelog.changelog(self.svfs)
self._tags = None
self.nodetagscache = None
self._branchcaches = branchmap.BranchMapCache()
self._revbranchcache = None
self.encodepats = None
self.decodepats = None
self._transref = None
def _restrictcapabilities(self, caps):
caps = super(statichttprepository, self)._restrictcapabilities(caps)
return caps.difference([b"pushkey"])
def url(self):
return self._url
def local(self):
return False
def peer(self):
return statichttppeer(self)
def wlock(self, wait=True):
raise error.LockUnavailable(
0,
_(b'lock not available'),
b'lock',
_(b'cannot lock static-http repository'),
)
def lock(self, wait=True):
raise error.Abort(_(b'cannot lock static-http repository'))
def _writecaches(self):
pass # statichttprepository are read only
def instance(ui, path, create, intents=None, createopts=None):
if create:
raise error.Abort(_(b'cannot create new static-http repository'))
return statichttprepository(ui, path[7:])