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
# urllibcompat.py - adapters to ease using urllib2 on Py2 and urllib on Py3
#
# Copyright 2017 Google, Inc.
#
# 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
from .pycompat import getattr
from . import pycompat
_sysstr = pycompat.sysstr
class _pycompatstub(object):
def __init__(self):
self._aliases = {}
def _registeraliases(self, origin, items):
"""Add items that will be populated at the first access"""
items = map(_sysstr, items)
self._aliases.update(
(item.replace('_', '').lower(), (origin, item)) for item in items
)
def _registeralias(self, origin, attr, name):
"""Alias ``origin``.``attr`` as ``name``"""
self._aliases[_sysstr(name)] = (origin, _sysstr(attr))
def __getattr__(self, name):
try:
origin, item = self._aliases[name]
except KeyError:
raise AttributeError(name)
self.__dict__[name] = obj = getattr(origin, item)
return obj
httpserver = _pycompatstub()
urlreq = _pycompatstub()
urlerr = _pycompatstub()
if pycompat.ispy3:
import urllib.parse
urlreq._registeraliases(
urllib.parse,
(
b"splitattr",
b"splitpasswd",
b"splitport",
b"splituser",
b"urlparse",
b"urlunparse",
),
)
urlreq._registeralias(urllib.parse, b"parse_qs", b"parseqs")
urlreq._registeralias(urllib.parse, b"parse_qsl", b"parseqsl")
urlreq._registeralias(urllib.parse, b"unquote_to_bytes", b"unquote")
import urllib.request
urlreq._registeraliases(
urllib.request,
(
b"AbstractHTTPHandler",
b"BaseHandler",
b"build_opener",
b"FileHandler",
b"FTPHandler",
b"ftpwrapper",
b"HTTPHandler",
b"HTTPSHandler",
b"install_opener",
b"pathname2url",
b"HTTPBasicAuthHandler",
b"HTTPDigestAuthHandler",
b"HTTPPasswordMgrWithDefaultRealm",
b"ProxyHandler",
b"Request",
b"url2pathname",
b"urlopen",
),
)
import urllib.response
urlreq._registeraliases(urllib.response, (b"addclosehook", b"addinfourl",))
import urllib.error
urlerr._registeraliases(urllib.error, (b"HTTPError", b"URLError",))
import http.server
httpserver._registeraliases(
http.server,
(
b"HTTPServer",
b"BaseHTTPRequestHandler",
b"SimpleHTTPRequestHandler",
b"CGIHTTPRequestHandler",
),
)
# urllib.parse.quote() accepts both str and bytes, decodes bytes
# (if necessary), and returns str. This is wonky. We provide a custom
# implementation that only accepts bytes and emits bytes.
def quote(s, safe='/'):
# bytestr has an __iter__ that emits characters. quote_from_bytes()
# does an iteration and expects ints. We coerce to bytes to appease it.
if isinstance(s, pycompat.bytestr):
s = bytes(s)
s = urllib.parse.quote_from_bytes(s, safe=safe)
return s.encode('ascii', 'strict')
# urllib.parse.urlencode() returns str. We use this function to make
# sure we return bytes.
def urlencode(query, doseq=False):
s = urllib.parse.urlencode(query, doseq=doseq)
return s.encode('ascii')
urlreq.quote = quote
urlreq.urlencode = urlencode
def getfullurl(req):
return req.full_url
def gethost(req):
return req.host
def getselector(req):
return req.selector
def getdata(req):
return req.data
def hasdata(req):
return req.data is not None
else:
import BaseHTTPServer
import CGIHTTPServer
import SimpleHTTPServer
import urllib2
import urllib
import urlparse
urlreq._registeraliases(
urllib,
(
b"addclosehook",
b"addinfourl",
b"ftpwrapper",
b"pathname2url",
b"quote",
b"splitattr",
b"splitpasswd",
b"splitport",
b"splituser",
b"unquote",
b"url2pathname",
b"urlencode",
),
)
urlreq._registeraliases(
urllib2,
(
b"AbstractHTTPHandler",
b"BaseHandler",
b"build_opener",
b"FileHandler",
b"FTPHandler",
b"HTTPBasicAuthHandler",
b"HTTPDigestAuthHandler",
b"HTTPHandler",
b"HTTPPasswordMgrWithDefaultRealm",
b"HTTPSHandler",
b"install_opener",
b"ProxyHandler",
b"Request",
b"urlopen",
),
)
urlreq._registeraliases(urlparse, (b"urlparse", b"urlunparse",))
urlreq._registeralias(urlparse, b"parse_qs", b"parseqs")
urlreq._registeralias(urlparse, b"parse_qsl", b"parseqsl")
urlerr._registeraliases(urllib2, (b"HTTPError", b"URLError",))
httpserver._registeraliases(
BaseHTTPServer, (b"HTTPServer", b"BaseHTTPRequestHandler",)
)
httpserver._registeraliases(
SimpleHTTPServer, (b"SimpleHTTPRequestHandler",)
)
httpserver._registeraliases(CGIHTTPServer, (b"CGIHTTPRequestHandler",))
def gethost(req):
return req.get_host()
def getselector(req):
return req.get_selector()
def getfullurl(req):
return req.get_full_url()
def getdata(req):
return req.get_data()
def hasdata(req):
return req.has_data()