view tests/test-fastannotate-revmap.py @ 44118:f81c17ec303c

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 ± σ):      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 ± σ):      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 ± σ):      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 ± σ):      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 ± σ):      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 ± σ):      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

from __future__ import absolute_import, print_function

import os
import tempfile

from mercurial import (
    pycompat,
    util,
)

from hgext.fastannotate import error, revmap

if pycompat.ispy3:
    xrange = range


def genhsh(i):
    return pycompat.bytechr(i) + b'\0' * 19


def gettemppath():
    fd, path = tempfile.mkstemp()
    os.close(fd)
    os.unlink(path)
    return path


def ensure(condition):
    if not condition:
        raise RuntimeError('Unexpected')


def testbasicreadwrite():
    path = gettemppath()

    rm = revmap.revmap(path)
    ensure(rm.maxrev == 0)
    for i in xrange(5):
        ensure(rm.rev2hsh(i) is None)
    ensure(rm.hsh2rev(b'\0' * 20) is None)

    paths = [
        b'',
        b'a',
        None,
        b'b',
        b'b',
        b'c',
        b'c',
        None,
        b'a',
        b'b',
        b'a',
        b'a',
    ]
    for i in xrange(1, 5):
        ensure(rm.append(genhsh(i), sidebranch=(i & 1), path=paths[i]) == i)

    ensure(rm.maxrev == 4)
    for i in xrange(1, 5):
        ensure(rm.hsh2rev(genhsh(i)) == i)
        ensure(rm.rev2hsh(i) == genhsh(i))

    # re-load and verify
    rm.flush()
    rm = revmap.revmap(path)
    ensure(rm.maxrev == 4)
    for i in xrange(1, 5):
        ensure(rm.hsh2rev(genhsh(i)) == i)
        ensure(rm.rev2hsh(i) == genhsh(i))
        ensure(bool(rm.rev2flag(i) & revmap.sidebranchflag) == bool(i & 1))

    # append without calling save() explicitly
    for i in xrange(5, 12):
        ensure(
            rm.append(genhsh(i), sidebranch=(i & 1), path=paths[i], flush=True)
            == i
        )

    # re-load and verify
    rm = revmap.revmap(path)
    ensure(rm.maxrev == 11)
    for i in xrange(1, 12):
        ensure(rm.hsh2rev(genhsh(i)) == i)
        ensure(rm.rev2hsh(i) == genhsh(i))
        ensure(rm.rev2path(i) == paths[i] or paths[i - 1])
        ensure(bool(rm.rev2flag(i) & revmap.sidebranchflag) == bool(i & 1))

    os.unlink(path)

    # missing keys
    ensure(rm.rev2hsh(12) is None)
    ensure(rm.rev2hsh(0) is None)
    ensure(rm.rev2hsh(-1) is None)
    ensure(rm.rev2flag(12) is None)
    ensure(rm.rev2path(12) is None)
    ensure(rm.hsh2rev(b'\1' * 20) is None)

    # illformed hash (not 20 bytes)
    try:
        rm.append(b'\0')
        ensure(False)
    except Exception:
        pass


def testcorruptformat():
    path = gettemppath()

    # incorrect header
    with open(path, 'wb') as f:
        f.write(b'NOT A VALID HEADER')
    try:
        revmap.revmap(path)
        ensure(False)
    except error.CorruptedFileError:
        pass

    # rewrite the file
    os.unlink(path)
    rm = revmap.revmap(path)
    rm.append(genhsh(0), flush=True)

    rm = revmap.revmap(path)
    ensure(rm.maxrev == 1)

    # corrupt the file by appending a byte
    size = os.stat(path).st_size
    with open(path, 'ab') as f:
        f.write(b'\xff')
    try:
        revmap.revmap(path)
        ensure(False)
    except error.CorruptedFileError:
        pass

    # corrupt the file by removing the last byte
    ensure(size > 0)
    with open(path, 'wb') as f:
        f.truncate(size - 1)
    try:
        revmap.revmap(path)
        ensure(False)
    except error.CorruptedFileError:
        pass

    os.unlink(path)


def testcopyfrom():
    path = gettemppath()
    rm = revmap.revmap(path)
    for i in xrange(1, 10):
        ensure(
            rm.append(genhsh(i), sidebranch=(i & 1), path=(b'%d' % (i // 3)))
            == i
        )
    rm.flush()

    # copy rm to rm2
    rm2 = revmap.revmap()
    rm2.copyfrom(rm)
    path2 = gettemppath()
    rm2.path = path2
    rm2.flush()

    # two files should be the same
    ensure(len(set(util.readfile(p) for p in [path, path2])) == 1)

    os.unlink(path)
    os.unlink(path2)


class fakefctx(object):
    def __init__(self, node, path=None):
        self._node = node
        self._path = path

    def node(self):
        return self._node

    def path(self):
        return self._path


def testcontains():
    path = gettemppath()

    rm = revmap.revmap(path)
    for i in xrange(1, 5):
        ensure(rm.append(genhsh(i), sidebranch=(i & 1)) == i)

    for i in xrange(1, 5):
        ensure(((genhsh(i), None) in rm) == ((i & 1) == 0))
        ensure((fakefctx(genhsh(i)) in rm) == ((i & 1) == 0))
    for i in xrange(5, 10):
        ensure(fakefctx(genhsh(i)) not in rm)
        ensure((genhsh(i), None) not in rm)

    # "contains" checks paths
    rm = revmap.revmap()
    for i in xrange(1, 5):
        ensure(rm.append(genhsh(i), path=(b'%d' % (i // 2))) == i)
    for i in xrange(1, 5):
        ensure(fakefctx(genhsh(i), path=(b'%d' % (i // 2))) in rm)
        ensure(fakefctx(genhsh(i), path=b'a') not in rm)


def testlastnode():
    path = gettemppath()
    ensure(revmap.getlastnode(path) is None)
    rm = revmap.revmap(path)
    ensure(revmap.getlastnode(path) is None)
    for i in xrange(1, 10):
        hsh = genhsh(i)
        rm.append(hsh, path=(b'%d' % (i // 2)), flush=True)
        ensure(revmap.getlastnode(path) == hsh)
        rm2 = revmap.revmap(path)
        ensure(rm2.rev2hsh(rm2.maxrev) == hsh)


testbasicreadwrite()
testcorruptformat()
testcopyfrom()
testcontains()
testlastnode()