view tests/test-absorb-filefixupstate.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 itertools
from mercurial import pycompat
from hgext import absorb


class simplefctx(object):
    def __init__(self, content):
        self.content = content

    def data(self):
        return self.content


def insertreturns(x):
    # insert "\n"s after each single char
    if isinstance(x, bytes):
        return b''.join(ch + b'\n' for ch in pycompat.bytestr(x))
    else:
        return pycompat.maplist(insertreturns, x)


def removereturns(x):
    # the revert of "insertreturns"
    if isinstance(x, bytes):
        return x.replace(b'\n', b'')
    else:
        return pycompat.maplist(removereturns, x)


def assertlistequal(lhs, rhs, decorator=lambda x: x):
    if lhs != rhs:
        raise RuntimeError(
            'mismatch:\n actual:   %r\n expected: %r'
            % tuple(map(decorator, [lhs, rhs]))
        )


def testfilefixup(oldcontents, workingcopy, expectedcontents, fixups=None):
    """([str], str, [str], [(rev, a1, a2, b1, b2)]?) -> None

    workingcopy is a string, of which every character denotes a single line.

    oldcontents, expectedcontents are lists of strings, every character of
    every string denots a single line.

    if fixups is not None, it's the expected fixups list and will be checked.
    """
    expectedcontents = insertreturns(expectedcontents)
    oldcontents = insertreturns(oldcontents)
    workingcopy = insertreturns(workingcopy)
    state = absorb.filefixupstate(
        pycompat.maplist(simplefctx, oldcontents), 'path'
    )
    state.diffwith(simplefctx(workingcopy))
    if fixups is not None:
        assertlistequal(state.fixups, fixups)
    state.apply()
    assertlistequal(state.finalcontents, expectedcontents, removereturns)


def buildcontents(linesrevs):
    # linesrevs: [(linecontent : str, revs : [int])]
    revs = set(itertools.chain(*[revs for line, revs in linesrevs]))
    return [b''] + [
        b''.join([l for l, rs in linesrevs if r in rs]) for r in sorted(revs)
    ]


# input case 0: one single commit
case0 = [b'', b'11']

# replace a single chunk
testfilefixup(case0, b'', [b'', b''])
testfilefixup(case0, b'2', [b'', b'2'])
testfilefixup(case0, b'22', [b'', b'22'])
testfilefixup(case0, b'222', [b'', b'222'])

# input case 1: 3 lines, each commit adds one line
case1 = buildcontents([(b'1', [1, 2, 3]), (b'2', [2, 3]), (b'3', [3]),])

# 1:1 line mapping
testfilefixup(case1, b'123', case1)
testfilefixup(case1, b'12c', [b'', b'1', b'12', b'12c'])
testfilefixup(case1, b'1b3', [b'', b'1', b'1b', b'1b3'])
testfilefixup(case1, b'1bc', [b'', b'1', b'1b', b'1bc'])
testfilefixup(case1, b'a23', [b'', b'a', b'a2', b'a23'])
testfilefixup(case1, b'a2c', [b'', b'a', b'a2', b'a2c'])
testfilefixup(case1, b'ab3', [b'', b'a', b'ab', b'ab3'])
testfilefixup(case1, b'abc', [b'', b'a', b'ab', b'abc'])

# non 1:1 edits
testfilefixup(case1, b'abcd', case1)
testfilefixup(case1, b'ab', case1)

# deletion
testfilefixup(case1, b'', [b'', b'', b'', b''])
testfilefixup(case1, b'1', [b'', b'1', b'1', b'1'])
testfilefixup(case1, b'2', [b'', b'', b'2', b'2'])
testfilefixup(case1, b'3', [b'', b'', b'', b'3'])
testfilefixup(case1, b'13', [b'', b'1', b'1', b'13'])

# replaces
testfilefixup(case1, b'1bb3', [b'', b'1', b'1bb', b'1bb3'])

# (confusing) replaces
testfilefixup(case1, b'1bbb', case1)
testfilefixup(case1, b'bbbb', case1)
testfilefixup(case1, b'bbb3', case1)
testfilefixup(case1, b'1b', case1)
testfilefixup(case1, b'bb', case1)
testfilefixup(case1, b'b3', case1)

# insertions at the beginning and the end
testfilefixup(case1, b'123c', [b'', b'1', b'12', b'123c'])
testfilefixup(case1, b'a123', [b'', b'a1', b'a12', b'a123'])

# (confusing) insertions
testfilefixup(case1, b'1a23', case1)
testfilefixup(case1, b'12b3', case1)

# input case 2: delete in the middle
case2 = buildcontents([(b'11', [1, 2]), (b'22', [1]), (b'33', [1, 2]),])

# deletion (optimize code should make it 2 chunks)
testfilefixup(
    case2, b'', [b'', b'22', b''], fixups=[(4, 0, 2, 0, 0), (4, 2, 4, 0, 0)]
)

# 1:1 line mapping
testfilefixup(case2, b'aaaa', [b'', b'aa22aa', b'aaaa'])

# non 1:1 edits
# note: unlike case0, the chunk is not "continuous" and no edit allowed
testfilefixup(case2, b'aaa', case2)

# input case 3: rev 3 reverts rev 2
case3 = buildcontents([(b'1', [1, 2, 3]), (b'2', [2]), (b'3', [1, 2, 3]),])

# 1:1 line mapping
testfilefixup(case3, b'13', case3)
testfilefixup(case3, b'1b', [b'', b'1b', b'12b', b'1b'])
testfilefixup(case3, b'a3', [b'', b'a3', b'a23', b'a3'])
testfilefixup(case3, b'ab', [b'', b'ab', b'a2b', b'ab'])

# non 1:1 edits
testfilefixup(case3, b'a', case3)
testfilefixup(case3, b'abc', case3)

# deletion
testfilefixup(case3, b'', [b'', b'', b'2', b''])

# insertion
testfilefixup(case3, b'a13c', [b'', b'a13c', b'a123c', b'a13c'])

# input case 4: a slightly complex case
case4 = buildcontents(
    [
        (b'1', [1, 2, 3]),
        (b'2', [2, 3]),
        (b'3', [1, 2,]),
        (b'4', [1, 3]),
        (b'5', [3]),
        (b'6', [2, 3]),
        (b'7', [2]),
        (b'8', [2, 3]),
        (b'9', [3]),
    ]
)

testfilefixup(case4, b'1245689', case4)
testfilefixup(case4, b'1a2456bbb', case4)
testfilefixup(case4, b'1abc5689', case4)
testfilefixup(case4, b'1ab5689', [b'', b'134', b'1a3678', b'1ab5689'])
testfilefixup(case4, b'aa2bcd8ee', [b'', b'aa34', b'aa23d78', b'aa2bcd8ee'])
testfilefixup(case4, b'aa2bcdd8ee', [b'', b'aa34', b'aa23678', b'aa24568ee'])
testfilefixup(case4, b'aaaaaa', case4)
testfilefixup(case4, b'aa258b', [b'', b'aa34', b'aa2378', b'aa258b'])
testfilefixup(case4, b'25bb', [b'', b'34', b'23678', b'25689'])
testfilefixup(case4, b'27', [b'', b'34', b'23678', b'245689'])
testfilefixup(case4, b'28', [b'', b'34', b'2378', b'28'])
testfilefixup(case4, b'', [b'', b'34', b'37', b''])

# input case 5: replace a small chunk which is near a deleted line
case5 = buildcontents([(b'12', [1, 2]), (b'3', [1]), (b'4', [1, 2]),])

testfilefixup(case5, b'1cd4', [b'', b'1cd34', b'1cd4'])

# input case 6: base "changeset" is immutable
case6 = [b'1357', b'0125678']

testfilefixup(case6, b'0125678', case6)
testfilefixup(case6, b'0a25678', case6)
testfilefixup(case6, b'0a256b8', case6)
testfilefixup(case6, b'abcdefg', [b'1357', b'a1c5e7g'])
testfilefixup(case6, b'abcdef', case6)
testfilefixup(case6, b'', [b'1357', b'157'])
testfilefixup(case6, b'0123456789', [b'1357', b'0123456789'])

# input case 7: change an empty file
case7 = [b'']

testfilefixup(case7, b'1', case7)