view tests/test-context.py @ 44363:f7459da77f23

nodemap: introduce an option to use mmap to read the nodemap mapping The performance and memory benefit is much greater if we don't have to copy all the data in memory for each information. So we introduce an option (on by default) to read the data using mmap. This changeset is the last one definition the API for index support nodemap data. (they have to be able to use the mmaping). Below are some benchmark comparing the best we currently have in 5.3 with the final step of this series (using the persistent nodemap implementation in Rust). The benchmark run `hg perfindex` with various revset and the following variants: Before: * do not use the persistent nodemap * use the CPython implementation of the index for nodemap * use mmapping of the changelog index After: * use the MixedIndex Rust code, with the NodeTree object for nodemap access (still in review) * use the persistent nodemap data from disk * access the persistent nodemap data through mmap * use mmapping of the changelog index The persistent nodemap greatly speed up most operation on very large repositories. Some of the previously very fast lookup end up a bit slower because the persistent nodemap has to be setup. However the absolute slowdown is very small and won't matters in the big picture. Here are some numbers (in seconds) for the reference copy of mozilla-try: Revset Before After abs-change speedup -10000: 0.004622 0.005532 0.000910 × 0.83 -10: 0.000050 0.000132 0.000082 × 0.37 tip 0.000052 0.000085 0.000033 × 0.61 0 + (-10000:) 0.028222 0.005337 -0.022885 × 5.29 0 0.023521 0.000084 -0.023437 × 280.01 (-10000:) + 0 0.235539 0.005308 -0.230231 × 44.37 (-10:) + :9 0.232883 0.000180 -0.232703 ×1293.79 (-10000:) + (:99) 0.238735 0.005358 -0.233377 × 44.55 :99 + (-10000:) 0.317942 0.005593 -0.312349 × 56.84 :9 + (-10:) 0.313372 0.000179 -0.313193 ×1750.68 :9 0.316450 0.000143 -0.316307 ×2212.93 On smaller repositories, the cost of nodemap related operation is not as big, so the win is much more modest. Yet it helps shaving a handful of millisecond here and there. Here are some numbers (in seconds) for the reference copy of mercurial: Revset Before After abs-change speedup -10: 0.000065 0.000097 0.000032 × 0.67 tip 0.000063 0.000078 0.000015 × 0.80 0 0.000561 0.000079 -0.000482 × 7.10 -10000: 0.004609 0.003648 -0.000961 × 1.26 0 + (-10000:) 0.005023 0.003715 -0.001307 × 1.35 (-10:) + :9 0.002187 0.000108 -0.002079 ×20.25 (-10000:) + 0 0.006252 0.003716 -0.002536 × 1.68 (-10000:) + (:99) 0.006367 0.003707 -0.002660 × 1.71 :9 + (-10:) 0.003846 0.000110 -0.003736 ×34.96 :9 0.003854 0.000099 -0.003755 ×38.92 :99 + (-10000:) 0.007644 0.003778 -0.003866 × 2.02 Differential Revision: https://phab.mercurial-scm.org/D7894
author Pierre-Yves David <pierre-yves.david@octobus.net>
date Tue, 11 Feb 2020 11:18:52 +0100
parents 2372284d9457
children 03ef0c8fa7d5
line wrap: on
line source

from __future__ import absolute_import, print_function
import os
import stat
import sys
from mercurial.node import hex
from mercurial import (
    context,
    diffutil,
    encoding,
    hg,
    scmutil,
    ui as uimod,
)

print_ = print


def print(*args, **kwargs):
    """print() wrapper that flushes stdout buffers to avoid py3 buffer issues

    We could also just write directly to sys.stdout.buffer the way the
    ui object will, but this was easier for porting the test.
    """
    print_(*args, **kwargs)
    sys.stdout.flush()


def printb(data, end=b'\n'):
    out = getattr(sys.stdout, 'buffer', sys.stdout)
    out.write(data + end)
    out.flush()


ui = uimod.ui.load()

repo = hg.repository(ui, b'test1', create=1)
os.chdir('test1')

# create 'foo' with fixed time stamp
f = open('foo', 'wb')
f.write(b'foo\n')
f.close()
os.utime('foo', (1000, 1000))

# add+commit 'foo'
repo[None].add([b'foo'])
repo.commit(text=b'commit1', date=b"0 0")

d = repo[None][b'foo'].date()
if os.name == 'nt':
    d = d[:2]
print("workingfilectx.date = (%d, %d)" % d)

# test memctx with non-ASCII commit message


def filectxfn(repo, memctx, path):
    return context.memfilectx(repo, memctx, b"foo", b"")


ctx = context.memctx(
    repo,
    [b'tip', None],
    encoding.tolocal(b"Gr\xc3\xbcezi!"),
    [b"foo"],
    filectxfn,
)
ctx.commit()
for enc in "ASCII", "Latin-1", "UTF-8":
    encoding.encoding = enc
    printb(b"%-8s: %s" % (enc.encode('ascii'), repo[b"tip"].description()))

# test performing a status


def getfilectx(repo, memctx, f):
    fctx = memctx.p1()[f]
    data, flags = fctx.data(), fctx.flags()
    if f == b'foo':
        data += b'bar\n'
    return context.memfilectx(
        repo, memctx, f, data, b'l' in flags, b'x' in flags
    )


ctxa = repo[0]
ctxb = context.memctx(
    repo,
    [ctxa.node(), None],
    b"test diff",
    [b"foo"],
    getfilectx,
    ctxa.user(),
    ctxa.date(),
)

print(ctxb.status(ctxa))

# test performing a diff on a memctx
diffopts = diffutil.diffallopts(repo.ui, {b'git': True})
for d in ctxb.diff(ctxa, opts=diffopts):
    printb(d, end=b'')

# test safeness and correctness of "ctx.status()"
print('= checking context.status():')

# ancestor "wcctx ~ 2"
actx2 = repo[b'.']

repo.wwrite(b'bar-m', b'bar-m\n', b'')
repo.wwrite(b'bar-r', b'bar-r\n', b'')
repo[None].add([b'bar-m', b'bar-r'])
repo.commit(text=b'add bar-m, bar-r', date=b"0 0")

# ancestor "wcctx ~ 1"
actx1 = repo[b'.']

repo.wwrite(b'bar-m', b'bar-m bar-m\n', b'')
repo.wwrite(b'bar-a', b'bar-a\n', b'')
repo[None].add([b'bar-a'])
repo[None].forget([b'bar-r'])

# status at this point:
#   M bar-m
#   A bar-a
#   R bar-r
#   C foo

from mercurial import scmutil

print('== checking workingctx.status:')

wctx = repo[None]
print('wctx._status=%s' % (str(wctx._status)))

print('=== with "pattern match":')
print(
    actx1.status(other=wctx, match=scmutil.matchfiles(repo, [b'bar-m', b'foo']))
)
print('wctx._status=%s' % (str(wctx._status)))
print(
    actx2.status(other=wctx, match=scmutil.matchfiles(repo, [b'bar-m', b'foo']))
)
print('wctx._status=%s' % (str(wctx._status)))

print('=== with "always match" and "listclean=True":')
print(actx1.status(other=wctx, listclean=True))
print('wctx._status=%s' % (str(wctx._status)))
print(actx2.status(other=wctx, listclean=True))
print('wctx._status=%s' % (str(wctx._status)))

print("== checking workingcommitctx.status:")

wcctx = context.workingcommitctx(
    repo,
    scmutil.status([b'bar-m'], [b'bar-a'], [], [], [], [], []),
    text=b'',
    date=b'0 0',
)
print('wcctx._status=%s' % (str(wcctx._status)))

print('=== with "always match":')
print(actx1.status(other=wcctx))
print('wcctx._status=%s' % (str(wcctx._status)))
print(actx2.status(other=wcctx))
print('wcctx._status=%s' % (str(wcctx._status)))

print('=== with "always match" and "listclean=True":')
print(actx1.status(other=wcctx, listclean=True))
print('wcctx._status=%s' % (str(wcctx._status)))
print(actx2.status(other=wcctx, listclean=True))
print('wcctx._status=%s' % (str(wcctx._status)))

print('=== with "pattern match":')
print(
    actx1.status(
        other=wcctx, match=scmutil.matchfiles(repo, [b'bar-m', b'foo'])
    )
)
print('wcctx._status=%s' % (str(wcctx._status)))
print(
    actx2.status(
        other=wcctx, match=scmutil.matchfiles(repo, [b'bar-m', b'foo'])
    )
)
print('wcctx._status=%s' % (str(wcctx._status)))

print('=== with "pattern match" and "listclean=True":')
print(
    actx1.status(
        other=wcctx,
        match=scmutil.matchfiles(repo, [b'bar-r', b'foo']),
        listclean=True,
    )
)
print('wcctx._status=%s' % (str(wcctx._status)))
print(
    actx2.status(
        other=wcctx,
        match=scmutil.matchfiles(repo, [b'bar-r', b'foo']),
        listclean=True,
    )
)
print('wcctx._status=%s' % (str(wcctx._status)))

os.chdir('..')

# test manifestlog being changed
print('== commit with manifestlog invalidated')

repo = hg.repository(ui, b'test2', create=1)
os.chdir('test2')

# make some commits
for i in [b'1', b'2', b'3']:
    with open(i, 'wb') as f:
        f.write(i)
    status = scmutil.status([], [i], [], [], [], [], [])
    ctx = context.workingcommitctx(
        repo, status, text=i, user=b'test@test.com', date=(0, 0)
    )
    ctx.p1().manifest()  # side effect: cache manifestctx
    n = repo.commitctx(ctx)
    printb(b'commit %s: %s' % (i, hex(n)))

    # touch 00manifest.i mtime so storecache could expire.
    # repo.__dict__['manifestlog'] is deleted by transaction releasefn.
    st = repo.svfs.stat(b'00manifest.i')
    repo.svfs.utime(
        b'00manifest.i', (st[stat.ST_MTIME] + 1, st[stat.ST_MTIME] + 1)
    )

    # read the file just committed
    try:
        if repo[n][i].data() != i:
            print('data mismatch')
    except Exception as ex:
        print('cannot read data: %r' % ex)

with repo.wlock(), repo.lock(), repo.transaction(b'test'):
    with open(b'4', 'wb') as f:
        f.write(b'4')
    repo.dirstate.normal(b'4')
    repo.commit(b'4')
    revsbefore = len(repo.changelog)
    repo.invalidate(clearfilecache=True)
    revsafter = len(repo.changelog)
    if revsbefore != revsafter:
        print('changeset lost by repo.invalidate()')