view tests/test-parse-date.t @ 51681:522b4d729e89

mmap: populate the mapping by default

Without pre-population, accessing all data through a mmap can result in many
pagefault, reducing performance significantly. If the mmap is prepopulated, the
performance can no longer get slower than a full read.
(See benchmark number below)

In some cases were very few data is read, prepopulating can be overkill and
slower than populating on access (through page fault). So that behavior can be
controlled when the caller can pre-determine the best behavior.
(See benchmark number below)

In addition, testing with populating in a secondary thread yield great result
combining the best of each approach. This might be implemented in later
changesets.

In all cases, using mmap has a great effect on memory usage when many processes
run in parallel on the same machine.

### Benchmarks

# What did I run

A couple of month back I ran a large benchmark campaign to assess the impact of
various approach for using mmap with the revlog (and other files), it
highlighted a few benchmarks that capture the impact of the changes well. So to
validate this change I checked the following:

- log command displaying various revisions
  (read the changelog index)
- log command displaying the patch of listed revisions
  (read the changelog index, the manifest index and a few files indexes)
- unbundling a few revisions
  (read and write changelog, manifest and few files indexes, and walk the graph
  to update some cache)
- pushing a few revisions
  (read and write changelog, manifest and few files indexes, walk the graph to
  update some cache, performs various accesses locally and remotely during
  discovery)

Benchmarks were run using the default module policy (c+py) and the rust one. No
significant difference were found between the two implementation, so we will
present result using the default policy (unless otherwise specified).

I ran them on a few repositories :

- mercurial: a "public changeset only" copy of mercurial from 2018-08-01 using
  zstd compression and sparse-revlog

- pypy: a copy of pypy from 2018-08-01 using zstd compression and sparse-revlog

- netbeans: a copy of netbeans from 2018-08-01 using zstd compression and
  sparse-revlog

- mozilla-try: a copy of mozilla-try from 2019-02-18 using zstd compression and
  sparse-revlog

- mozilla-try persistent-nodemap: Same as the above but with a persistent
  nodemap. Used for the log --patch benchmark only

# Results

For the smaller repositories (mercurial, pypy), the impact of mmap is almost
imperceptible, other cost dominating the operation. The impact of prepopulating
is undiscernible in the benchmark we ran.

For larger repositories the benchmark support explanation given above:

On netbeans, the log can be about 1% faster without repopulation (for a
difference < 100ms) but unbundle becomes a bit slower, even when small.

### data-env-vars.name                             = netbeans-2018-08-01-zstd-sparse-revlog
  # benchmark.name                                 = hg.command.unbundle
  # benchmark.variants.issue6528                   = disabled
  # benchmark.variants.reuse-external-delta-parent = yes
  # benchmark.variants.revs                        = any-1-extra-rev
  # benchmark.variants.source                      = unbundle
  # benchmark.variants.verbosity                   = quiet
with-populate: 0.240157
no-populate:   0.265087  (+10.38%, +0.02)
  # benchmark.variants.revs                        = any-100-extra-rev
with-populate: 1.459518
no-populate:   1.481290   (+1.49%, +0.02)
 ## benchmark.name                                 = hg.command.push
  # benchmark.variants.explicit-rev                = none
  # benchmark.variants.issue6528                   = disabled
  # benchmark.variants.protocol                    = ssh
  # benchmark.variants.reuse-external-delta-parent = yes
  # benchmark.variants.revs                        = any-1-extra-rev
with-populate: 0.771919
no-populate:   0.792025   (+2.60%, +0.02)
  # benchmark.variants.revs                        = any-100-extra-rev
with-populate: 1.459518
no-populate:   1.481290   (+1.49%, +0.02)


For mozilla-try, the "slow down" from pre-populate for small `hg log` is more
visible, but still small in absolute time. (using rust value for the persistent
nodemap value to be relevant).

### data-env-vars.name            = mozilla-try-2019-02-18-ds2-pnm
  # benchmark.name                = hg.command.log
  # bin-env-vars.hg.flavor        = rust
  # benchmark.variants.patch      = yes
  # benchmark.variants.limit-rev  = 1
with-populate: 0.237813
no-populate:   0.229452   (-3.52%, -0.01)
  # benchmark.variants.limit-rev  = 10
  # benchmark.variants.patch      = yes
with-populate: 1.213578
no-populate:   1.205189
### data-env-vars.name            = mozilla-try-2019-02-18-zstd-sparse-revlog
  # benchmark.variants.limit-rev  = 1000
  # benchmark.variants.patch      = no
  # benchmark.variants.rev        = tip
with-populate: 0.198607
no-populate:   0.195038   (-1.80%, -0.00)

However pre-populating provide a significant boost on more complex operations
like unbundle or push:

### data-env-vars.name                             = mozilla-try-2019-02-18-zstd-sparse-revlog
  # benchmark.name                                 = hg.command.push
  # benchmark.variants.explicit-rev                = none
  # benchmark.variants.issue6528                   = disabled
  # benchmark.variants.protocol                    = ssh
  # benchmark.variants.reuse-external-delta-parent = yes
  # benchmark.variants.revs                        = any-1-extra-rev
with-populate: 4.798632
no-populate:   4.953295   (+3.22%, +0.15)
  # benchmark.variants.revs                        = any-100-extra-rev
with-populate: 4.903618
no-populate:   5.014963   (+2.27%, +0.11)
 ## benchmark.name                                 = hg.command.unbundle
  # benchmark.variants.revs                        = any-1-extra-rev
with-populate: 1.423411
no-populate:   1.585365  (+11.38%, +0.16)
  # benchmark.variants.revs                        = any-100-extra-rev
with-populate: 1.537909
no-populate:   1.688489   (+9.79%, +0.15)

This runs with TZ="GMT"

  $ hg init
  $ echo "test-parse-date" > a
  $ hg add a
  $ hg ci -d "2006-02-01 13:00:30" -m "rev 0"
  $ echo "hi!" >> a
  $ hg ci -d "2006-02-01 13:00:30 -0500" -m "rev 1"
  $ hg tag -d "2006-04-15 13:30" "Hi"
  $ hg backout --merge -d "2006-04-15 13:30 +0200" -m "rev 3" 1
  reverting a
  created new head
  changeset 3:107ce1ee2b43 backs out changeset 1:25a1420a55f8
  merging with changeset 3:107ce1ee2b43
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  (branch merge, don't forget to commit)
  $ hg ci -d "1150000000 14400" -m "rev 4 (merge)"
  $ echo "fail" >> a
  $ hg ci -d "should fail" -m "fail"
  hg: parse error: invalid date: 'should fail'
  [10]
  $ hg ci -d "100000000000000000 1400" -m "fail"
  hg: parse error: date exceeds 32 bits: 100000000000000000
  [10]
  $ hg ci -d "100000 1400000" -m "fail"
  hg: parse error: impossible time zone offset: 1400000
  [10]

Check with local timezone other than GMT and with DST

  $ TZ="PST+8PDT+7,M4.1.0/02:00:00,M10.5.0/02:00:00"
  $ export TZ

PST=UTC-8 / PDT=UTC-7
Summer time begins on April's first Sunday at 2:00am,
and ends on October's last Sunday at 2:00am.

  $ hg debugrebuildstate
  $ echo "a" > a
  $ hg ci -d "2006-07-15 13:30" -m "summer@UTC-7"
  $ hg debugrebuildstate
  $ echo "b" > a
  $ hg ci -d "2006-07-15 13:30 +0500" -m "summer@UTC+5"
  $ hg debugrebuildstate
  $ echo "c" > a
  $ hg ci -d "2006-01-15 13:30" -m "winter@UTC-8"
  $ hg debugrebuildstate
  $ echo "d" > a
  $ hg ci -d "2006-01-15 13:30 +0500" -m "winter@UTC+5"
  $ hg log --template '{date|date}\n'
  Sun Jan 15 13:30:00 2006 +0500
  Sun Jan 15 13:30:00 2006 -0800
  Sat Jul 15 13:30:00 2006 +0500
  Sat Jul 15 13:30:00 2006 -0700
  Sun Jun 11 00:26:40 2006 -0400
  Sat Apr 15 13:30:00 2006 +0200
  Sat Apr 15 13:30:00 2006 +0000
  Wed Feb 01 13:00:30 2006 -0500
  Wed Feb 01 13:00:30 2006 +0000

Test issue1014 (fractional timezones)

  $ hg debugdate "1000000000 -16200" # 0430
  internal: 1000000000 -16200
  standard: Sun Sep 09 06:16:40 2001 +0430
  $ hg debugdate "1000000000 -15300" # 0415
  internal: 1000000000 -15300
  standard: Sun Sep 09 06:01:40 2001 +0415
  $ hg debugdate "1000000000 -14400" # 0400
  internal: 1000000000 -14400
  standard: Sun Sep 09 05:46:40 2001 +0400
  $ hg debugdate "1000000000 0"      # GMT
  internal: 1000000000 0
  standard: Sun Sep 09 01:46:40 2001 +0000
  $ hg debugdate "1000000000 14400"  # -0400
  internal: 1000000000 14400
  standard: Sat Sep 08 21:46:40 2001 -0400
  $ hg debugdate "1000000000 15300"  # -0415
  internal: 1000000000 15300
  standard: Sat Sep 08 21:31:40 2001 -0415
  $ hg debugdate "1000000000 16200"  # -0430
  internal: 1000000000 16200
  standard: Sat Sep 08 21:16:40 2001 -0430
  $ hg debugdate "Sat Sep 08 21:16:40 2001 +0430"
  internal: 999967600 -16200
  standard: Sat Sep 08 21:16:40 2001 +0430
  $ hg debugdate "Sat Sep 08 21:16:40 2001 -0430"
  internal: 1000000000 16200
  standard: Sat Sep 08 21:16:40 2001 -0430

Test 12-hours times

  $ hg debugdate "2006-02-01 1:00:30PM +0000"
  internal: 1138798830 0
  standard: Wed Feb 01 13:00:30 2006 +0000
  $ hg debugdate "1:00:30PM" > /dev/null

Normal range

  $ hg log -d -1

Negative range

  $ hg log -d "--2"
  abort: -2 must be nonnegative (see 'hg help dates')
  [10]

Whitespace only

  $ hg log -d " "
  abort: dates cannot consist entirely of whitespace
  [10]

Test date formats with '>' or '<' accompanied by space characters

  $ hg log -d '>' --template '{date|date}\n'
  abort: invalid day spec, use '>DATE'
  [10]
  $ hg log -d '<' --template '{date|date}\n'
  abort: invalid day spec, use '<DATE'
  [10]

  $ hg log -d ' >' --template '{date|date}\n'
  abort: invalid day spec, use '>DATE'
  [10]
  $ hg log -d ' <' --template '{date|date}\n'
  abort: invalid day spec, use '<DATE'
  [10]

  $ hg log -d '> ' --template '{date|date}\n'
  abort: invalid day spec, use '>DATE'
  [10]
  $ hg log -d '< ' --template '{date|date}\n'
  abort: invalid day spec, use '<DATE'
  [10]

  $ hg log -d ' > ' --template '{date|date}\n'
  abort: invalid day spec, use '>DATE'
  [10]
  $ hg log -d ' < ' --template '{date|date}\n'
  abort: invalid day spec, use '<DATE'
  [10]

  $ hg log -d '>02/01' --template '{date|date}\n'
  $ hg log -d '<02/01' --template '{date|date}\n'
  Sun Jan 15 13:30:00 2006 +0500
  Sun Jan 15 13:30:00 2006 -0800
  Sat Jul 15 13:30:00 2006 +0500
  Sat Jul 15 13:30:00 2006 -0700
  Sun Jun 11 00:26:40 2006 -0400
  Sat Apr 15 13:30:00 2006 +0200
  Sat Apr 15 13:30:00 2006 +0000
  Wed Feb 01 13:00:30 2006 -0500
  Wed Feb 01 13:00:30 2006 +0000

  $ hg log -d ' >02/01' --template '{date|date}\n'
  $ hg log -d ' <02/01' --template '{date|date}\n'
  Sun Jan 15 13:30:00 2006 +0500
  Sun Jan 15 13:30:00 2006 -0800
  Sat Jul 15 13:30:00 2006 +0500
  Sat Jul 15 13:30:00 2006 -0700
  Sun Jun 11 00:26:40 2006 -0400
  Sat Apr 15 13:30:00 2006 +0200
  Sat Apr 15 13:30:00 2006 +0000
  Wed Feb 01 13:00:30 2006 -0500
  Wed Feb 01 13:00:30 2006 +0000

  $ hg log -d '> 02/01' --template '{date|date}\n'
  $ hg log -d '< 02/01' --template '{date|date}\n'
  Sun Jan 15 13:30:00 2006 +0500
  Sun Jan 15 13:30:00 2006 -0800
  Sat Jul 15 13:30:00 2006 +0500
  Sat Jul 15 13:30:00 2006 -0700
  Sun Jun 11 00:26:40 2006 -0400
  Sat Apr 15 13:30:00 2006 +0200
  Sat Apr 15 13:30:00 2006 +0000
  Wed Feb 01 13:00:30 2006 -0500
  Wed Feb 01 13:00:30 2006 +0000

  $ hg log -d ' > 02/01' --template '{date|date}\n'
  $ hg log -d ' < 02/01' --template '{date|date}\n'
  Sun Jan 15 13:30:00 2006 +0500
  Sun Jan 15 13:30:00 2006 -0800
  Sat Jul 15 13:30:00 2006 +0500
  Sat Jul 15 13:30:00 2006 -0700
  Sun Jun 11 00:26:40 2006 -0400
  Sat Apr 15 13:30:00 2006 +0200
  Sat Apr 15 13:30:00 2006 +0000
  Wed Feb 01 13:00:30 2006 -0500
  Wed Feb 01 13:00:30 2006 +0000

  $ hg log -d '>02/01 ' --template '{date|date}\n'
  $ hg log -d '<02/01 ' --template '{date|date}\n'
  Sun Jan 15 13:30:00 2006 +0500
  Sun Jan 15 13:30:00 2006 -0800
  Sat Jul 15 13:30:00 2006 +0500
  Sat Jul 15 13:30:00 2006 -0700
  Sun Jun 11 00:26:40 2006 -0400
  Sat Apr 15 13:30:00 2006 +0200
  Sat Apr 15 13:30:00 2006 +0000
  Wed Feb 01 13:00:30 2006 -0500
  Wed Feb 01 13:00:30 2006 +0000

  $ hg log -d ' >02/01 ' --template '{date|date}\n'
  $ hg log -d ' <02/01 ' --template '{date|date}\n'
  Sun Jan 15 13:30:00 2006 +0500
  Sun Jan 15 13:30:00 2006 -0800
  Sat Jul 15 13:30:00 2006 +0500
  Sat Jul 15 13:30:00 2006 -0700
  Sun Jun 11 00:26:40 2006 -0400
  Sat Apr 15 13:30:00 2006 +0200
  Sat Apr 15 13:30:00 2006 +0000
  Wed Feb 01 13:00:30 2006 -0500
  Wed Feb 01 13:00:30 2006 +0000

  $ hg log -d '> 02/01 ' --template '{date|date}\n'
  $ hg log -d '< 02/01 ' --template '{date|date}\n'
  Sun Jan 15 13:30:00 2006 +0500
  Sun Jan 15 13:30:00 2006 -0800
  Sat Jul 15 13:30:00 2006 +0500
  Sat Jul 15 13:30:00 2006 -0700
  Sun Jun 11 00:26:40 2006 -0400
  Sat Apr 15 13:30:00 2006 +0200
  Sat Apr 15 13:30:00 2006 +0000
  Wed Feb 01 13:00:30 2006 -0500
  Wed Feb 01 13:00:30 2006 +0000

  $ hg log -d ' > 02/01 ' --template '{date|date}\n'
  $ hg log -d ' < 02/01 ' --template '{date|date}\n'
  Sun Jan 15 13:30:00 2006 +0500
  Sun Jan 15 13:30:00 2006 -0800
  Sat Jul 15 13:30:00 2006 +0500
  Sat Jul 15 13:30:00 2006 -0700
  Sun Jun 11 00:26:40 2006 -0400
  Sat Apr 15 13:30:00 2006 +0200
  Sat Apr 15 13:30:00 2006 +0000
  Wed Feb 01 13:00:30 2006 -0500
  Wed Feb 01 13:00:30 2006 +0000

Test issue 3764 (interpreting 'today' and 'yesterday')
  $ echo "hello" >> a
  >>> import datetime
  >>> today = datetime.date.today().strftime("%b %d")
  >>> yesterday = (datetime.date.today() - datetime.timedelta(days=1)).strftime("%b %d")
  >>> dates = open('dates', 'w')
  >>> dates.write(today + '\n') and None
  >>> dates.write(yesterday + '\n') and None
  >>> dates.close()
  $ hg ci -d "`sed -n '1p' dates`" -m "today is a good day to code"
  $ hg log -d today --template '{desc}\n'
  today is a good day to code
  $ echo "goodbye" >> a
  $ hg ci -d "`sed -n '2p' dates`" -m "the time traveler's code"
  $ hg log -d yesterday --template '{desc}\n'
  the time traveler's code
  $ echo "foo" >> a
  $ hg commit -d now -m 'Explicitly committed now.'
  $ hg log -d today --template '{desc}\n'
  Explicitly committed now.
  today is a good day to code

Test parsing various ISO8601 forms

  $ hg debugdate "2016-07-27T12:10:21"
  internal: 1469646621 * (glob)
  standard: Wed Jul 27 12:10:21 2016 -0700
  $ hg debugdate "2016-07-27T12:10:21Z"
  internal: 1469621421 0
  standard: Wed Jul 27 12:10:21 2016 +0000
  $ hg debugdate "2016-07-27T12:10:21+00:00"
  internal: 1469621421 0
  standard: Wed Jul 27 12:10:21 2016 +0000
  $ hg debugdate "2016-07-27T121021Z"
  internal: 1469621421 0
  standard: Wed Jul 27 12:10:21 2016 +0000

  $ hg debugdate "2016-07-27 12:10:21"
  internal: 1469646621 * (glob)
  standard: Wed Jul 27 12:10:21 2016 -0700
  $ hg debugdate "2016-07-27 12:10:21Z"
  internal: 1469621421 0
  standard: Wed Jul 27 12:10:21 2016 +0000
  $ hg debugdate "2016-07-27 12:10:21+00:00"
  internal: 1469621421 0
  standard: Wed Jul 27 12:10:21 2016 +0000
  $ hg debugdate "2016-07-27 121021Z"
  internal: 1469621421 0
  standard: Wed Jul 27 12:10:21 2016 +0000

Test parsing months

  $ for i in Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec; do
  >   hg log -d "$i 2018" -r null
  > done