view tests/test-bookflow.t @ 44950:f9734b2d59cc

py3: make stdout line-buffered if connected to a TTY

Status messages that are to be shown on the terminal should be written to the
file descriptor before anything further is done, to keep the user updated.

One common way to achieve this is to make stdout line-buffered if it is
connected to a TTY. This is done on Python 2 (except on Windows, where libc,
which the CPython 2 streams depend on, does not properly support this).

Python 3 rolls it own I/O streams. On Python 3, buffered binary streams can't be
set line-buffered. The previous code (added in 227ba1afcb65) incorrectly
assumed that on Python 3, pycompat.stdout (sys.stdout.buffer) is already
line-buffered. However the interpreter initializes it with a block-buffered
stream or an unbuffered stream (when the -u option or the PYTHONUNBUFFERED
environment variable is set), never with a line-buffered stream.

One example where the current behavior is unacceptable is when running
`hg pull https://www.mercurial-scm.org/repo/hg` on Python 3, where the line
"pulling from https://www.mercurial-scm.org/repo/hg" does not appear on the
terminal before the hg process blocks while waiting for the server.

Various approaches to fix this problem are possible, including:

1. Weaken the contract of procutil.stdout to not give any guarantees about
   buffering behavior. In this case, users of procutil.stdout need to be
   changed to do enough flushes. In particular,
  1. either ui must insert enough flushes for ui.write() and friends, or
  2. ui.write() and friends get split into flushing and fully buffered
     methods, or
  3. users of ui.write() and friends must flush explicitly.
2. Make stdout unbuffered.
3. Make stdout line-buffered. Since Python 3 does not natively support that for
   binary streams, we must implement it ourselves.

(2.) is problematic because using unbuffered I/O changes the performance
characteristics significantly compared to line-buffered (which is used on
Python 2) and this would be a regression.

(1.2.) and (1.3) are a substantial amount of work. It’s unclear whether the
added complexity would be justified, given that raw performance doesn’t matter
that much when writing to a terminal much faster than the user could read it.

(1.1.) pushes complexity into the ui class instead of separating the concern of
how stdout is buffered. Other users of procutil.stdout would still need to take
care of the flushes.

This patch implements (3.). The general performance considerations are very
similar to (1.1.). The extra method invocation and method forwarding add a
little more overhead if the class is used. In exchange, it doesn’t add overhead
if not used.


For the benchmarks, I compared the previous implementation (incorrect on Python
3), (1.1.), (3.) and (2.). The command was chosen so that the streams were
configured as if they were writing to a TTY, but actually write to a pager,
which is also the default:
HGRCPATH=/dev/null python3 ./hg --cwd ~/vcs/mozilla-central --time --pager yes --config pager.pager='cat > /dev/null' status --all

previous:
time: real 7.880 secs (user 7.290+0.050 sys 0.580+0.170)
time: real 7.830 secs (user 7.220+0.070 sys 0.590+0.140)
time: real 7.800 secs (user 7.210+0.050 sys 0.570+0.170)

(1.1.) using Yuya Nishihara’s patch:
time: real 9.860 secs (user 8.670+0.350 sys 1.160+0.830)
time: real 9.540 secs (user 8.430+0.370 sys 1.100+0.770)
time: real 9.830 secs (user 8.630+0.370 sys 1.180+0.840)

(3.) using this patch:
time: real 9.580 secs (user 8.480+0.350 sys 1.090+0.770)
time: real 9.670 secs (user 8.480+0.330 sys 1.170+0.860)
time: real 9.640 secs (user 8.500+0.350 sys 1.130+0.810)

(2.) using a previous patch by me:
time: real 10.480 secs (user 8.850+0.720 sys 1.590+1.500)
time: real 10.490 secs (user 8.750+0.750 sys 1.710+1.470)
time: real 10.240 secs (user 8.600+0.700 sys 1.590+1.510)


As expected, there’s no difference on Python 2, as exactly the same code paths
are used:

previous:
time: real 6.950 secs (user 5.870+0.330 sys 1.070+0.770)
time: real 7.040 secs (user 6.040+0.360 sys 0.980+0.750)
time: real 7.070 secs (user 5.950+0.360 sys 1.100+0.760)

this patch:
time: real 7.010 secs (user 5.900+0.390 sys 1.070+0.730)
time: real 7.000 secs (user 5.850+0.350 sys 1.120+0.760)
time: real 7.000 secs (user 5.790+0.380 sys 1.170+0.710)

initialize
  $ make_changes() {
  >     d=`pwd`
  >     [ ! -z $1 ] && cd $1
  >     echo "test `basename \`pwd\``" >> test
  >     hg commit -Am"${2:-test}"
  >     r=$?
  >     cd $d
  >     return $r
  > }
  $ ls -1a
  .
  ..
  $ hg init a
  $ cd a
  $ echo 'test' > test; hg commit -Am'test'
  adding test

clone to b

  $ mkdir ../b
  $ cd ../b
  $ hg clone ../a .
  updating to branch default
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ echo "[extensions]" >> .hg/hgrc
  $ echo "bookflow=" >> .hg/hgrc
  $ hg branch X
  abort: creating named branches is disabled and you should use bookmarks
  (see 'hg help bookflow')
  [255]
  $ hg bookmark X
  $ hg bookmarks
  * X                         0:* (glob)
  $ hg bookmark X
  abort: bookmark X already exists, to move use the --rev option
  [255]
  $ make_changes
  $ hg push ../a -q

  $ hg bookmarks
   \* X                         1:* (glob)

change a
  $ cd ../a
  $ hg up
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ echo 'test' >> test; hg commit -Am'test'


pull in b
  $ cd ../b
  $ hg pull -u
  pulling from $TESTTMP/a
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 1 changesets with 1 changes to 1 files
  new changesets * (glob)
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  (leaving bookmark X)
  $ hg status
  $ hg bookmarks
     X                         1:* (glob)

check protection of @ bookmark
  $ hg bookmark @
  $ hg bookmarks
   \* @                         2:* (glob)
     X                         1:* (glob)
  $ make_changes
  abort: cannot commit, bookmark @ is protected
  [255]

  $ hg status
  M test
  $ hg bookmarks
   \* @                         2:* (glob)
     X                         1:* (glob)

  $ hg --config bookflow.protect= commit  -Am"Updated test"

  $ hg bookmarks
   \* @                         3:* (glob)
     X                         1:* (glob)

check requirement for an active bookmark
  $ hg bookmark -i
  $ hg bookmarks
     @                         3:* (glob)
     X                         1:* (glob)
  $ make_changes
  abort: cannot commit without an active bookmark
  [255]
  $ hg revert test
  $ rm test.orig
  $ hg status


make the bookmark move by updating it on a, and then pulling
# add a commit to a
  $ cd ../a
  $ hg bookmark X
  $ hg bookmarks
   \* X                         2:* (glob)
  $ make_changes
  $ hg bookmarks
   * X                         3:81af7977fdb9

# go back to b, and check out X
  $ cd ../b
  $ hg up X
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  (activating bookmark X)
  $ hg bookmarks
     @                         3:* (glob)
   \* X                         1:* (glob)

# pull, this should move the bookmark forward, because it was changed remotely
  $ hg pull -u | grep "updating to active bookmark X"
  updating to active bookmark X

  $ hg bookmarks
     @                         3:* (glob)
   * X                         4:81af7977fdb9

the bookmark should not move if it diverged from remote
  $ hg -R ../a status
  $ hg -R ../b status
  $ make_changes ../a
  $ make_changes ../b
  $ hg -R ../a status
  $ hg -R ../b status
  $ hg -R ../a bookmarks
   * X                         4:238292f60a57
  $ hg -R ../b bookmarks
     @                         3:* (glob)
   * X                         5:096f7e86892d
  $ cd ../b
  $ # make sure we cannot push after bookmarks diverged
  $ hg push -B X | grep abort
  abort: push creates new remote head * with bookmark 'X'! (glob)
  (pull and merge or see 'hg help push' for details about pushing new heads)
  [1]
  $ hg pull -u | grep divergent
  divergent bookmark X stored as X@default
  1 other divergent bookmarks for "X"
  $ hg bookmarks
     @                         3:* (glob)
   * X                         5:096f7e86892d
     X@default                 6:238292f60a57
  $ hg id -in
  096f7e86892d 5
  $ make_changes
  $ hg status
  $ hg bookmarks
     @                         3:* (glob)
   * X                         7:227f941aeb07
     X@default                 6:238292f60a57

now merge with the remote bookmark
  $ hg merge X@default --tool :local -q
  $ hg status
  M test
  $ hg commit -m"Merged with X@default"
  $ hg bookmarks
     @                         3:* (glob)
   * X                         8:26fed9bb3219
  $ hg push -B X | grep bookmark
  pushing to $TESTTMP/a (?)
  updating bookmark X
  $ cd ../a
  $ hg up -q
  $ hg bookmarks
   * X                         7:26fed9bb3219

test hg pull when there is more than one descendant
  $ cd ../a
  $ hg bookmark Z
  $ hg bookmark Y
  $ make_changes . YY
  $ hg up Z -q
  $ make_changes . ZZ
  created new head
  $ hg bookmarks
     X                         7:26fed9bb3219
     Y                         8:131e663dbd2a
   * Z                         9:b74a4149df25
  $ hg log -r 'p1(Y)' -r 'p1(Z)' -T '{rev}\n' # prove that Y and Z share the same parent
  7
  $ hg log -r 'Y%Z' -T '{rev}\n'  # revs in Y but not in Z
  8
  $ hg log -r 'Z%Y' -T '{rev}\n'  # revs in Z but not in Y
  9
  $ cd ../b
  $ hg pull -uq
  $ hg id
  b74a4149df25 tip Z
  $ hg bookmarks | grep \*  # no active bookmark
  [1]


test shelving
  $ cd ../a
  $ echo anotherfile > anotherfile # this change should not conflict
  $ hg add anotherfile
  $ hg commit -m"Change in a"
  $ cd ../b
  $ hg up Z | grep Z
  (activating bookmark Z)
  $ hg book | grep \* # make sure active bookmark
   \* Z                         10:* (glob)
  $ echo "test b" >> test
  $ hg diff --stat
   test |  1 +
   1 files changed, 1 insertions(+), 0 deletions(-)
  $ hg shelve
  shelved as Z
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ hg pull -uq
  $ hg unshelve
  unshelving change 'Z'
  rebasing shelved changes
  $ hg diff --stat
   test |  1 +
   1 files changed, 1 insertions(+), 0 deletions(-)


make the bookmark move by updating it on a, and then pulling with a local change
# add a commit to a
  $ cd ../a
  $ hg up -C X |fgrep  "activating bookmark X"
  (activating bookmark X)
# go back to b, and check out X
  $ cd ../b
  $ hg up -C X |fgrep  "activating bookmark X"
  (activating bookmark X)
# update and push from a
  $ make_changes ../a
  created new head
  $ echo "more" >> test
  $ hg pull -u 2>&1 | fgrep -v TESTTMP| fgrep -v "searching for changes" | fgrep -v adding
  pulling from $TESTTMP/a
  updating bookmark X
  added 1 changesets with 0 changes to 0 files (+1 heads)
  new changesets * (glob)
  updating to active bookmark X
  merging test
  warning: conflicts while merging test! (edit, then use 'hg resolve --mark')
  0 files updated, 0 files merged, 0 files removed, 1 files unresolved
  use 'hg resolve' to retry unresolved file merges
  $ hg update -Cq
  $ rm test.orig

make sure that commits aren't possible if working directory is not pointing to active bookmark
  $ hg -R ../a status
  $ hg -R ../b status
  $ hg -R ../a id -i
  36a6e592ec06
  $ hg -R ../a book | grep X
   \* X                         \d+:36a6e592ec06 (re)
  $ hg -R ../b id -i
  36a6e592ec06
  $ hg -R ../b book | grep X
   \* X                         \d+:36a6e592ec06 (re)
  $ make_changes ../a
  $ hg -R ../a book | grep X
   \* X                         \d+:f73a71c992b8 (re)
  $ cd ../b
  $ hg pull  2>&1 | grep -v add | grep -v pulling | grep -v searching | grep -v changeset
  updating bookmark X
  (run 'hg update' to get a working copy)
  working directory out of sync with active bookmark, run 'hg up X'
  $ hg id -i # we're still on the old commit
  36a6e592ec06
  $ hg book | grep X # while the bookmark moved
   \* X                         \d+:f73a71c992b8 (re)
  $ make_changes
  abort: cannot commit, working directory out of sync with active bookmark
  (run 'hg up X')
  [255]
  $ hg up -Cq -r .  # cleanup local changes
  $ hg status
  $ hg id -i # we're still on the old commit
  36a6e592ec06
  $ hg up X -q
  $ hg id -i # now we're on X
  f73a71c992b8
  $ hg book | grep X
   \* X                         \d+:f73a71c992b8 (re)