view tests/test-fastannotate-hg.t @ 40021:c537144fdbef

wireprotov2: support response caching

One of the things I've learned from managing VCS servers over the
years is that they are hard to scale. It is well known that some
companies have very beefy (read: very expensive) servers to power
their VCS needs. It is also known that specialized servers for
various VCS exist in order to facilitate scaling servers. (Mercurial
is in this boat.)

One of the aspects that make a VCS server hard to scale is the
high CPU load incurred by constant client clone/pull operations.

To alleviate the scaling pain associated with data retrieval
operations, I want to integrate caching into the Mercurial wire
protocol server as robustly as possible such that servers can
aggressively cache responses and defer as much server load as
possible.

This commit represents the initial implementation of a general
caching layer in wire protocol version 2.

We define a new interface and behavior for a wire protocol cacher
in repository.py. (This is probably where a reviewer should look
first to understand what is going on.)

The bulk of the added code is in wireprotov2server.py, where we
define how a command can opt in to being cached and integrate
caching into command dispatching.

From a very high-level:

* A command can declare itself as cacheable by providing a callable
  that can be used to derive a cache key.
* At dispatch time, if a command is cacheable, we attempt to
  construct a cacher and use it for serving the request and/or
  caching the request.
* The dispatch layer handles the bulk of the business logic for
  caching, making cachers mostly "dumb content stores."
* The mechanism for invalidating cached entries (one of the harder
  parts about caching in general) is by varying the cache key when
  state changes. As such, cachers don't need to be concerned with
  cache invalidation.

Initially, we've hooked up support for caching "manifestdata" and
"filedata" commands. These are the simplest to cache, as they should
be immutable over time. Caching of commands related to changeset
data is a bit harder (because cache validation is impacted by
changes to bookmarks, phases, etc). This will be implemented later.
(Strictly speaking, censoring a file should invalidate caches. I've
added an inline TODO to track this edge case.)

To prove it works, this commit implements a test-only extension
providing in-memory caching backed by an lrucachedict. A new test
showing this extension behaving properly is added. FWIW, the
cacher is ~50 lines of code, demonstrating the relative ease with
which a cache can be added to a server.

While the test cacher is not suitable for production workloads, just
for kicks I performed a clone of just the changeset and manifest data
for the mozilla-unified repository. With a fully warmed cache (of just
the manifest data since changeset data is not cached), server-side
CPU usage dropped from ~73s to ~28s. That's pretty significant and
demonstrates the potential that response caching has on server
scalability!

Differential Revision: https://phab.mercurial-scm.org/D4773

(this file is backported from core hg tests/test-annotate.t)

  $ cat >> $HGRCPATH << EOF
  > [diff]
  > git=1
  > [extensions]
  > fastannotate=
  > [fastannotate]
  > modes=fctx
  > forcefollow=False
  > mainbranch=.
  > EOF

  $ HGMERGE=true; export HGMERGE

init

  $ hg init repo
  $ cd repo

commit

  $ echo 'a' > a
  $ hg ci -A -m test -u nobody -d '1 0'
  adding a

annotate -c

  $ hg annotate -c a
  8435f90966e4: a

annotate -cl

  $ hg annotate -cl a
  8435f90966e4:1: a

annotate -d

  $ hg annotate -d a
  Thu Jan 01 00:00:01 1970 +0000: a

annotate -n

  $ hg annotate -n a
  0: a

annotate -nl

  $ hg annotate -nl a
  0:1: a

annotate -u

  $ hg annotate -u a
  nobody: a

annotate -cdnu

  $ hg annotate -cdnu a
  nobody 0 8435f90966e4 Thu Jan 01 00:00:01 1970 +0000: a

annotate -cdnul

  $ hg annotate -cdnul a
  nobody 0 8435f90966e4 Thu Jan 01 00:00:01 1970 +0000:1: a

annotate (JSON)

  $ hg annotate -Tjson a
  [
   {
    "lines": [{"line": "a\n", "rev": 0}],
    "path": "a"
   }
  ]

  $ hg annotate -Tjson -cdfnul a
  [
   {
    "lines": [{"date": [1.0, 0], "line": "a\n", "lineno": 1, "node": "8435f90966e442695d2ded29fdade2bac5ad8065", "path": "a", "rev": 0, "user": "nobody"}],
    "path": "a"
   }
  ]

  $ cat <<EOF >>a
  > a
  > a
  > EOF
  $ hg ci -ma1 -d '1 0'
  $ hg cp a b
  $ hg ci -mb -d '1 0'
  $ cat <<EOF >> b
  > b4
  > b5
  > b6
  > EOF
  $ hg ci -mb2 -d '2 0'

annotate -n b

  $ hg annotate -n b
  0: a
  1: a
  1: a
  3: b4
  3: b5
  3: b6

annotate --no-follow b

  $ hg annotate --no-follow b
  2: a
  2: a
  2: a
  3: b4
  3: b5
  3: b6

annotate -nl b

  $ hg annotate -nl b
  0:1: a
  1:2: a
  1:3: a
  3:4: b4
  3:5: b5
  3:6: b6

annotate -nf b

  $ hg annotate -nf b
  0 a: a
  1 a: a
  1 a: a
  3 b: b4
  3 b: b5
  3 b: b6

annotate -nlf b

  $ hg annotate -nlf b
  0 a:1: a
  1 a:2: a
  1 a:3: a
  3 b:4: b4
  3 b:5: b5
  3 b:6: b6

  $ hg up -C 2
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ cat <<EOF >> b
  > b4
  > c
  > b5
  > EOF
  $ hg ci -mb2.1 -d '2 0'
  created new head
  $ hg merge
  merging b
  0 files updated, 1 files merged, 0 files removed, 0 files unresolved
  (branch merge, don't forget to commit)
  $ hg ci -mmergeb -d '3 0'

annotate after merge
(note: the first one falls back to the vanilla annotate which does not use linelog)

  $ hg annotate -nf b --debug
  fastannotate: b: rebuilding broken cache
  fastannotate: b: 5 new changesets in the main branch
  0 a: a
  1 a: a
  1 a: a
  3 b: b4
  4 b: c
  3 b: b5

(difference explained below)

  $ hg annotate -nf b --debug
  fastannotate: b: using fast path (resolved fctx: False)
  0 a: a
  1 a: a
  1 a: a
  4 b: b4
  4 b: c
  4 b: b5

annotate after merge with -l
(fastannotate differs from annotate)

  $ hg log -Gp -T '{rev}:{node}' -r '2..5'
  @    5:64afcdf8e29e063c635be123d8d2fb160af00f7e
  |\
  | o  4:5fbdc1152d97597717021ad9e063061b200f146bdiff --git a/b b/b
  | |  --- a/b
  | |  +++ b/b
  | |  @@ -1,3 +1,6 @@
  | |   a
  | |   a
  | |   a
  | |  +b4
  | |  +c
  | |  +b5
  | |
  o |  3:37ec9f5c3d1f99572d7075971cb4876e2139b52fdiff --git a/b b/b
  |/   --- a/b
  |    +++ b/b
  |    @@ -1,3 +1,6 @@
  |     a
  |     a
  |     a
  |    +b4
  |    +b5
  |    +b6
  |
  o  2:3086dbafde1ce745abfc8d2d367847280aabae9ddiff --git a/a b/b
  |  copy from a
  ~  copy to b
  

(in this case, "b4", "b5" could be considered introduced by either rev 3, or rev 4.
 and that causes the rev number difference)

  $ hg annotate -nlf b --config fastannotate.modes=
  0 a:1: a
  1 a:2: a
  1 a:3: a
  3 b:4: b4
  4 b:5: c
  3 b:5: b5

  $ hg annotate -nlf b
  0 a:1: a
  1 a:2: a
  1 a:3: a
  4 b:4: b4
  4 b:5: c
  4 b:6: b5

  $ hg up -C 1
  0 files updated, 0 files merged, 1 files removed, 0 files unresolved
  $ hg cp a b
  $ cat <<EOF > b
  > a
  > z
  > a
  > EOF
  $ hg ci -mc -d '3 0'
  created new head
  $ hg merge
  merging b
  0 files updated, 1 files merged, 0 files removed, 0 files unresolved
  (branch merge, don't forget to commit)
  $ cat <<EOF >> b
  > b4
  > c
  > b5
  > EOF
  $ echo d >> b
  $ hg ci -mmerge2 -d '4 0'

annotate after rename merge

  $ hg annotate -nf b
  0 a: a
  6 b: z
  1 a: a
  3 b: b4
  4 b: c
  3 b: b5
  7 b: d

annotate after rename merge with -l
(fastannotate differs from annotate)

  $ hg log -Gp -T '{rev}:{node}' -r '0+1+6+7'
  @    7:6284bb6c38fef984a929862a53bbc71ce9eafa81diff --git a/b b/b
  |\   --- a/b
  | :  +++ b/b
  | :  @@ -1,3 +1,7 @@
  | :   a
  | :   z
  | :   a
  | :  +b4
  | :  +c
  | :  +b5
  | :  +d
  | :
  o :  6:b80e3e32f75a6a67cd4ac85496a11511e9112816diff --git a/a b/b
  :/   copy from a
  :    copy to b
  :    --- a/a
  :    +++ b/b
  :    @@ -1,3 +1,3 @@
  :    -a (?)
  :     a
  :    +z
  :     a
  :    -a (?)
  :
  o  1:762f04898e6684ff713415f7b8a8d53d33f96c92diff --git a/a b/a
  |  --- a/a
  |  +++ b/a
  |  @@ -1,1 +1,3 @@
  |   a
  |  +a
  |  +a
  |
  o  0:8435f90966e442695d2ded29fdade2bac5ad8065diff --git a/a b/a
     new file mode 100644
     --- /dev/null
     +++ b/a
     @@ -0,0 +1,1 @@
     +a
  

(note on question marks:
 the upstream bdiff change (96f2f50d923f+3633403888ae+8c0c75aa3ff4+5c4e2636c1a9
 +38ed54888617) alters the output so deletion is not always at the end of the
 output. for example:
 | a | b | old | new | # old: e1d6aa0e4c3a, new: 8836f13e3c5b
 |-------------------|
 | a | a |  a  | -a  |
 | a | z | +z  |  a  |
 | a | a |  a  | +z  |
 |   |   | -a  |  a  |
 |-------------------|
 | a | a |     a     |
 | a | a |     a     |
 | a |   |    -a     |
 this leads to more question marks below)

(rev 1 adds two "a"s and rev 6 deletes one "a".
 the "a" that rev 6 deletes could be either the first or the second "a" of those two "a"s added by rev 1.
 and that causes the line number difference)

  $ hg annotate -nlf b --config fastannotate.modes=
  0 a:1: a
  6 b:2: z
  1 a:3: a
  3 b:4: b4
  4 b:5: c
  3 b:5: b5
  7 b:7: d

  $ hg annotate -nlf b
  0 a:1: a (?)
  1 a:2: a (?)
  6 b:2: z
  1 a:2: a (?)
  1 a:3: a (?)
  3 b:4: b4
  4 b:5: c
  3 b:5: b5
  7 b:7: d

Issue2807: alignment of line numbers with -l
(fastannotate differs from annotate, same reason as above)

  $ echo more >> b
  $ hg ci -mmore -d '5 0'
  $ echo more >> b
  $ hg ci -mmore -d '6 0'
  $ echo more >> b
  $ hg ci -mmore -d '7 0'
  $ hg annotate -nlf b
   0 a: 1: a (?)
   1 a: 2: a (?)
   6 b: 2: z
   1 a: 2: a (?)
   1 a: 3: a (?)
   3 b: 4: b4
   4 b: 5: c
   3 b: 5: b5
   7 b: 7: d
   8 b: 8: more
   9 b: 9: more
  10 b:10: more

linkrev vs rev

  $ hg annotate -r tip -n a
  0: a
  1: a
  1: a

linkrev vs rev with -l

  $ hg annotate -r tip -nl a
  0:1: a
  1:2: a
  1:3: a

Issue589: "undelete" sequence leads to crash

annotate was crashing when trying to --follow something

like A -> B -> A

generate ABA rename configuration

  $ echo foo > foo
  $ hg add foo
  $ hg ci -m addfoo
  $ hg rename foo bar
  $ hg ci -m renamefoo
  $ hg rename bar foo
  $ hg ci -m renamebar

annotate after ABA with follow

  $ hg annotate --follow foo
  foo: foo

missing file

  $ hg ann nosuchfile
  abort: nosuchfile: no such file in rev e9e6b4fa872f
  [255]

annotate file without '\n' on last line

  $ printf "" > c
  $ hg ci -A -m test -u nobody -d '1 0'
  adding c
  $ hg annotate c
  $ printf "a\nb" > c
  $ hg ci -m test
  $ hg annotate c
  [0-9]+: a (re)
  [0-9]+: b (re)

Issue3841: check annotation of the file of which filelog includes
merging between the revision and its ancestor

to reproduce the situation with recent Mercurial, this script uses (1)
"hg debugsetparents" to merge without ancestor check by "hg merge",
and (2) the extension to allow filelog merging between the revision
and its ancestor by overriding "repo._filecommit".

  $ cat > ../legacyrepo.py <<EOF
  > from mercurial import error, node
  > def reposetup(ui, repo):
  >     class legacyrepo(repo.__class__):
  >         def _filecommit(self, fctx, manifest1, manifest2,
  >                         linkrev, tr, changelist):
  >             fname = fctx.path()
  >             text = fctx.data()
  >             flog = self.file(fname)
  >             fparent1 = manifest1.get(fname, node.nullid)
  >             fparent2 = manifest2.get(fname, node.nullid)
  >             meta = {}
  >             copy = fctx.renamed()
  >             if copy and copy[0] != fname:
  >                 raise error.Abort('copying is not supported')
  >             if fparent2 != node.nullid:
  >                 changelist.append(fname)
  >                 return flog.add(text, meta, tr, linkrev,
  >                                 fparent1, fparent2)
  >             raise error.Abort('only merging is supported')
  >     repo.__class__ = legacyrepo
  > EOF

  $ cat > baz <<EOF
  > 1
  > 2
  > 3
  > 4
  > 5
  > EOF
  $ hg add baz
  $ hg commit -m "baz:0"

  $ cat > baz <<EOF
  > 1 baz:1
  > 2
  > 3
  > 4
  > 5
  > EOF
  $ hg commit -m "baz:1"

  $ cat > baz <<EOF
  > 1 baz:1
  > 2 baz:2
  > 3
  > 4
  > 5
  > EOF
  $ hg debugsetparents 17 17
  $ hg --config extensions.legacyrepo=../legacyrepo.py  commit -m "baz:2"
  $ hg debugindexdot baz
  digraph G {
  	-1 -> 0
  	0 -> 1
  	1 -> 2
  	1 -> 2
  }
  $ hg annotate baz
  17: 1 baz:1
  18: 2 baz:2
  16: 3
  16: 4
  16: 5

  $ cat > baz <<EOF
  > 1 baz:1
  > 2 baz:2
  > 3 baz:3
  > 4
  > 5
  > EOF
  $ hg commit -m "baz:3"

  $ cat > baz <<EOF
  > 1 baz:1
  > 2 baz:2
  > 3 baz:3
  > 4 baz:4
  > 5
  > EOF
  $ hg debugsetparents 19 18
  $ hg --config extensions.legacyrepo=../legacyrepo.py  commit -m "baz:4"
  $ hg debugindexdot baz
  digraph G {
  	-1 -> 0
  	0 -> 1
  	1 -> 2
  	1 -> 2
  	2 -> 3
  	3 -> 4
  	2 -> 4
  }
  $ hg annotate baz
  17: 1 baz:1
  18: 2 baz:2
  19: 3 baz:3
  20: 4 baz:4
  16: 5

annotate clean file

  $ hg annotate -ncr "wdir()" foo
  11 472b18db256d : foo

annotate modified file

  $ echo foofoo >> foo
  $ hg annotate -r "wdir()" foo
  11 : foo
  20+: foofoo

  $ hg annotate -cr "wdir()" foo
  472b18db256d : foo
  b6bedd5477e7+: foofoo

  $ hg annotate -ncr "wdir()" foo
  11 472b18db256d : foo
  20 b6bedd5477e7+: foofoo

  $ hg annotate --debug -ncr "wdir()" foo
  11 472b18db256d1e8282064eab4bfdaf48cbfe83cd : foo
  20 b6bedd5477e797f25e568a6402d4697f3f895a72+: foofoo

  $ hg annotate -udr "wdir()" foo
  test Thu Jan 01 00:00:00 1970 +0000: foo
  test [A-Za-z0-9:+ ]+: foofoo (re)

  $ hg annotate -ncr "wdir()" -Tjson foo
  [
   {
    "lines": [{"line": "foo\n", "node": "472b18db256d1e8282064eab4bfdaf48cbfe83cd", "rev": 11}, {"line": "foofoo\n", "node": "ffffffffffffffffffffffffffffffffffffffff", "rev": 2147483647}],
    "path": "foo"
   }
  ]

annotate added file

  $ echo bar > bar
  $ hg add bar
  $ hg annotate -ncr "wdir()" bar
  20 b6bedd5477e7+: bar

annotate renamed file

  $ hg rename foo renamefoo2
  $ hg annotate -ncr "wdir()" renamefoo2
  11 472b18db256d : foo
  20 b6bedd5477e7+: foofoo

annotate missing file

  $ rm baz
  $ hg annotate -ncr "wdir()" baz
  abort: $TESTTMP/repo/baz: $ENOENT$ (windows !)
  abort: $ENOENT$: $TESTTMP/repo/baz (no-windows !)
  [255]

annotate removed file

  $ hg rm baz
  $ hg annotate -ncr "wdir()" baz
  abort: $TESTTMP/repo/baz: $ENOENT$ (windows !)
  abort: $ENOENT$: $TESTTMP/repo/baz (no-windows !)
  [255]

Test annotate with whitespace options

  $ cd ..
  $ hg init repo-ws
  $ cd repo-ws
  $ cat > a <<EOF
  > aa
  > 
  > b b
  > EOF
  $ hg ci -Am "adda"
  adding a
  $ sed 's/EOL$//g' > a <<EOF
  > a  a
  > 
  >  EOL
  > b  b
  > EOF
  $ hg ci -m "changea"

Annotate with no option

  $ hg annotate a
  1: a  a
  0: 
  1:  
  1: b  b

Annotate with --ignore-space-change

  $ hg annotate --ignore-space-change a
  1: a  a
  1: 
  0:  
  0: b  b

Annotate with --ignore-all-space

  $ hg annotate --ignore-all-space a
  0: a  a
  0: 
  1:  
  0: b  b

Annotate with --ignore-blank-lines (similar to no options case)

  $ hg annotate --ignore-blank-lines a
  1: a  a
  0: 
  1:  
  1: b  b

  $ cd ..

Annotate with linkrev pointing to another branch
------------------------------------------------

create history with a filerev whose linkrev points to another branch

  $ hg init branchedlinkrev
  $ cd branchedlinkrev
  $ echo A > a
  $ hg commit -Am 'contentA'
  adding a
  $ echo B >> a
  $ hg commit -m 'contentB'
  $ hg up --rev 'desc(contentA)'
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ echo unrelated > unrelated
  $ hg commit -Am 'unrelated'
  adding unrelated
  created new head
  $ hg graft -r 'desc(contentB)'
  grafting 1:fd27c222e3e6 "contentB"
  $ echo C >> a
  $ hg commit -m 'contentC'
  $ echo W >> a
  $ hg log -G
  @  changeset:   4:072f1e8df249
  |  tag:         tip
  |  user:        test
  |  date:        Thu Jan 01 00:00:00 1970 +0000
  |  summary:     contentC
  |
  o  changeset:   3:ff38df03cc4b
  |  user:        test
  |  date:        Thu Jan 01 00:00:00 1970 +0000
  |  summary:     contentB
  |
  o  changeset:   2:62aaf3f6fc06
  |  parent:      0:f0932f74827e
  |  user:        test
  |  date:        Thu Jan 01 00:00:00 1970 +0000
  |  summary:     unrelated
  |
  | o  changeset:   1:fd27c222e3e6
  |/   user:        test
  |    date:        Thu Jan 01 00:00:00 1970 +0000
  |    summary:     contentB
  |
  o  changeset:   0:f0932f74827e
     user:        test
     date:        Thu Jan 01 00:00:00 1970 +0000
     summary:     contentA
  

Annotate should list ancestor of starting revision only

  $ hg annotate a
  0: A
  3: B
  4: C

  $ hg annotate a -r 'wdir()'
  0 : A
  3 : B
  4 : C
  4+: W

Even when the starting revision is the linkrev-shadowed one:

  $ hg annotate a -r 3
  0: A
  3: B

  $ cd ..

Issue5360: Deleted chunk in p1 of a merge changeset

  $ hg init repo-5360
  $ cd repo-5360
  $ echo 1 > a
  $ hg commit -A a -m 1
  $ echo 2 >> a
  $ hg commit -m 2
  $ echo a > a
  $ hg commit -m a
  $ hg update '.^' -q
  $ echo 3 >> a
  $ hg commit -m 3 -q
  $ hg merge 2 -q
  $ cat > a << EOF
  > b
  > 1
  > 2
  > 3
  > a
  > EOF
  $ hg resolve --mark -q
  $ hg commit -m m
  $ hg annotate a
  4: b
  0: 1
  1: 2
  3: 3
  2: a

  $ cd ..