view tests/test-rebase-dest.t @ 40326:fed697fa1734

sqlitestore: file storage backend using SQLite

This commit provides an extension which uses SQLite to store file
data (as opposed to revlogs).

As the inline documentation describes, there are still several
aspects to the extension that are incomplete. But it's a start.
The extension does support basic clone, checkout, and commit
workflows, which makes it suitable for simple use cases.

One notable missing feature is support for "bundlerepos." This is
probably responsible for the most test failures when the extension
is activated as part of the test suite.

All revision data is stored in SQLite. Data is stored as zstd
compressed chunks (default if zstd is available), zlib compressed
chunks (default if zstd is not available), or raw chunks (if
configured or if a compressed delta is not smaller than the raw
delta). This makes things very similar to revlogs.

Unlike revlogs, the extension doesn't yet enforce a limit on delta
chain length. This is an obvious limitation and should be addressed.
This is somewhat mitigated by the use of zstd, which is much faster
than zlib to decompress.

There is a dedicated table for storing deltas. Deltas are stored
by the SHA-1 hash of their uncompressed content. The "fileindex" table
has columns that reference the delta for each revision and the base
delta that delta should be applied against. A recursive SQL query
is used to resolve the delta chain along with the delta data.

By storing deltas by hash, we are able to de-duplicate delta storage!
With revlogs, the same deltas in different revlogs would result in
duplicate storage of that delta. In this scheme, inserting the
duplicate delta is a no-op and delta chains simply reference the
existing delta.

When initially implementing this extension, I did not have
content-indexed deltas and deltas could be duplicated across files
(just like revlogs). When I implemented content-indexed deltas, the
size of the SQLite database for a full clone of mozilla-unified
dropped:

before: 2,554,261,504 bytes
after:  2,488,754,176 bytes

Surprisingly, this is still larger than the bytes size of revlog
files:

revlog files: 2,104,861,230 bytes
du -b:        2,254,381,614

I would have expected storage to be smaller since we're not limiting
delta chain length and since we're using zstd instead of zlib. I
suspect the SQLite indexes and per-column overhead account for the
bulk of the differences. (Keep in mind that revlog uses a 64-byte
packed struct for revision index data and deltas are stored without
padding. Aside from the 12 unused bytes in the 32 byte node field,
revlogs are pretty efficient.) Another source of overhead is file
name storage. With revlogs, file names are stored in the filesystem.
But with SQLite, we need to store file names in the database. This is
roughly equivalent to the size of the fncache file, which for the
mozilla-unified repository is ~34MB.

Since the SQLite database isn't append-only and since delta chains
can reference any delta, this opens some interesting possibilities.
For example, we could store deltas in reverse, such that fulltexts
are stored for newer revisions and deltas are applied to reconstruct
older revisions. This is likely a more optimal storage strategy for
version control, as new data tends to be more frequently accessed
than old data. We would obviously need wire protocol support for
transferring revision data from newest to oldest. And we would
probably need some kind of mechanism for "re-encoding" stores. But
it should be doable.

This extension is very much experimental quality. There are a handful
of features that don't work. It probably isn't suitable for day-to-day
use. But it could be used in limited cases (e.g. read-only checkouts
like in CI). And it is also a good proving ground for alternate
storage backends. As we continue to define interfaces for all things
storage, it will be useful to have a viable alternate storage backend
to see how things shake out in practice.

test-storage.py passes on Python 2 and introduces no new test failures on
Python 3. Having the storage-level unit tests has proved to be insanely
useful when developing this extension. Those tests caught numerous bugs
during development and I'm convinced this style of testing is the way
forward for ensuring alternate storage backends work as intended. Of
course, test coverage isn't close to what it needs to be. But it is
a start. And what coverage we have gives me confidence that basic store
functionality is implemented properly.

Differential Revision: https://phab.mercurial-scm.org/D4928

Require a destination
  $ cat >> $HGRCPATH <<EOF
  > [extensions]
  > rebase =
  > [commands]
  > rebase.requiredest = True
  > EOF
  $ hg init repo
  $ cd repo
  $ echo a >> a
  $ hg commit -qAm aa
  $ echo b >> b
  $ hg commit -qAm bb
  $ hg up ".^"
  0 files updated, 0 files merged, 1 files removed, 0 files unresolved
  $ echo c >> c
  $ hg commit -qAm cc
  $ hg rebase
  abort: you must specify a destination
  (use: hg rebase -d REV)
  [255]
  $ hg rebase -d 1
  rebasing 2:5db65b93a12b "cc" (tip)
  saved backup bundle to $TESTTMP/repo/.hg/strip-backup/5db65b93a12b-4fb789ec-rebase.hg
  $ hg rebase -d 0 -r . -q
  $ HGPLAIN=1 hg rebase
  rebasing 2:889b0bc6a730 "cc" (tip)
  saved backup bundle to $TESTTMP/repo/.hg/strip-backup/889b0bc6a730-41ec4f81-rebase.hg
  $ hg rebase -d 0 -r . -q
  $ hg --config commands.rebase.requiredest=False rebase
  rebasing 2:279de9495438 "cc" (tip)
  saved backup bundle to $TESTTMP/repo/.hg/strip-backup/279de9495438-ab0a5128-rebase.hg

Requiring dest should not break continue or other rebase options
  $ hg up 1 -q
  $ echo d >> c
  $ hg commit -qAm dc
  $ hg log -G -T '{rev} {desc}'
  @  3 dc
  |
  | o  2 cc
  |/
  o  1 bb
  |
  o  0 aa
  
  $ hg rebase -d 2
  rebasing 3:0537f6b50def "dc" (tip)
  merging c
  warning: conflicts while merging c! (edit, then use 'hg resolve --mark')
  unresolved conflicts (see hg resolve, then hg rebase --continue)
  [1]
  $ echo d > c
  $ hg resolve --mark --all
  (no more unresolved files)
  continue: hg rebase --continue
  $ hg rebase --continue
  rebasing 3:0537f6b50def "dc" (tip)
  saved backup bundle to $TESTTMP/repo/.hg/strip-backup/0537f6b50def-be4c7386-rebase.hg

  $ cd ..

Check rebase.requiredest interaction with pull --rebase
  $ hg clone repo clone
  updating to branch default
  3 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ cd repo
  $ echo e > e
  $ hg commit -qAm ee
  $ cd ..
  $ cd clone
  $ echo f > f
  $ hg commit -qAm ff
  $ hg pull --rebase
  abort: rebase destination required by configuration
  (use hg pull followed by hg rebase -d DEST)
  [255]

Setup rebase with multiple destinations

  $ cd $TESTTMP

  $ cat >> $TESTTMP/maprevset.py <<EOF
  > from __future__ import absolute_import
  > from mercurial import registrar, revset, revsetlang, smartset
  > revsetpredicate = registrar.revsetpredicate()
  > cache = {}
  > @revsetpredicate(b'map')
  > def map(repo, subset, x):
  >     """(set, mapping)"""
  >     setarg, maparg = revsetlang.getargs(x, 2, 2, b'')
  >     rset = revset.getset(repo, smartset.fullreposet(repo), setarg)
  >     mapstr = revsetlang.getstring(maparg, b'')
  >     map = dict(a.split(b':') for a in mapstr.split(b','))
  >     rev = rset.first()
  >     desc = repo[rev].description()
  >     newdesc = map.get(desc)
  >     if newdesc == b'null':
  >         revs = [-1]
  >     else:
  >         query = revsetlang.formatspec(b'desc(%s)', newdesc)
  >         revs = repo.revs(query)
  >     return smartset.baseset(revs)
  > EOF

  $ cat >> $HGRCPATH <<EOF
  > [ui]
  > allowemptycommit=1
  > [extensions]
  > drawdag=$TESTDIR/drawdag.py
  > [phases]
  > publish=False
  > [alias]
  > tglog = log -G --template "{rev}: {node|short} {desc} {instabilities}" -r 'sort(all(), topo)'
  > [extensions]
  > maprevset=$TESTTMP/maprevset.py
  > [experimental]
  > evolution=true
  > EOF

  $ rebasewithdag() {
  >   N=`"$PYTHON" -c "print($N+1)"`
  >   hg init repo$N && cd repo$N
  >   hg debugdrawdag
  >   hg rebase "$@" > _rebasetmp
  >   r=$?
  >   grep -v 'saved backup bundle' _rebasetmp
  >   [ $r -eq 0 ] && rm -f .hg/localtags && hg tglog
  >   cd ..
  >   return $r
  > }

Destination resolves to an empty set:

  $ rebasewithdag -s B -d 'SRC - SRC' <<'EOS'
  > C
  > |
  > B
  > |
  > A
  > EOS
  nothing to rebase - empty destination
  [1]

Multiple destinations and --collapse are not compatible:

  $ rebasewithdag -s C+E -d 'SRC^^' --collapse <<'EOS'
  > C F
  > | |
  > B E
  > | |
  > A D
  > EOS
  abort: --collapse does not work with multiple destinations
  [255]

Multiple destinations cannot be used with --base:

  $ rebasewithdag -b B+E -d 'SRC^^' --collapse <<'EOS'
  > B E
  > | |
  > A D
  > EOS
  abort: unknown revision 'SRC'!
  [255]

Rebase to null should work:

  $ rebasewithdag -r A+C+D -d 'null' <<'EOS'
  > C D
  > | |
  > A B
  > EOS
  already rebased 0:426bada5c675 "A" (A)
  already rebased 2:dc0947a82db8 "C" (C)
  rebasing 3:004dc1679908 "D" (D tip)
  o  4: d8d8601abd5e D
  
  o  2: dc0947a82db8 C
  |
  | o  1: fc2b737bb2e5 B
  |
  o  0: 426bada5c675 A
  
Destination resolves to multiple changesets:

  $ rebasewithdag -s B -d 'ALLSRC+SRC' <<'EOS'
  > C
  > |
  > B
  > |
  > Z
  > EOS
  abort: rebase destination for f0a671a46792 is not unique
  [255]

Destination is an ancestor of source:

  $ rebasewithdag -s B -d 'SRC' <<'EOS'
  > C
  > |
  > B
  > |
  > Z
  > EOS
  abort: source and destination form a cycle
  [255]

Switch roots:

  $ rebasewithdag -s 'all() - roots(all())' -d 'roots(all()) - ::SRC' <<'EOS'
  > C  F
  > |  |
  > B  E
  > |  |
  > A  D
  > EOS
  rebasing 2:112478962961 "B" (B)
  rebasing 4:26805aba1e60 "C" (C)
  rebasing 3:cd488e83d208 "E" (E)
  rebasing 5:0069ba24938a "F" (F tip)
  o  9: d150ff263fc8 F
  |
  o  8: 66f30a1a2eab E
  |
  | o  7: 93db94ffae0e C
  | |
  | o  6: d0071c3b0c88 B
  | |
  | o  1: 058c1e1fb10a D
  |
  o  0: 426bada5c675 A
  
Different destinations for merge changesets with a same root:

  $ rebasewithdag -s B -d '((parents(SRC)-B-A)::) - (::ALLSRC)' <<'EOS'
  > C G
  > |\|
  > | F
  > |
  > B E
  > |\|
  > A D
  > EOS
  rebasing 3:a4256619d830 "B" (B)
  rebasing 6:8e139e245220 "C" (C tip)
  o    8: 51e2ce92e06a C
  |\
  | o    7: 2ed0c8546285 B
  | |\
  o | |  5: 8fdb2c1feb20 G
  | | |
  | | o  4: cd488e83d208 E
  | | |
  o | |  2: a6661b868de9 F
   / /
  | o  1: 058c1e1fb10a D
  |
  o  0: 426bada5c675 A
  
Move to a previous parent:

  $ rebasewithdag -s E+F+G -d 'SRC^^' <<'EOS'
  >     H
  >     |
  >   D G
  >   |/
  >   C F
  >   |/
  >   B E  # E will be ignored, since E^^ is empty
  >   |/
  >   A
  > EOS
  rebasing 4:33441538d4aa "F" (F)
  rebasing 6:cf43ad9da869 "G" (G)
  rebasing 7:eef94f3b5f03 "H" (H tip)
  o  10: b3d84c6666cf H
  |
  | o  5: f585351a92f8 D
  |/
  o  3: 26805aba1e60 C
  |
  | o  9: f7c28a1a15e2 G
  |/
  o  1: 112478962961 B
  |
  | o  8: 02aa697facf7 F
  |/
  | o  2: 7fb047a69f22 E
  |/
  o  0: 426bada5c675 A
  
Source overlaps with destination:

  $ rebasewithdag -s 'B+C+D' -d 'map(SRC, "B:C,C:D")' <<'EOS'
  > B C D
  >  \|/
  >   A
  > EOS
  rebasing 2:dc0947a82db8 "C" (C)
  rebasing 1:112478962961 "B" (B)
  o  5: 5fe9935d5222 B
  |
  o  4: 12d20731b9e0 C
  |
  o  3: b18e25de2cf5 D
  |
  o  0: 426bada5c675 A
  
Detect cycles early:

  $ rebasewithdag -r 'all()-Z' -d 'map(SRC, "A:B,B:C,C:D,D:B")' <<'EOS'
  > A B C
  >  \|/
  >   | D
  >   |/
  >   Z
  > EOS
  abort: source and destination form a cycle
  [255]

Detect source is ancestor of dest in runtime:

  $ rebasewithdag -r 'C+B' -d 'map(SRC, "C:B,B:D")' -q <<'EOS'
  >   D
  >   |
  > B C
  >  \|
  >   A
  > EOS
  abort: source is ancestor of destination
  [255]

"Already rebased" fast path still works:

  $ rebasewithdag -r 'all()' -d 'SRC^' <<'EOS'
  >   E F
  >  /| |
  > B C D
  >  \|/
  >   A
  > EOS
  already rebased 1:112478962961 "B" (B)
  already rebased 2:dc0947a82db8 "C" (C)
  already rebased 3:b18e25de2cf5 "D" (D)
  already rebased 4:312782b8f06e "E" (E)
  already rebased 5:ad6717a6a58e "F" (F tip)
  o  5: ad6717a6a58e F
  |
  o  3: b18e25de2cf5 D
  |
  | o    4: 312782b8f06e E
  | |\
  +---o  2: dc0947a82db8 C
  | |
  | o  1: 112478962961 B
  |/
  o  0: 426bada5c675 A
  
Massively rewrite the DAG:

  $ rebasewithdag -r 'all()' -d 'map(SRC, "A:I,I:null,H:A,B:J,J:C,C:H,D:E,F:G,G:K,K:D,E:B")' <<'EOS'
  > D G K
  > | | |
  > C F J
  > | | |
  > B E I
  >  \| |
  >   A H
  > EOS
  rebasing 4:701514e1408d "I" (I)
  rebasing 0:426bada5c675 "A" (A)
  rebasing 1:e7050b6e5048 "H" (H)
  rebasing 5:26805aba1e60 "C" (C)
  rebasing 7:cf89f86b485b "J" (J)
  rebasing 2:112478962961 "B" (B)
  rebasing 3:7fb047a69f22 "E" (E)
  rebasing 8:f585351a92f8 "D" (D)
  rebasing 10:ae41898d7875 "K" (K tip)
  rebasing 9:711f53bbef0b "G" (G)
  rebasing 6:64a8289d2492 "F" (F)
  o  21: 3735afb3713a F
  |
  o  20: 07698142d7a7 G
  |
  o  19: 33aba52e7e72 K
  |
  o  18: 9fdae89dc5a1 D
  |
  o  17: 277dda9a65ee E
  |
  o  16: 9c74fd8657ad B
  |
  o  15: 6527eb0688bb J
  |
  o  14: e94d655b928d C
  |
  o  13: 620d6d349459 H
  |
  o  12: a569a116758f A
  |
  o  11: 2bf1302f5c18 I
  
Resolve instability:

  $ rebasewithdag <<'EOF' -r 'orphan()-obsolete()' -d 'max((successors(max(roots(ALLSRC) & ::SRC)^)-obsolete())::)'
  >      F2
  >      |
  >    J E E2
  >    | |/
  > I2 I | E3
  >   \| |/
  >    H | G
  >    | | |
  >   B2 D F
  >    | |/         # rebase: B -> B2
  >    N C          # amend: E -> E2
  >    | |          # amend: E2 -> E3
  >    M B          # rebase: F -> F2
  >     \|          # amend: I -> I2
  >      A
  > EOF
  6 new orphan changesets
  rebasing 16:5c432343bf59 "J" (J tip)
  rebasing 3:26805aba1e60 "C" (C)
  rebasing 6:f585351a92f8 "D" (D)
  rebasing 10:ffebc37c5d0b "E3" (E3)
  rebasing 13:fb184bcfeee8 "F2" (F2)
  rebasing 11:dc838ab4c0da "G" (G)
  o  22: 174f63d574a8 G
  |
  o  21: c9d9fbe76705 F2
  |
  o  20: 0a03c2ede755 E3
  |
  o  19: 228d9d2541b1 D
  |
  o  18: cd856b400c95 C
  |
  o  17: 9148200c858c J
  |
  o  15: eb74780f5094 I2
  |
  o  12: 78309edd643f H
  |
  o  5: 4b4531bd8e1d B2
  |
  o  4: 337c285c272b N
  |
  o  2: 699bc4b6fa22 M
  |
  o  0: 426bada5c675 A