view tests/test-mq-qimport.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

  $ cat > writelines.py <<EOF
  > import sys
  > path = sys.argv[1]
  > args = sys.argv[2:]
  > assert (len(args) % 2) == 0
  > 
  > f = open(path, 'wb')
  > for i in range(len(args)//2):
  >    count, s = args[2*i:2*i+2]
  >    count = int(count)
  >    s = s.decode('string_escape')
  >    f.write(s*count)
  > f.close()
  > 
  > EOF
  > cat <<EOF >> $HGRCPATH
  > [extensions]
  > mq =
  > [diff]
  > git = 1
  > EOF
  $ hg init repo
  $ cd repo

qimport without file or revision

  $ hg qimport
  abort: no files or revisions specified
  [255]

qimport non-existing-file

  $ hg qimport non-existing-file
  abort: unable to read file non-existing-file
  [255]

qimport null revision

  $ hg qimport -r null
  abort: revision -1 is not mutable
  (see 'hg help phases' for details)
  [255]
  $ hg qseries

import email

  $ hg qimport --push -n email - <<EOF
  > From: Username in email <[e-mail]>
  > Subject: [PATCH] Message in email
  > Date: Fri, 02 Jan 1970 00:00:00 +0000
  > 
  > Text before patch.
  > 
  > # HG changeset patch
  > # User Username in patch <[e-mail]>
  > # Date 0 0
  > # Node ID 1a706973a7d84cb549823634a821d9bdf21c6220
  > # Parent  0000000000000000000000000000000000000000
  > First line of commit message.
  > 
  > More text in commit message.
  > --- confuse the diff detection
  > 
  > diff --git a/x b/x
  > new file mode 100644
  > --- /dev/null
  > +++ b/x
  > @@ -0,0 +1,1 @@
  > +new file
  > Text after patch.
  > 
  > EOF
  adding email to series file
  applying email
  now at: email

hg tip -v

  $ hg tip -v
  changeset:   0:1a706973a7d8
  tag:         email
  tag:         qbase
  tag:         qtip
  tag:         tip
  user:        Username in patch <[e-mail]>
  date:        Thu Jan 01 00:00:00 1970 +0000
  files:       x
  description:
  First line of commit message.
  
  More text in commit message.
  
  
  $ hg qpop
  popping email
  patch queue now empty
  $ hg qdelete email

import URL

  $ echo foo >> foo
  $ hg add foo
  $ hg diff > url.diff
  $ hg revert --no-backup foo
  $ rm foo

Under unix: file:///foobar/blah
Under windows: file:///c:/foobar/blah

  $ patchurl=`pwd | tr '\\\\' /`/url.diff
  $ expr "$patchurl" : "\/" > /dev/null || patchurl="/$patchurl"
  $ hg qimport file://"$patchurl"
  adding url.diff to series file
  $ rm url.diff
  $ hg qun
  url.diff

import patch that already exists

  $ echo foo2 >> foo
  $ hg add foo
  $ hg diff > ../url.diff
  $ hg revert --no-backup foo
  $ rm foo
  $ hg qimport ../url.diff
  abort: patch "url.diff" already exists
  [255]
  $ hg qpush
  applying url.diff
  now at: url.diff
  $ cat foo
  foo
  $ hg qpop
  popping url.diff
  patch queue now empty

qimport -f

  $ hg qimport -f ../url.diff
  adding url.diff to series file
  $ hg qpush
  applying url.diff
  now at: url.diff
  $ cat foo
  foo2
  $ hg qpop
  popping url.diff
  patch queue now empty

build diff with CRLF

  $ "$PYTHON" ../writelines.py b 5 'a\n' 5 'a\r\n'
  $ hg ci -Am addb
  adding b
  $ "$PYTHON" ../writelines.py b 2 'a\n' 10 'b\n' 2 'a\r\n'
  $ hg diff > b.diff
  $ hg up -C
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved

qimport CRLF diff

  $ hg qimport b.diff
  adding b.diff to series file
  $ hg qpush
  applying b.diff
  now at: b.diff

try to import --push

  $ cat > appendfoo.diff <<EOF
  > append foo
  > 
  > diff -r 07f494440405 -r 261500830e46 baz
  > --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
  > +++ b/baz	Thu Jan 01 00:00:00 1970 +0000
  > @@ -0,0 +1,1 @@
  > +foo
  > EOF

  $ cat > appendbar.diff <<EOF
  > append bar
  > 
  > diff -r 07f494440405 -r 261500830e46 baz
  > --- a/baz	Thu Jan 01 00:00:00 1970 +0000
  > +++ b/baz	Thu Jan 01 00:00:00 1970 +0000
  > @@ -1,1 +1,2 @@
  >  foo
  > +bar
  > EOF

  $ hg qimport --push appendfoo.diff appendbar.diff
  adding appendfoo.diff to series file
  adding appendbar.diff to series file
  applying appendfoo.diff
  applying appendbar.diff
  now at: appendbar.diff
  $ hg qfin -a
  patch b.diff finalized without changeset message
  $ touch .hg/patches/append_foo
  $ hg qimport -r 'p1(.)::'
  $ hg qapplied
  append_foo__1
  append_bar
  $ hg qfin -a
  $ rm .hg/patches/append_foo
  $ hg qimport -r 'p1(.)::' -P
  $ hg qpop -a
  popping append_bar
  popping append_foo
  patch queue now empty
  $ hg qdel append_foo
  $ hg qdel -k append_bar

qimport -e

  $ hg qimport -e append_bar
  adding append_bar to series file
  $ hg qdel -k append_bar

qimport -e --name newname oldexisitingpatch

  $ hg qimport -e --name this-name-is-better append_bar
  renaming append_bar to this-name-is-better
  adding this-name-is-better to series file
  $ hg qser
  this-name-is-better
  url.diff

qimport -e --name without --force

  $ cp .hg/patches/this-name-is-better .hg/patches/3.diff
  $ hg qimport -e --name this-name-is-better 3.diff
  abort: patch "this-name-is-better" already exists
  [255]
  $ hg qser
  this-name-is-better
  url.diff

qimport -e --name with --force

  $ hg qimport --force -e --name this-name-is-better 3.diff
  renaming 3.diff to this-name-is-better
  adding this-name-is-better to series file
  $ hg qser
  this-name-is-better
  url.diff

import patch of bad filename

  $ touch '../ bad.diff'
  $ hg qimport '../ bad.diff'
  abort: patch name cannot begin or end with whitespace
  [255]
  $ touch '.hg/patches/ bad.diff'
  $ hg qimport -e ' bad.diff'
  abort: patch name cannot begin or end with whitespace
  [255]

qimport with bad name, should abort before reading file

  $ hg qimport non-existent-file --name .hg
  abort: patch name cannot begin with ".hg"
  [255]
  $ hg qimport non-existent-file --name ' foo'
  abort: patch name cannot begin or end with whitespace
  [255]
  $ hg qimport non-existent-file --name 'foo '
  abort: patch name cannot begin or end with whitespace
  [255]

qimport http:// patch with leading slashes in url

set up hgweb

  $ cd ..
  $ hg init served
  $ cd served
  $ echo a > a
  $ hg ci -Am patch
  adding a
  $ hg serve -p $HGPORT -d --pid-file=hg.pid -A access.log -E errors.log
  $ cat hg.pid >> $DAEMON_PIDS

  $ cd ../repo
  $ hg qimport http://localhost:$HGPORT/raw-rev/0///
  adding 0 to series file

check qimport phase:

  $ hg -q qpush
  now at: 0
  $ hg phase qparent
  1: draft
  $ hg qimport -r qparent
  $ hg phase qbase
  1: draft
  $ hg qfinish qbase
  $ echo '[mq]' >> $HGRCPATH
  $ echo 'secret=true' >> $HGRCPATH
  $ hg qimport -r qparent
  $ hg phase qbase
  1: secret

  $ cd ..

  $ killdaemons.py

check patch name generation for non-alpha-numeric summary line

  $ cd repo

  $ hg qpop -a -q
  patch queue now empty
  $ hg qseries -v
  0 U imported_patch_b_diff
  1 U 0
  2 U this-name-is-better
  3 U url.diff

  $ echo bb >> b
  $ hg commit -m '==++--=='

  $ hg qimport -r tip
  $ hg qseries -v
  0 A 1.diff
  1 U imported_patch_b_diff
  2 U 0
  3 U this-name-is-better
  4 U url.diff

check reserved patch names

  $ hg qpop -qa
  patch queue now empty
  $ echo >> b
  $ hg commit -m 'status'
  $ echo >> b
  $ hg commit -m '.'
  $ echo >> b
  $ hg commit -m 'taken'
  $ mkdir .hg/patches/taken
  $ touch .hg/patches/taken__1
  $ hg qimport -r -3::
  $ hg qap
  1.diff__1
  2.diff
  taken__2

check very long patch name

  $ hg qpop -qa
  patch queue now empty
  $ echo >> b
  $ hg commit -m 'abcdefghi pqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghi pqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghi pqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghi pqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghi pqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghi pqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
  $ hg qimport -r .
  $ hg qap
  abcdefghi_pqrstuvwxyzabcdefghijklmnopqrstuvwxyzabcdefghi_pqrstuvwxyzabcdefg