##// END OF EJS Templates
transaction: issue "new obsmarkers" message at the end of the transaction...
transaction: issue "new obsmarkers" message at the end of the transaction Instead of making bundle2 code responsible for this, it seems better to have it handled and the transaction level. First, it means the message will be more consistently printed. Second it means we won't spam the message over and over if the data arrive in multiple piece. Third, we are planning to move other similar message at the same level (for the same reason) so having them all at the same location will help us to control the order they are displayed.

File last commit:

r42819:db963eda default
r43164:38392d5b default
Show More
test-rebase-legacy.t
89 lines | 2.1 KiB | text/troff | Tads3Lexer
/ tests / test-rebase-legacy.t
Taapas Agrawal
continue: added support for rebase...
r42832 #testcases continuecommand continueflag
Jun Wu
rebase: change internal format to support destination map...
r34006 Test rebase --continue with rebasestate written by legacy client
$ cat >> $HGRCPATH <<EOF
> [extensions]
> rebase=
> drawdag=$TESTDIR/drawdag.py
> EOF
Taapas Agrawal
continue: added support for rebase...
r42832 #if continueflag
$ cat >> $HGRCPATH <<EOF
> [alias]
> continue = rebase --continue
> EOF
#endif
Jun Wu
rebase: change internal format to support destination map...
r34006 $ hg init
$ hg debugdrawdag <<'EOF'
> D H
> | |
> C G
> | |
> B F
> | |
> Z A E
> \|/
> R
> EOF
rebasestate generated by a legacy client running "hg rebase -r B+D+E+G+H -d Z"
$ touch .hg/last-message.txt
$ cat > .hg/rebasestate <<EOF
> 0000000000000000000000000000000000000000
> f424eb6a8c01c4a0c0fba9f863f79b3eb5b4b69f
> 0000000000000000000000000000000000000000
> 0
> 0
> 0
>
> 21a6c45028857f500f56ae84fbf40689c429305b:-2
> de008c61a447fcfd93f808ef527d933a84048ce7:0000000000000000000000000000000000000000
> c1e6b162678d07d0b204e5c8267d51b4e03b633c:0000000000000000000000000000000000000000
> aeba276fcb7df8e10153a07ee728d5540693f5aa:-3
> bd5548558fcf354d37613005737a143871bf3723:-3
> d2fa1c02b2401b0e32867f26cce50818a4bd796a:0000000000000000000000000000000000000000
> 6f7a236de6852570cd54649ab62b1012bb78abc8:0000000000000000000000000000000000000000
> 6582e6951a9c48c236f746f186378e36f59f4928:0000000000000000000000000000000000000000
> EOF
Taapas Agrawal
continue: added support for rebase...
r42832 #if continuecommand
$ hg continue --dry-run
rebase in progress, will be resumed
#endif
$ hg continue
Jun Wu
rebase: change internal format to support destination map...
r34006 rebasing 4:c1e6b162678d "B" (B)
rebasing 8:6f7a236de685 "D" (D)
rebasing 2:de008c61a447 "E" (E)
rebasing 7:d2fa1c02b240 "G" (G)
rebasing 9:6582e6951a9c "H" (H tip)
warning: orphaned descendants detected, not stripping c1e6b162678d, de008c61a447
Matt Harbison
tests: remove (glob) annotations that were only for '\' matches...
r35394 saved backup bundle to $TESTTMP/.hg/strip-backup/6f7a236de685-9880a3dc-rebase.hg
Jun Wu
rebase: change internal format to support destination map...
r34006
$ hg log -G -T '{rev}:{node|short} {desc}\n'
o 11:721b8da0a708 H
|
o 10:9d65695ec3c2 G
|
o 9:21c8397a5d68 E
|
| o 8:fc52970345e8 D
| |
| o 7:eac96551b107 B
|/
| o 6:bd5548558fcf C
| |
| | o 5:aeba276fcb7d F
| | |
| o | 4:c1e6b162678d B
| | |
o | | 3:f424eb6a8c01 Z
| | |
+---o 2:de008c61a447 E
| |
| o 1:21a6c4502885 A
|/
o 0:b41ce7760717 R