view tests/test-evolve-issue6246.t @ 6783:ccc8352c7133

topic: add more branchcache entries to test-topic-flow-publish-flag.t

It seems these new entries started appearing after the recent-ish refactoring
of branchcache in core. Now the server and the client have the same branch
cache data on disk, which is probably for the best.

The actual output with changeset hashes is:
branch2-base:c0155e8ec55a9b46d415d5b13a319d0c80e26e9d o double//slash
branch2-immutable:c0155e8ec55a9b46d415d5b13a319d0c80e26e9d o double//slash
branch2-served:256a2cb64fd8ada4ff18d57606d212f094cb50fb o double//slash

c0155e8ec55a is public, so it's correct to be in "base" and "immutable" repo
views. 256a2cb64fd8 is draft (without topic), so it's expected to be in the
"served" view.

Failure to open evoext_stablerange_v2.sqlite shouldn't affect operations (issue6246)
https://bz.mercurial-scm.org/show_bug.cgi?id=6246

  $ . $TESTDIR/testlib/common.sh

  $ cat << EOF >> $HGRCPATH
  > [extensions]
  > evolve =
  > EOF

  $ hg init issue6246
  $ cd issue6246
  $ hg debugbuilddag '.+6'

making a cache file that sqlite cannot open shouldn't break stablerange cache

  $ touch .hg/cache/evoext_stablerange_v2.sqlite
  $ chmod 0000 .hg/cache/evoext_stablerange_v2.sqlite

  $ hg debug::evo-ext-stable-range --method default --verify --subranges --rev 1 --debug
  stable-range cache: unable to load, regenerating
  66f7d451a68b-0 (1, 2, 2) [complete] - 1ea73414a91b-0 (0, 1, 1), 66f7d451a68b-1 (1, 2, 1)
  1ea73414a91b-0 (0, 1, 1) [leaf] - 
  66f7d451a68b-1 (1, 2, 1) [leaf] - 

  $ hg debugobshashrange --rev tip --debug
  stable-range cache: unable to load, regenerating
           rev         node        index         size        depth      obshash
  obshashrange cache: unable to load, regenerating
             6 f69452c5b1af            0            7            7 000000000000

  $ cd ..