phases: avoid N² behavior in `advanceboundary`
We allowed duplicated entries in the deque, which each entry could potentially
insert all its ancestors. So advancing boundary for the full repository would
mean each revision would walk all its ancestors, resulting in O(N²) iteration.
For repository of any decent size, N² is quickly insane.
We introduce a simple set to avoid this and get back to reasonable performance.
$ hg init repo
$ cd repo
$ touch a.html b.html c.py d.py
$ cat > frontend.sparse << EOF
> [include]
> *.html
> EOF
$ hg -q commit -A -m initial
$ echo 1 > a.html
$ echo 1 > c.py
$ hg commit -m 'commit 1'
Enable sparse profile
$ hg debugrequires
dotencode
dirstate-v2 (dirstate-v2 !)
fncache
generaldelta
persistent-nodemap (rust !)
revlog-compression-zstd (zstd !)
revlogv1
share-safe
sparserevlog
store
testonly-simplestore (reposimplestore !)
$ hg debugsparse --config extensions.sparse= --enable-profile frontend.sparse
$ ls -A
.hg
a.html
b.html
Requirement for sparse added when sparse is enabled
$ hg debugrequires --config extensions.sparse=
dotencode
dirstate-v2 (dirstate-v2 !)
exp-sparse
fncache
generaldelta
persistent-nodemap (rust !)
revlog-compression-zstd (zstd !)
revlogv1
share-safe
sparserevlog
store
testonly-simplestore (reposimplestore !)
Client without sparse enabled reacts properly
$ hg files
abort: repository is using sparse feature but sparse is not enabled; enable the "sparse" extensions to access
[255]
Requirement for sparse is removed when sparse is disabled
$ hg debugsparse --reset --config extensions.sparse=
$ hg debugrequires
dotencode
dirstate-v2 (dirstate-v2 !)
fncache
generaldelta
persistent-nodemap (rust !)
revlog-compression-zstd (zstd !)
revlogv1
share-safe
sparserevlog
store
testonly-simplestore (reposimplestore !)
And client without sparse can access
$ hg files
a.html
b.html
c.py
d.py
frontend.sparse