view tests/test-filecache.py.out @ 27784:432242f41d9f

obsolete: make _computeobsoleteset much faster

This patch makes _computeobsoleteset much faster by looping
over the draft and secrets as opposed to looping over the
successors.
This works because "number of draft and secret" is typically
way smaller(<100) than the number of successor in the repo (~90k in
my checkout of core mercurial as of today). And also because
it is very fast to compute "not public()".

I timed the code with the following setup:
"""
from mercurial import hg, ui, obsolete
ui = ui.ui()
repo = hg.repository(ui, "~/hg")
l = repo.obsstore.successors # This caches the result
"""

With about 90k successors.

k=obsolete._computeobsoleteset(repo) before this patch:
10 loops, best of 3: 33.9 ms per loop

k=obsolete._computeobsoleteset(repo) after this patch:
10000 loops, best of 3: 83.3 µs per loop

basic:

* neither file exists
creating
* neither file still exists
* empty file x created
creating
* file x changed size
creating
* nothing changed with either file
* file x changed inode
creating
* empty file y created
creating
* file y changed size
creating
* file y changed inode
creating
* both files changed inode
creating

fakeuncacheable:

* neither file exists
creating
* neither file still exists
creating
* empty file x created
creating
* file x changed size
creating
* nothing changed with either file
creating
* file x changed inode
creating
* empty file y created
creating
* file y changed size
creating
* file y changed inode
creating
* both files changed inode
creating
repository tip rolled back to revision -1 (undo commit)
working directory now based on revision -1
repository tip rolled back to revision -1 (undo commit)
working directory now based on revision -1

setbeforeget:

* neither file exists
string set externally
* file x created
creating
string from function
* string set externally again
string 2 set externally
* file y created
creating
string from function