rspec
/
rspec-mocks
/
1033
/
8
98%
master: 63%
master: 63%
|
Ran
|
Files
34
|
Run time
–
|
Badge
Embed ▾
README BADGES
|
coverage: 99.435%. First build
push
travis-ci
myronmarstonRefactor to improve perf. - Don't get the original method handle for a test double (it doesn't have an original method as its not a partial mock). - Leverage InstanceMethodStasher; since it has to stash the method anyway, it can store it so we don't have to waste effort getting it. - Delay getting the original_method until the last possible moment: right before we redefine it with a new definition. As a side benefit, this change makes it so that we no longer "leak" odd obfuscated aliased methods onto partial mock objects. This is a good thing. I ran benchmarks of this commits vs the prior (unoptimized fix) vs what we originally had in master using this shell command: time (for i in {1..20}; do bin/rspec -fp --order default; done) Here are the totals: - this commit: 39.070 seconds - prior unoptimized fix (c864e0a53): 39.936 - master (80d80d919): 37.783
1759 of 1769 relevant lines covered (99.43%)
115.49 hits per line
