fix(bufferCount): will behave as expected when startBufferEvery is …

[benlesh]

…greater than bufferSize

• fixed issue where internal buffers store was keeping an additional buffer for no good reason
• improved logic and performance around updating internal buffers list
• adds a test to ensure proper behavior

fixes #2062

[benlesh]

Worth noting, performance improves here as well:

This branch:

                                         |                     RxJS 4.1.0 |                RxJS 5.0.0-rc.1 |          factor |      % improved
---------------------------------------------------------------------------------------------------------------------------------------------------
                 buffercount - immediate |                24,789 (±0.76%) |               455,692 (±0.57%) |          18.38x |        1,738.3%
                             buffercount |                19,820 (±1.55%) |               242,544 (±1.45%) |          12.24x |        1,123.7%

master:

                                         |                     RxJS 4.1.0 |                RxJS 5.0.0-rc.1 |          factor |      % improved
---------------------------------------------------------------------------------------------------------------------------------------------------
                 buffercount - immediate |                25,716 (±0.73%) |               403,000 (±0.87%) |          15.67x |        1,467.1%
                             buffercount |                19,664 (±1.47%) |               226,951 (±1.35%) |          11.54x |        1,054.1%

[jayphelps]

@Blesh #2062 is actually the opposite, startBufferEvery is less than bufferSize. (the PR description says this fix is for when startBufferEvery is greater than bufferSize)

[benlesh]

haha... @jayphelps oops.

[benlesh]

Also, I checked windowCount, and it does not have the same issue.

[coveralls]

Coverage decreased (-0.003%) to 97.214% when pulling 1787b6e on blesh:bufferCount-fix into 260d335 on ReactiveX:master.

[coveralls]

Changes Unknown when pulling eb11bb8 on blesh:bufferCount-fix into * on ReactiveX:master*.

[coveralls]

Coverage decreased (-0.003%) to 97.214% when pulling dde5d8d on blesh:bufferCount-fix into 0271fab on ReactiveX:master.

[benlesh]

FYI: I think the coverage zero tolerance is getting us here with an unfortunate mathematical jump.

[lock]

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.