Codecov Report

Merging #72 into master will decrease coverage by 0.46%.
The diff coverage is 33.33%.

@@            Coverage Diff            @@
##           master     #72      +/-   ##
=========================================
- Coverage   55.86%   55.4%   -0.47%     
=========================================
  Files          11      11              
  Lines         571     583      +12     
=========================================
+ Hits          319     323       +4     
- Misses        252     260       +8

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 66bd967...98b49e4. Read the comment docs.

@aisk Please review when you get a chance!

Thanks @letmutx ! I'll review this as soon as posible!

@aisk Just a heads up - please do not merge this yet. I added more tests and a particular test is failing. I am debugging. Will try and fix over the weekend.

@aisk Fixed now. Should be good to review.

The failing test is because of a debug_assert! I added in parse_header_only_response. In some cases, a body can follow after a store command. (for ex: when using cas, add, replace commands, if the key doesn't exist, an error string follows after the response header)

There are no tests for these cases. We should handle them in a different issue. I will remove the debug_assert and add an issue.

Hi sorry late for the review, I'm a little busy these days.

I saw you added a FromMemcacheValueExt trait which implements FromMemcacheValue, and the origin FromMemcacheValue trait is not directly used now. Can we merge the two trait into one and keep the old name? Or do you have some other consideration?

The Ext trait is to keep backwards compatibility. Because the definition of from_memcache_value changes in the new trait, users who have implemented the trait for their custom types would've had to implement the new trait.

But because there is a blanket impl of FromMemcacheValueExtfor all types which implement FromMemcacheValue, they no longer need to do that.

Yes you are right, sorry I don't considered the backwards compatibility issue. We shoul keep it, thanks @letmutx

Hi @letmutx I approved this PR, except I want to remove the CasID type as mentioned above. I think maybe we should add a type or trait some day if we have a chance to refactor the API.

And can you rebase the master branch since there are some changes. Seems there was no conflict but I think a rebase with CI test is needed.

Really thanks @letmutx 👍

Thanks for the review.

@aisk rebased and pushed the changes. good to merge now, i think.

@letmutx merged, thanks!