Keep csv files open

[lhupfeldt]

The intent of is to reduce os calls and disk IO.
This code goes on top of my other PR #1504, so submitting as draft.

[codecov]

Codecov Report

Merging #1522 into master will increase coverage by 0.35%.
The diff coverage is 93.20%.

@@            Coverage Diff             @@
##           master    #1522      +/-   ##
==========================================
+ Coverage   81.58%   81.94%   +0.35%     
==========================================
  Files          28       28              
  Lines        2466     2509      +43     
  Branches      382      381       -1     
==========================================
+ Hits         2012     2056      +44     
+ Misses        359      358       -1     
  Partials       95       95              

Impacted Files	Coverage Δ
locust/main.py	20.44% <33.33%> (+0.44%)	⬆️
locust/stats.py	90.02% <98.82%> (+0.98%)	⬆️
locust/env.py	96.72% <100.00%> (ø)
locust/web.py	90.15% <100.00%> (+0.15%)	⬆️
locust/argument_parser.py	77.19% <0.00%> (+0.20%)	⬆️
locust/runners.py	83.23% <0.00%> (+0.38%)	⬆️
locust/user/task.py	96.75% <0.00%> (+0.54%)	⬆️

---

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 12bfb26...e0e35dc. Read the comment docs.

[cyberw]

Does this also split stats into stats and stats_csv? It might make sense, but as I dont really know the web interface code very well, I can't really judge if that is a good idea.

[lhupfeldt]

Does this also split stats into stats and stats_csv? It might make sense, but as I dont really know the web interface code very well, I can't really judge if that is a good idea.

Yes, it splits out the csv generation. I think it gives better readability, but I can merge it back into stats if you prefer.

[cyberw]

Yes, I would prefer keeping them in the same file (but we can change it later maybe)

[lhupfeldt]

Let me rebase on master first.

…

On Tue, 18 Aug 2020, 10:55 Lars Holmberg, ***@***.***> wrote: Is this ok to merge now? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#1522 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAMCZSN7JHC3Y5O5SZAMQPDSBI6ZPANCNFSM4QAX2LHA> .

[cyberw]

Good to go? oh right, still waiting to un-split stats/stats_csv (preferably by rewriting the history so that git-blame is not distorted)

[lhupfeldt]

Good to go? oh right, still waiting to un-split stats/stats_csv (preferably by rewriting the history so that git-blame is not distorted)

I think I will be to blame for the stats-csv stuff regardless how it is un-split, since all lines will be indented under the class.
So the easiest will be to simply copy stats_csv content back into stats.
Any objection to that?

[lhupfeldt]

Created a new single commit and force pushed. Should keep the blame changes to a minimum, although not sure there is much difference.

[cyberw]

LGTM. Thanks!