fix shifting of moving* datapoints by one step into the future

[zivillian]

This is the quick fix from the discussion in #2632

fixes #2632

[deniszh]

Hi @zivillian ,

I'm afraid that's quick and incorrect fix, though... As you can see many function tests are broken now, because of extra 'None' inserted.

[zivillian]

If I'm correct, there is no extra None introduced, but it's missing and still present in the expectedData lists. Is my assumption correct, that the original and moved series should contain the same number of datapoints? If so, I would update the test data to reflect this.

[deniszh]

Ah, mea culpa, you're rightm @zivillian E.g. https://github.com/graphite-project/graphite-web/blob/97f4d85750b3bff2c478c2d3b98cf4368ca84756/webapp/tests/test_functions.py#L4687 Then tests need to be correcred, indeed

[zivillian]

I'll take a look into this, and also add an assertion that the number of datapoints in both series is equal

[codecov-io]

Codecov Report

Merging #2682 (1805baa) into master (528d197) will not change coverage. The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master    #2682   +/-   ##
=======================================
  Coverage   80.11%   80.11%           
=======================================
  Files          88       88           
  Lines        9403     9403           
  Branches     2002     2002           
=======================================
  Hits         7533     7533           
  Misses       1582     1582           
  Partials      288      288           

Impacted Files	Coverage Δ
webapp/graphite/render/functions.py	95.37% <100.00%> (ø)

---

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 528d197...1805baa. Read the comment docs.

[deniszh]

I'm still bit hesitating to merge this TBH. Opinions? /cc @ploxiln

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[zivillian]

@ploxiln @deniszh any thoughts?

[deniszh]

Thanks, @ploxiln for reviewing this! That's indeed quite tricky. Merging. And thanks to @zivillian for issue and PR too, of course!

[deniszh]

💔 Some backports could not be created

Status	Branch	Result
❌	1.1.x	Cherrypick failed because the selected commit was a merge commit. Please try again by specifying the parent with the mainline argument:

backport --mainline

or:

backport --mainline

Or refer to the git documentation for more information: https://git-scm.com/docs/git-cherry-pick#Documentation/git-cherry-pick.txt---mainlineparent-number|

To backport manually run: node scripts/backport --pr 2682. For more info read the Backport documentation

[deniszh]

💔 Some backports could not be created

Status	Branch	Result
❌	1.1.x	Cherrypick failed because the selected commit was a merge commit. Please try again by specifying the parent with the mainline argument:

backport --mainline

or:

backport --mainline

Or refer to the git documentation for more information: https://git-scm.com/docs/git-cherry-pick#Documentation/git-cherry-pick.txt---mainlineparent-number|

To backport manually run: node scripts/backport --pr 2682. For more info read the Backport documentation

[deniszh]

💔 Some backports could not be created

Status	Branch	Result
❌	1.1.x	Cherrypick failed because the selected commit was a merge commit. Please try again by specifying the parent with the mainline argument:

backport --mainline

or:

backport --mainline

Or refer to the git documentation for more information: https://git-scm.com/docs/git-cherry-pick#Documentation/git-cherry-pick.txt---mainlineparent-number|

To backport manually run: node scripts/backport --pr 2682. For more info read the Backport documentation