Improve efficiency of the default date/time parsing methods (16% speed-up)

[sequba]

Test case

A spreadsheet with 500k rows and 10 columns filled with string data. Total of 5M data cells, no formulas, no date/time values.

Script:

const hf = HyperFormula.buildFromArray([], {
  licenseKey: 'gpl-v3',
  maxRows: 500100,
  useStats: true,
  chooseAddressMappingPolicy: new AlwaysDense(),
})

const data = Array(500000).fill(0).map(() => Array(10).fill(0).map(() => 'A500000'))

var ty1 = (new Date()).getTime()
hf.setSheetContent(0, data)
var ty2 = (new Date()).getTime()
console.log(ty2 - ty1)
console.log(hf.getStats())

Ideas for improvement

• Function getTopSortedWithSccSubgraphFrom (total time 42.7%), which is the iterative implementation of the Tarjan algorithm that performs the topological sorting of the dependency graph and finds cycles (SCCs) in the graph. In this test case, the dependency graph is trivial; it contains only isolated nodes without any edges. It seems that it can be done more efficiently.
• Function parseDateTimeFromConfigFormats (total time 21.2%), which tries to parse all string data as date/time values. This test case contains no date/time values, so there might be some way of saving time by determining it quickly and avoiding running the heavy parsing operations.

This PR focuses on optimizing date/time parsing functions

1. Date format string and time format string (provided in the engine's configuration) need to be pre-processed before using them to parse date/time values from the input strings. I extracted this pre-processing code and applied memorization to it so that it is being run only once per configured format instead of once for every data cell in the spreadsheet.
2. I introduced a quick regexp check to detect early the strings that certainly couldn't be parsed to date/time values and avoid running expensive parsing operations.

Results

Total time:
Before: 32264ms
After: 26391ms

Function parseDateTimeFromConfigFormats:
Before: 21.2%
After: 6.3%

Profiler:
Before:

After:

Profiler: Chrome Dev Tools

This result corresponds to ~16% speed-up of HyperFormula for this use-case.

How did you test your changes?

• correctness verified by the full suite of unit tests
• verified the improvement on the test case described above
• run regular performance benchmarks (results included in this PR as a comment below)

Types of changes

• Breaking change (a fix or a feature because of which an existing functionality doesn't work as expected anymore)
• New feature or improvement (a non-breaking change that adds functionality)
• Bug fix (a non-breaking change that fixes an issue)
• Additional language file, or a change to an existing language file (translations)
• Change to the documentation

Related issues:

1. Slow graph building with 500k of rows without formulas #876

Checklist:

• I have reviewed the guidelines about Contributing to HyperFormula and I confirm that my code follows the code style of this project.
• I have signed the Contributor License Agreement.
• My change is compliant with the OpenDocument standard.
• My change is compatible with Microsoft Excel.
• My change is compatible with Google Sheets.
• I described my changes in the CHANGELOG.md file.
• My changes require a documentation update.
• My changes require a migration guide.

[github-actions]

Performance comparison of head (4224b3a) vs base (276f731)

                                     testName |   base |  head |  change
------------------------------------------------------------------------
                                      Sheet A | 1003.6 | 981.2 |  -2.23%
                                      Sheet B |  358.7 | 380.3 |  +6.02%
                                      Sheet T |  397.8 |   308 | -22.57%
                                Column ranges |  851.4 | 885.9 |  +4.05%
Sheet A:  change value, add/remove row/column |     51 |    56 |  +9.80%
 Sheet B: change value, add/remove row/column |    424 |   423 |  -0.24%
                   Column ranges - add column |    330 |   586 | +77.58%
                Column ranges - without batch |    866 |   912 |  +5.31%
                        Column ranges - batch |    240 |   453 | +88.75%

[budnix]

Looks good 👌

[codecov]

Codecov Report

Attention: Patch coverage is 98.80952% with 1 line in your changes missing coverage. Please review.

Project coverage is 97.23%. Comparing base (276f731) to head (4224b3a).
Report is 81 commits behind head on develop.

Files with missing lines	Patch %	Lines
src/DateTimeDefault.ts	98.70%	1 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #1298      +/-   ##
===========================================
+ Coverage    97.20%   97.23%   +0.03%     
===========================================
  Files          167      167              
  Lines        14299    14304       +5     
  Branches      3064     3065       +1     
===========================================
+ Hits         13899    13909      +10     
  Misses         395      395              
+ Partials         5        0       -5     

Files with missing lines	Coverage Δ
src/DateTimeHelper.ts	96.19% <100.00%> (+0.54%)	⬆️
src/format/format.ts	99.30% <100.00%> (ø)
src/DateTimeDefault.ts	97.43% <98.70%> (+3.68%)	⬆️

... and 3 files with indirect coverage changes