fix stackoverflow from spread operation

[BrianHung]

Context

Ran into a stack overflow error when trying to build a sheet with 65536 (2^16) rows.

How did you test your changes?

By incrementing adding vertices to a set instead of creating arrays and then flattening them at the end, we save on memory.

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)
• [x] 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:

Checklist:

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

[sequba]

@BrianHung, thank you for your contribution. I'll look into it soon.

Did you test how much memory does this way save? How many rows are you able to add after this change?

[BrianHung]

thank you for your contribution. I'll look into it soon.

Thanks! Not sure if it's the most efficient way so change it if you find something better.

Did you test how much memory does this way save? How many rows are you able to add after this change?

I modified the hyperformula benchmark suite to run with 2^16 rows since I wanted to see how fast it would take to initialize a spreadsheet of that size. This was on a 32 gb macbook. Don't know how much memory it saved but it allowed me to run the benchmark without crashing into OOM.

[sequba]

@BrianHung could you share the code that produced the OOM error?

[BrianHung]

@sequba

I modified cruds-benchmark.ts to use 2^16 rows. Unsure if this OOM was system specific since I only tested it only my macbook with 32gb of ram.

  batch(result,
    () => benchmarkCruds('Sheet A:  change value, add/remove row/column', sheetAGenerator(66536), (engine: HyperFormula) => {
      engine.setCellContents(adr('A1'), '123')
      engine.addRows(0, [5000, 1])
      engine.removeRows(0, [8000, 1])
      engine.addColumns(0, [0, 1])
      engine.removeColumns(0, [0, 1])
    }, [
      {address: 'E7000', value: -1.17344394901827e+23},
    ], {
      engineConfig: {
        maxRows: 131072
      }
    }),

    () => benchmarkCruds('Sheet B: change value, add/remove row/column', sheetBGenerator(66536), (engine: HyperFormula) => {
      engine.setCellContents(adr('A1'), '123')
      engine.addRows(0, [2000, 1])
      engine.removeRows(0, [3000, 1])
      engine.addColumns(0, [0, 1])
      engine.removeColumns(0, [0, 1])
    }, [
      {address: 'E50', value: 1347},
      {address: 'E2002', value: 2001122},
    ], {
      engineConfig: {
        maxRows: 131072
      }
    }),

[sequba]

Thank you @BrianHung. I'm merging this PR

[codecov[bot]]

Codecov Report

Merging #1321 (64f540e) into feature/issue-1332 (a008c1d) will decrease coverage by 0.03%. The diff coverage is 100.00%.

:exclamation: Current head 64f540e differs from pull request most recent head 60533c9. Consider uploading reports for the commit 60533c9 to get more accurate results

@@                  Coverage Diff                   @@
##           feature/issue-1332    #1321      +/-   ##
======================================================
- Coverage               97.27%   97.25%   -0.03%     
======================================================
  Files                     169      169              
  Lines                   14422    14409      -13     
  Branches                 3018     3092      +74     
======================================================
- Hits                    14029    14013      -16     
+ Misses                    393      391       -2     
- Partials                    0        5       +5     

Files	Coverage Δ
src/DependencyGraph/DependencyGraph.ts	97.94% <100.00%> (+<0.01%)	:arrow_up:

... and 9 files with indirect coverage changes

[AMBudnik]

Hi @BrianHung

We just closed the linked issue https://github.com/handsontable/hyperformula/issues/1332 and released it under HyperFormula v2.6.1