Add tests for saved search creation and loading for query enhancement

[angle943]

Description

Adds cypress tests for query enhanced env. Creates a saved search and loads it.

Issues Resolved

closes #8949
closes #8948

Changelog

• test: add tests for saving search and loading it

Check List

• All tests pass
  • yarn test:jest
  • yarn test:jest_integration
• New functionality includes testing.
• New functionality has been documented.
• Update CHANGELOG.md
• Commits are signed per the DCO using --signoff

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 60.90%. Comparing base (87011f7) to head (36c75d9).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #9112      +/-   ##
==========================================
- Coverage   60.90%   60.90%   -0.01%     
==========================================
  Files        3808     3808              
  Lines       91196    91196              
  Branches    14400    14400              
==========================================
- Hits        55547    55543       -4     
- Misses      32096    32099       +3     
- Partials     3553     3554       +1     

Flag	Coverage Δ
Linux_1	29.00% <ø> (ø)
Linux_2	56.38% <ø> (ø)
Linux_3	38.01% <ø> (-0.01%)	⬇️
Linux_4	29.01% <ø> (ø)
Windows_1	29.01% <ø> (ø)
Windows_2	56.34% <ø> (ø)
Windows_3	38.01% <ø> (ø)
Windows_4	29.01% <ø> (-0.01%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[angle943]

here i am randomizing to avoid conflict in parallel tests for the workspace name and datasource

[ananzh]

nit: since we capsuled a lot of test items in, some item like sampleTableData is not quite clear to me. shall we add a comment at the beginning of indexPatternTestConfigurations.

[angle943]

yeah can do

[ananzh]

Should we make mapLanguageToApiResponseString and mapDatasetToType to constants for type safe and single source of truth? For example, instead of running a func mapDatasetToType we just do

const DatasetTypes = {
  INDEX_PATTERN: {
    name: 'INDEX_PATTERN',
    value: INDEX_PATTERN_WITH_TIME,
    supportedLanguages: [QueryLanguages.DQL, QueryLanguages.LUCENE, QueryLanguages.SQL, QueryLanguages.PPL],
  },
  INDEXES: {
    name: 'INDEXES',
    value: INDEX_WITH_TIME_1,
    supportedLanguages: [QueryLanguages.SQL, QueryLanguages.PPL],
  },
};

and

const QueryLanguages = {
  DQL: {
    name: 'DQL',
    apiName: 'kuery',  // The mapping is part of the definition
    supports: {
      filters: true,
      histogram: true,
      // ...
    }
  },
  // ...
};

and if (language === QueryLanguages.DQL.name) instead of if (language === 'DQL'). So adding a new query language only requires updating QueryLanguages and relevant templates and
Adding a new dataset type is only add to DatasetTypes.

Then we can have
TestCaseGenerator: Handles test case creation
TestExecutor: Manages test execution flow

export const generateTestCase = (dataset, language) => {
  const baseConfig = {
    name: `${dataset.name}-${language.name}`,
    dataset: dataset.value,
    language: language.name,
    queryString: createQueryString(language.name, dataset.value),
    ...language.supports,
  };

  return {
    ...baseConfig,
    hitCount: getExpectedHitCount(dataset.name, language.name),
    sampleTableData: getExpectedTableData(dataset.name, language.name),
    saveName: `${language.name.toLowerCase()}-${dataset.name.toLowerCase()}-01`,
  };
};

export const generateAllTestCases = () => {
  return Object.values(DatasetTypes).flatMap(dataset => 
    dataset.supportedLanguages.map(language => 
      generateTestCase(dataset, language)
    )
  );

then

// Main test suite
generateAllTestCases().forEach(testCase => {
      it(`should handle saved search for ${testCase.name}`, async () => {
        // Set up the search
        await navigateToDiscoverPage(cy, testSetup.workspaceName);
        await setDataset(cy, {
          ...testCase,
          datasourceName: testSetup.datasourceName
        });
        await setSearchConfigurations(cy, testCase);

        // Verify the search state
        verifyDiscoverPageState(cy, testCase);
        
        // Save and verify the search
        await saveSearch(cy, testCase.saveName);
        verifySavedSearchInAssetsPage(cy, {
          ...testCase,
          workspaceName: testSetup.workspaceName
        });
      });
    });

I think we can move some functions here to util. I am thinking how to make the tests are readable and type safe and no big duplication.

[angle943]

For:

const DatasetTypes = {
  INDEX_PATTERN: {
    name: 'INDEX_PATTERN',
    value: INDEX_PATTERN_WITH_TIME,
    supportedLanguages: [QueryLanguages.DQL, QueryLanguages.LUCENE, QueryLanguages.SQL, QueryLanguages.PPL],
  },
  INDEXES: {
    name: 'INDEXES',
    value: INDEX_WITH_TIME_1,
    supportedLanguages: [QueryLanguages.SQL, QueryLanguages.PPL],
  },
};

Im afraid of restricting the value to specific INDEX_PATTERN_WQITH_TIME and INDEX_WITH_TIME_1. For example, what if another dev wanted to use INDEX_WITHOUT_TIME_1?

But I completely agree with abstracting things out that can be abstracted, and I will do so

[angle943]

I implemented something very similar to what you have described. Could you take a look?

[ananzh]

I don't quite get it. For DQL, sort should happen after waitForLoad. Do you mind adding a video for this?

[angle943]

I can't always replicate it, but sometimes:

• The "Selected fields" are not applied. The headers for the selected fields are there, but all the data is dumped into the first column
• The Sort sometimes doesn't happen, even after loading for a while

I'll grab videos of these if i can replicate it

[angle943]

I got a video of the latter case, but its too big to upload here :D . I'll share it with you in private