Fix handling of common sub expressions and encoded input

[bikramSingh91]

This patch fixes the following:

• Handling common sub expressions running over disjoint set of active rows
• Tracking of rows that have been pre-computed by the common expressions
• Handling evaluation of encoded input

Test Plan: Added unit tests

[netlify[bot]]

✅ Deploy Preview for meta-velox canceled.

Name	Link
🔨 Latest commit	4a6a577c71827942d898d997be78fff9124930d6
🔍 Latest deploy log	https://app.netlify.com/sites/meta-velox/deploys/631a5122d65450000816d7f3

[facebook-github-bot]

@bikramSingh91 has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

[facebook-github-bot]

@bikramSingh91 has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

[bikramSingh91]

@mbasmanova request you to please take another look when you have some free cycles

[bikramSingh91]

Just to confirm, do you mean test coverage with respect to adding a case where final selection is not a superset of sharedSubExpressionRows?

[mbasmanova]

Just to confirm, do you mean test coverage with respect to adding a case where final selection is not a superset of sharedSubExpressionRows?

I'm thinking that we might be missing test cases that would cover the following scenario:

I wonder if it is possible for sharedSubexprRows_ to have rows that are not present in context->finalSelection(). I think this may happen if some rows are needed for evaluating one expression, but not another. To cover this case we may need to set finalSelection to a union of finalSelection and sharedSubexprRows_.

[bikramSingh91]

Just to confirm, do you mean test coverage with respect to adding a case where final selection is not a superset of sharedSubExpressionRows?

I'm thinking that we might be missing test cases that would cover the following scenario:

I wonder if it is possible for sharedSubexprRows_ to have rows that are not present in context->finalSelection(). I think this may happen if some rows are needed for evaluating one expression, but not another. To cover this case we may need to set finalSelection to a union of finalSelection and sharedSubexprRows_.

Got it, thanks for confirming

[facebook-github-bot]

@bikramSingh91 has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

[facebook-github-bot]

@bikramSingh91 has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.