SNOW-1865904 fix query gen when nested cte node is partitioned

[sfc-gh-aalam]

1. Which Jira issue is this PR addressing? Make sure that there is an accompanying issue to your PR.

   Fixes SNOW-1865904

2. Fill out the following pre-review checklist:

   • I am adding a new automated test(s) to verify correctness of my new code
     • If this test skips Local Testing mode, I'm requesting review from @snowflakedb/local-testing
   • I am adding new logging messages
   • I am adding a new telemetry message
   • I am adding new credentials
   • I am adding a new dependency
   • If this is a new feature/behavior, I'm adding the Local Testing parity changes.
   • I acknowledge that I have ensured my changes to be thread-safe. Follow the link for more information: Thread-safe Developer Guidelines

3. Please describe how your code solves the related issue.

   Update query generation for with query block where cte query is updated if the child plan is re-resolved during optimization process.

[sfc-gh-helmeleegy]

LGTM, just a couple of nits. Thanks!

[sfc-gh-yzou]

can you give some more detailed description about what is the problem if we skip the node update here ? the cte optimizaiton is doing a level to level update, so it can skip the node update if it is already updated, but for large query breakdown, i recall you were doing a dfs, right?

[sfc-gh-aalam]

consider a tree below

     node1
        |
     node2
   /       \
node3      node4

suppose we start by re-resolving node3 -> node2 -> node1. In this process, node2 and node1 are marked in updated_nodes. Now, when we go updating the ancestors of node4, re-resolution of node2 and node1 would be skipped. This is not ideal if node4 update can also trigger a re-update of node2 and node1.

For example, this could be problematic is when node3 and node4 before the update had referenced_cte map. After the first update node3 -> node2 -> node1, node2 will be resolved with an older version of node4. If after a re-resolved node4 there are no referenced_ctes, then node2 will not be updated with this information.

[sfc-gh-yzou]

so it is because you are doing a dfs, if you are doing updated in the order of node3, node4 then node2, last node1 you shouldn't run into problems.

[sfc-gh-yzou]

do we have a bug before?

[sfc-gh-aalam]

no. I just noticed an unnecessary pass so I cleaned it up.