[SPARK-57447][SQL] Escape the index name in (H2|MySQL|Postgres)Dialect.indexExists

[dongjoon-hyun]

What changes were proposed in this pull request?

indexExists in H2Dialect, MySQLDialect, and PostgresDialect embeds the
index name as a SQL string literal in its lookup query. This PR wraps the index
name with the existing JdbcDialect.escapeSql helper so that single quotes are
properly escaped.

Note that the scope of this PR is only index names. For table names or TABLE_SCHEMA, we are going to handle independently.

Why are the changes needed?

If an index name contains a single quote, the lookup query becomes malformed
(e.g. ... INDEX_NAME = 'a'b'). JdbcUtils.checkIfIndexExists swallows the
resulting exception and returns false, so an existing index whose name
contains a single quote is incorrectly reported as not existing.

Does this PR introduce any user-facing change?

No.

How was this patch tested?

Pass the CIs.

Was this patch authored or co-authored using generative AI tooling?

Generated-by: Claude Code

[dongjoon-hyun]

cc @huaxingao

[dongjoon-hyun]

All tests passed.

[dongjoon-hyun]

Could you review this PR when you have some time, please, @viirya ?

[viirya]

The fix itself is correct — escapeSql doubles single quotes, the three call sites all interpolate indexName inside a '...' literal, and I confirmed JdbcUtils.checkIfIndexExists swallows the exception and returns false, so the false-negative described in the PR is real. The Mockito test is nice: it captures the executed SQL and asserts the escaped form, so all three dialects are covered in plain CI without a live DB.

My one concern is scope vs. the bug class. The same false-negative pattern still exists in these methods for the other interpolated literals:

• H2Dialect.indexExists still has TABLE_SCHEMA = '${tableIdent.namespace().last}' and TABLE_NAME = '${tableIdent.name()}' unescaped.
• PostgresDialect.indexExists still has tablename = '${tableIdent.name()}' unescaped.

A table named o'brien reproduces the identical malformed-query -> swallowed-exception -> false "index does not exist" behavior. The same raw-literal pattern also appears in listIndexes (H2Dialect and PostgresDialect build ... = '${tableIdent.name()}' the same way).

Could we either escape the table/schema names in these queries too, or note in the PR description that they're intentionally out of scope for a follow-up? Leaning toward fixing them here since it's just a few more escapeSql calls and avoids a near-identical bug report later. Not blocking — the change as-is is safe and correct.

[huaxingao]

LGTM

[dongjoon-hyun]

Thank you, @viirya and @huaxingao .

To @viirya , I revise the PR description by mentioning the scope is only index names and the table names and TABLE_SCHEMA will be handled independently.

Could we either escape the table/schema names in these queries too, or note in the PR description that they're intentionally out of scope for a follow-up?

[dongjoon-hyun]

Merged to master/4.x for now because we have Apache Spark 4.2.0 RC3 vote currently.

If RC3 fails, I will consider to cherry-pick this to branch-4.2 and older.