[SPARK-35581][SPARK-36567][SQL][DOCS][FOLLOWUP] Update the SQL migration guide about foldable special datetime values

### What changes were proposed in this pull request?
In the PR, I propose to update an existing item in the SQL migration guide, and mention that Spark 3.2 supports foldable special datetime values as well.
<img width="1292" alt="Screenshot 2021-08-25 at 23 29 51" src="https://user-images.githubusercontent.com/1580697/130860184-27f0ba56-6c2d-4a5a-91a8-195f2f8aa5da.png">

### Why are the changes needed?
To inform users about actual Spark SQL behavior introduced by #33816

### Does this PR introduce _any_ user-facing change?
No.

### How was this patch tested?
By generating docs, and checking results manually.

Closes #33840 from MaxGekk/special-datetime-cast-migr-guide.

Authored-by: Max Gekk <[email protected]>
Signed-off-by: Wenchen Fan <[email protected]>

Author: MaxGekk

docs/sql-migration-guide.md

@@ -117,7 +117,7 @@ license: |
 
   - In Spark 3.2, `CREATE TABLE AS SELECT` with non-empty `LOCATION` will throw `AnalysisException`. To restore the behavior before Spark 3.2, you can set `spark.sql.legacy.allowNonEmptyLocationInCTAS` to `true`.
 
-  - In Spark 3.2, special datetime values such as `epoch`, `today`, `yesterday`, `tomorrow`, and `now` are supported in typed literals only, for instance, `select timestamp'now'`. In Spark 3.1 and 3.0, such special values are supported in any casts of strings to dates/timestamps. To keep these special values as dates/timestamps in Spark 3.1 and 3.0, you should replace them manually, e.g. `if (c in ('now', 'today'), current_date(), cast(c as date))`.
+  - In Spark 3.2, special datetime values such as `epoch`, `today`, `yesterday`, `tomorrow`, and `now` are supported in typed literals or in cast of foldable strings only, for instance, `select timestamp'now'` or `select cast('today' as date)`. In Spark 3.1 and 3.0, such special values are supported in any casts of strings to dates/timestamps. To keep these special values as dates/timestamps in Spark 3.1 and 3.0, you should replace them manually, e.g. `if (c in ('now', 'today'), current_date(), cast(c as date))`.
 
   - In Spark 3.2, `FloatType` is mapped to `FLOAT` in MySQL. Prior to this, it used to be mapped to `REAL`, which is by default a synonym to `DOUBLE PRECISION` in MySQL.