Implement ReadableMetrics for Entries table

[dramaticlly]

close #7364

This enable query readable metrics on entry metadata table, similar to #5376 by @szehon-ho

Given table schema like below

table {
  1: id: optional int
  2: data: optional string
}

we can query its per column min/max for each manifest entry

spark
.sql("select snapshot_id, readable_metrics.id.lower_bound, readable_metrics.data.upper_bound")
.show(false)

+-------------------+-----------+-----------+
|snapshot_id        |lower_bound|upper_bound|
+-------------------+-----------+-----------+
|8728349059432389037|1          |1          |
+-------------------+-----------+-----------+

[szehon-ho]

Great job, just some comments about code - reuse

[szehon-ho]

Merged, thanks @dramaticlly for the change!

[szehon-ho]

Hi, @dramaticlly as I didnt want to block the change, I had some follow up that I made in #7613 including what we discussed in #7539 (comment).

I think it makes the code between BaseFilesTable and BaseEntriesTable even more similar and with more re-use, if you wanted to take a look

[bwliu62]

Hello, I want to get upper/lower bounds from the manifest file like:

val icebergTalbe = catalog.loadTable(tableIdentifier)

val filter = Expressions.greaterThan("id", 10)

val scan = icebergTable.newScan().filter(filter)

scan.includeColumnStats().planFiles().forEach { task =>

    val upperBounds = task.file().upperBounds()

    val lowerBounds = task.file().lowerBounds()

    log.info(s"upperBounds: $upperBounds")

    log.info(s"upperBounds: $upperBounds")

}

The log will be byteBuffer, I see @szehon-ho pr for a readable metrics, I am wondering do we have a java api we can use, or I have to use DDL?