Log format unification for easier grep

Enhancement

As there is no fields support in the current logger, when meeting fields like segment_id, or table_name, we are now outputting them in different ways in the log. For example, it may be like segment [1], or may be like [id:1]. This makes the grep process hard to catch all occurrences, because we need to grep for different keywords. Additionally, notice that [id:1] is not friendly for grep because it may be conflicted with other "id".

We should arrange these fields in a unified way, so that searching a specific keyword will lead and only lead to related log lines.

The following are tasks so that we can pinpoint logs related with different things:

• Related with a segment: grep by segment_id=XX\> https://github.com/pingcap/tiflash/pull/5728
• Related with segment changes: grep by Split, MergeDelta, Merge, Flush https://github.com/pingcap/tiflash/pull/5728
• Related with a table: grep by table_id=XX\> https://github.com/pingcap/tiflash/pull/6055
• Related with a DMFile: TBD
• Related with a stable layer: TBD
• Related with a delta layer: TBD
• Related with a dm store: TBD
• Related with a region: TBD

Start by reviewing the existing logging changes in pull requests #5728 and #6055, then identify how the unchecked DMFile, stable layer, delta layer, dm store, and region log fields are currently emitted. Done means defining and applying a consistent grep-friendly format for the remaining areas, with the relevant log lines searchable by their specific fields.