[C++][Parquet] Improve parquet reading performance for String/Binary type based on Buffer operations instead of BinaryArrayBuilder

[asfimport]

Improve parquet reading performance for String/Binary type based on Buffer operations instead of BinaryArrayBuilder.

 

Currently, parquet String/Binary type reading performance is not good enough. For fixedwidth types, the Reader has adopted Buffer operations  instead of ArrayBuilder interfaces.

 

So it‘s better to use Buffer operations in String/Binary type reading.

 

Reporter: xiangxiang Shen / @zhixingheyi-tian
Assignee: xiangxiang Shen / @zhixingheyi-tian

PRs and other links:

• GitHub Pull Request #14353

_{Note: This issue was originally created as ARROW-17735. Please see the migration documentation for further details.}

[asfimport]

xiangxiang Shen / @zhixingheyi-tian:
Hi @pitrou , @jorisvandenbossche 

Please give a review, thanks.

[asfimport]

Antoine Pitrou / @pitrou:
@zhixingheyi-tian Which implementation is this for? C++?

[asfimport]

xiangxiang Shen / @zhixingheyi-tian:
Hi @pitrou , yes, C++.

Related code:

   cpp/src/parquet/column_reader.cc

   cpp/src/parquet/encoding.cc

   cpp/src/parquet/arrow/reader_internal.cc

  

Also, I can submit one pull request

 

[asfimport]

Antoine Pitrou / @pitrou:

Also, I can submit one pull request

Great. Hopefully that will improve performance then :-)

[thisisnic]

This issue has been marked as stale because it has had no activity in the past 2 years. Please remove the stale label or comment below, or this issue will be closed in 14 days.