fix: no need to expect mem table return Err

[BowenXiao1999]

What's changed and what's your intention?

Ideas from:
#3068 (comment)

Memtable do not return Error, so no need to call .next()?.

Checklist

• I have written necessary docs and comments
• I have added necessary unit tests and integration tests
• All checks passed in ./risedev check (or alias, ./risedev c)

Refer to a related PR or issue link (optional)

[skyzh]

This would cause wrong result. We only peeked memtable without calling .next.

Better to rewrite this part with async_stream to make everything clear.

[BowenXiao1999]

This would cause wrong result. We only peeked memtable without calling .next.

Better to rewrite this part with async_stream to make everything clear.

Actually it should return Error at (state_table.next()). If not(I cant imagine why), it will directly panic. So I guess It will not make the code malfunction.

[codecov]

Codecov Report

Merging #3071 (6732f6b) into main (4ffe791) will decrease coverage by 0.09%.
The diff coverage is n/a.

❗ Current head 6732f6b differs from pull request most recent head ebe0086. Consider uploading reports for the commit ebe0086 to get more accurate results

@@            Coverage Diff             @@
##             main    #3071      +/-   ##
==========================================
- Coverage   73.62%   73.52%   -0.10%     
==========================================
  Files         736      732       -4     
  Lines      101467    99431    -2036     
==========================================
- Hits        74701    73106    -1595     
+ Misses      26766    26325     -441     

Flag	Coverage Δ
rust	73.52% <ø> (-0.10%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
src/storage/src/table/state_table.rs	87.97% <ø> (+1.18%)	⬆️
src/meta/src/hummock/compaction_group/mod.rs	55.55% <0.00%> (-38.89%)	⬇️
src/common/src/catalog/physical_table.rs	0.00% <0.00%> (-30.00%)	⬇️
src/frontend/src/planner/relation.rs	76.00% <0.00%> (-16.81%)	⬇️
src/expr/src/vector_op/bitwise_op.rs	91.93% <0.00%> (-8.07%)	⬇️
src/frontend/src/catalog/source_catalog.rs	60.86% <0.00%> (-7.56%)	⬇️
src/storage/src/hummock/state_store.rs	78.14% <0.00%> (-7.09%)	⬇️
src/expr/src/vector_op/cmp.rs	61.73% <0.00%> (-6.49%)	⬇️
src/expr/src/vector_op/cast.rs	59.88% <0.00%> (-6.15%)	⬇️
src/meta/src/stream/source_manager.rs	14.28% <0.00%> (-5.31%)	⬇️
... and 175 more

📣 Codecov can now indicate which changes are the most critical in Pull Requests. Learn more

[BowenXiao1999]

But indeed. This part logic looks strange

[BugenZhao]

This would cause wrong result. We only peeked memtable without calling .next.

Better to rewrite this part with async_stream to make everything clear.

The match goes to this arm only if we peeked an error from cell_based_table_iter, so there should be no problem.

It may look strange because we cannot take the error out of the peeked reference. 😢