refactor(pdk): serialize log msg with string.buffer

[chronolaw]

Summary

Use string.buffer to optimize string operation.

Here I simply replace table.insert and table.concat, but not sure the serializers[n]'s effect, so just keep them.

Checklist

• The Pull Request has tests
• A changelog file has been created under changelog/unreleased/kong or skip-changelog label added on PR if changelog is unnecessary. README.md
• There is a user-facing docs PR against https://github.com/Kong/docs.konghq.com - PUT DOCS PR HERE

Full changelog

• [Implement ...]

Issue reference

Fix #[issue number]

[chobits]

I have reviewed the source code and confirmed that the new refactoring logic aligns with the original.

However I'm uncertain if this PR can improve the perfomance of string serialization. Its beneficial to run some simple tests on random string sets and compare the time consumption. I noticed that the new refactoring logic uses more efficient string.buf library, but it also has introduced a higher number of function calls of :put() in serialization function [1~n]

[zhongweiy]

Do we have any performance change data after this optimisation?

And the Jira ticket is missing in the PR message.

[chronolaw]

Do we have any performance change data after this optimisation?

And the Jira ticket is missing in the PR message.

Not all PR need a Jira ticket, especially for some small changes.

[kikito]

@chronolaw does this refactor provide a perf improvement?

[chronolaw]

@chronolaw does this refactor provide a perf improvement?

I have no proper data to prove the improvement of performance, but I am sure it will be helpful for perf.
It seems that it is a little hard to test so small change of code, do we need to do this?

[zhongweiy]

I checked there are already a bunch of similar changes done before:

https://github.com/Kong/kong-ee/pulls?q=is%3Apr+is%3Aclosed+author%3Achronolaw+string

Also there is similar discussion like #11304 (comment), which provides performance improvement like 5x time better when replacing table with string buffer in similar usage.

Based on this, I feel it is OK to do similar replacement. @chronolaw Sorry for keeping the review process long and I'm a little bit conservative before checking the previous similar changes.