regex: Speed-optimized PCRE implementation.

[penguindark]

What's inside

• optimizations for speed
• less memory used
• avoided every recursion
• stack depth selection with function change_stack_depth() for particular case

What's good

• code with less of 700 rows of code
• good separation between compiler and executor
• stack depth configurable

What's not good

• large use of goto in order to increase the speed

What's next

• more tests vs C/C++ and Python libs in order to compare the speed achieved.

Optimization

• Pre-allocated Machine workspace: All search operations are zero-allocation.
• Raw pointer access: Bypasses bounds checking for string and instruction access.
• Fast ASCII Path: Characters < 128 bypass heavy UTF-8 decoding.
• Instruction merging: Consecutive character matches are merged into string blocks.
• Bitmap lookups: ASCII character classes use a 128-bit bitset for O(1) matching.

AI support: Gemini 3 Pro preview and a lot of my personal effort because AI is too stupid to understand same basic concepts. 😢

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: ba95cbb5ae

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

[spytheman]

Thank you @penguindark 🙇🏻 .
Excellent work.