Allow larger autocorrect dictionaries via configurable link size

[MustafaNisar2002]

Description

This PR relaxes the size limit on autocorrect dictionaries so tables larger than 64KB can be generated and used safely.

• Allow the autocorrect runtime to read tries with either 2- or 3-byte node links, using AUTOCORRECT_LINK_BYTE_COUNT defined in the generated header.
• Update the autocorrect data generator to compute entry sizes, try 2-byte links first, and automatically switch to 3-byte links when the serialized table would exceed 64KB.
• Emit AUTOCORRECT_LINK_BYTE_COUNT in generated autocorrect_data.h files and annotate the default autocorrect dictionary with its link size.
• Document the new link format behaviour in docs/features/autocorrect.md.

Types of Changes

• Core
• Enhancement/optimization
• Documentation

Issues Fixed or Closed by This PR

• Fixes [Feature Request] Increase autocorrect dictionary size limit #25823

Checklist

• My code follows the code style of this project: C, Python
• I have read the PR Checklist document and have made the appropriate changes.
• My change requires a change to the documentation.
• I have updated the documentation accordingly.
• I have added tests or basic checks to cover my changes (e.g. python -m compileall lib/python/qmk/cli/generate/autocorrect_data.py and firmware builds with a generated dictionary).
• I have tested the changes and verified that they work and don't break anything (as well as I can manage).

[zvecr]

We would not accept platform specific logic in core, when there is an abstraction in place that already exists.

platforms/progmem.h would need to be updated to add pgm_read_byte_far compatibility.

Suggested change

	#if defined(__AVR__) && DICTIONARY_SIZE > UINT16_MAX
	#if DICTIONARY_SIZE > UINT16_MAX

[zvecr]

You also have both c and py code that needs to pass formatting checks.