fix(functions): insert underscore between a digit and a capital letter in idToSnakeCaseFast

[sarmah-rup]

What

idToSnakeCaseFast in libs/functions/src/lib/strings.ts does not insert an underscore between a digit and a following capital letter. An id like field1Name becomes field1name instead of field1_name.

Why I hit this

I run events into a warehouse through Jitsu and noticed some of my columns came out mis-named. Properties like field1Name, address2Line and plan9FromOuterSpace landed as field1name, address2line and plan9from_outer_space rather than the snake_case I expected. That breaks downstream models that reference field1_name.

Root cause

In the loop, needUnderscore is only turned on after a latin letter, so the digit to capital boundary is missed:

// we add underscore only between latin letters
needUnderscore = (c >= aCode && c <= zCode) || (c >= ACode && c <= ZCode);

The comment even says "only between latin letters", but the canonical reference implementation idToSnakeCaseRegex in libs/core-functions-lib/src/functions/lib/strings.ts uses a (?<=[a-zA-Z0-9]) lookbehind, which includes digits. So the fast and regex versions disagree exactly on digit boundaries.

Fix

Include the digit range in needUnderscore, reusing the zeroCode / nineCode constants already declared at the top of the file:

needUnderscore = (c >= aCode && c <= zCode) || (c >= ACode && c <= ZCode) || (c >= zeroCode && c <= nineCode);

Tests

Added regression cases to libs/core-functions-lib/__tests__/strings.test.ts asserting the corrected output (field1Name becomes field1_name, address2Line becomes address2_line, plan9FromOuterSpace becomes plan9_from_outer_space) and that idToSnakeCaseFast now matches idToSnakeCaseRegex. The two new tests fail on newjitsu before the change and pass after. The full core-functions-lib suite is green at 42 tests.

[vklimontovich]

Hey @sarmah-rup, thank you for the PR! If we merge it, would it mean that columns containing digits will be renamed? May be too disruptive