fix: stopNodes should preserve text content as string (#795)

[abhu85]

Summary

• Fixes stopNodes content being parsed as numbers/booleans instead of remaining as strings
• <a>6</a> with stopNodes: ["*.a"] now returns { a: "6" } instead of { a: 6 }

Problem

When using stopNodes option, numeric and boolean-like text content was being converted to JavaScript numbers/booleans. According to the documentation, stopNodes should preserve content as raw text.

Reproduction (from #795):

const parser = new XMLParser({ stopNodes: ["*.a"] });
console.log(typeof parser.parse("<a>6</a>").a); // "number" - should be "string"

Solution

Added doNotParseTagValue parameter to parseTextData function and set it to true when processing stopNode content. This skips the parseValue call that converts strings to numbers/booleans, while still allowing tagValueProcessor to work for user customization.

Test Plan

• All 282 existing tests pass
• Added regression test covering: numeric strings, boolean strings ("true"/"false"), negative/decimal numbers, scientific notation

Compatibility

• No breaking changes
• Backwards compatible - existing tagValueProcessor behavior preserved

Fixes #795

[abhu85]

Hi @AmitagPro 👋

Friendly ping on this PR. It fixes the stopNodes option to preserve text content as a string (#795), with tests covering the fix.

All CI checks are passing. Would appreciate a review when you have a moment. Thanks!