sica-fondt/brain/reasoning/momoa/assets/prompts/tool-instructions/reg-ex-validation-tool-string.md

name
reg-ex-validation-tool-string

RegEx Validation Tool (${strings/tool-prefix}REGEX/VALIDATE)

• Purpose: To build, test, and debug regular expressions before integrating them into source code. Use this tool whenever you are creating or modifying a regular expression.
• Syntax: ${strings/tool-prefix}REGEX/VALIDATE{ "regExString": "{SoRegEx}<raw_regex_pattern>{EoRegEx}", "flags": "<optional_flags>", "testCases": [ { "type": "validate", "input": "", "expected": <true_or_false> }, { "type": "extract", "input": "", "expected": "<extracted_string>", "group": 1 } ] }
• Rules and Usage:
  • Core Principle: Isolate and Test the Smallest Part
    • Think in Units: Treat RegEx validation like unit testing. When modifying a regular expression that is built from multiple parts in the code, you MUST identify and test only the specific component you are changing.
    • DO NOT Reconstruct Complex Patterns: Avoid recreating large, multi-part regular expressions from the source code. This is error-prone and leads to overly complex tests. Your goal is to verify your specific change, not to re-implement the entire RegEx logic within the tool.
    • Example: If the code is final_regex = part_A + part_B and you only need to modify part_A, your test should focus exclusively on part_A.
  • CRITICAL: RegEx Delimiters: To prevent parsing errors, the value of the regExString field MUST be wrapped in the {SoRegEx...EoRegEx} delimiters. The tool is designed to handle this specific format.
    • regExString (string): The Regular Expression pattern string to be tested, wrapped with the required delimiters. Do not escape any part of this, this should be the raw regex string to be evaluated.
    • flags (string, optional): Flags for the RegEx engine (e.g., "i" for case-insensitive, "g" for global). Defaults to an empty string.
    • testCases (array): An array of test case objects. Each object defines a single test and must contain:
      • type: The type of test. Must be one of the following:
        • validate: Returns true if the pattern is found, false otherwise.
        • extract: Extracts a string using capture groups.
        • replace: Replaces matched text with a new string.
        • findAll: Finds all non-overlapping matches.
        • split: Splits the input string using the regex as a delimiter.
      • input: The string input for the test case.
      • expected: The expected outcome. The data type depends on the type (e.g., boolean for validate, string for extract, array for findAll).
      • group (number, optional): For type: 'extract', the capture group index. Defaults to 0 (the full match).
      • replacement (string, optional): For type: 'replace', the string to replace matched patterns with.
  • Return Value: The tool returns a JSON object with isRegexValid, a summary of test results, and a testResults array detailing the outcome of each individual test case.