Beginner guide

1. What StringKit-FP is

TStringKit is a collection of static functions for everyday text work: cleaning input, splitting strings, converting identifier case, validating common formats, encoding data, approximate comparison, and simple text analysis. It uses Free Pascal RTL/FCL components only.

Start with the static API. The optional string helper is a second spelling of most string-first calls, not a different implementation.

2. Add StringKit-FP to a project

For an FPC project, put src/ on the unit search path and add StringKit to uses.

fpc -Fupath/to/stringkit-fp/src your_program.pas

In Lazarus, open packages/lazarus/stringkit_fp.lpk, compile it, then choose Use → Add to Project. Alternatively add the repository’s src/ directory under Project Options → Compiler Options → Paths → Other Unit Files.

3. Make the first call

The first call program converts an identifier and prints:

hello_world

4. Choose static or helper style

The static style is explicit and works with only StringKit:

Clean := TStringKit.Trim(Input);

Add StringKitHelper when you prefer a pipeline-like spelling:

Clean := Input.Trim;

Both calls use the same library logic. Read Static API vs helper API before enabling selective helper feature flags.

5. Clean and transform text

Use Trim for surrounding whitespace and CollapseWhitespace for runs inside text. The compiled normalise input program prints Ada Lovelace. Use ToLower, ToUpper, or ToTitleCase for display-oriented changes. For code-like names, use the identifier converters described in Case conversion.

6. Split and join

Split returns TStringDynArray, declared in the Types unit. The compiled split and join program shows the full uses list and prints a three-item join. This is useful for simple delimited text, not quoted CSV fields.

7. Convert identifier case

Use ToCamelCase, ToPascalCase, ToSnakeCase, or ToKebabCase when converting a name such as HelloWorld or snake_case. Start with the identifier case program, then see the case conversion guide for acronyms and digits.

8. Validate common input

IsValidEmail, IsValidURL, IsValidIPv4, IsValidIPv6, and IsValidDate are practical syntax checks. They do not prove that an address exists, a URL is reachable, or input meets every RFC edge case. See Validation.

9. Encode text for the right context

Use HTMLEncode before inserting text into HTML. Use PercentEncode for URI components and FormURLEncode for HTML form data. URLEncode is intentionally a legacy-compatible form alias. The encoding guide explains the difference.

10. Prefer safe parsing for external input

When malformed data is normal, prefer TryHexDecode, TryDecode64, and TryFromRoman. They return False and clear the out value on malformed input. See Parsing and Try APIs.

11. Compare similar text

IsFuzzyMatch can compare close spellings such as colour and color. Its typed TFuzzyMethod selector makes the chosen algorithm explicit. See Fuzzy matching.

12. Analyse readability and text

CountWords, FleschReadingEase, FleschKincaidGradeLevel, and GenerateNGrams provide lightweight text analysis. The English readability calculations use heuristic syllable counts; they are useful signals, not editorial verdicts. See Text analysis.

13. Where to go next

Choose a complete task from Recipes, keep the Cheat Sheet as a reminder, and consult Contracts and limitations whenever input format or indexing matters.