Beginner recipes

Every Pascal program below is compiled with FPC 3.2.2 and run against its stated output by tools/test_docs_examples.py.

Trim and normalise user input#

Problem: Turn accidental surrounding and repeated whitespace into one readable value.

Recommended API: Trim and CollapseWhitespace.

program NormalizeInput;

{$mode objfpc}{$H+}

uses
  StringKit;

begin
  Writeln(TStringKit.CollapseWhitespace(TStringKit.Trim('  Ada   Lovelace  ')));
end.

Expected output: Ada Lovelace

Caveat: This handles the library’s ordinary byte whitespace characters; it is not Unicode whitespace normalisation. Source program.

Convert HelloWorld to hello_world#

Problem: Change a code identifier to snake case.

Recommended API: ToSnakeCase.

program IdentifierCases;

{$mode objfpc}{$H+}

uses
  StringKit;

begin
  Writeln(TStringKit.ToSnakeCase('HelloWorld'));
  Writeln(TStringKit.ToPascalCase('snake_case'));
end.

Expected output:

hello_world
SnakeCase

Caveat: Identifier tokenisation is ASCII/byte-oriented. Source program.

Convert snake_case to PascalCase#

Problem: Create a Pascal type or class-like name from a snake-case name.

Recommended API: ToPascalCase.

program IdentifierCases;

{$mode objfpc}{$H+}

uses
  StringKit;

begin
  Writeln(TStringKit.ToSnakeCase('HelloWorld'));
  Writeln(TStringKit.ToPascalCase('snake_case'));
end.

Expected output:

hello_world
SnakeCase

Caveat: The same program demonstrates both identifier conversions. Source program.

Split a simple delimited string and join strings#

Problem: Handle a short, unquoted delimiter-separated value.

Recommended API: Split and Join.

program SplitAndJoin;

{$mode objfpc}{$H+}

uses
  Types,
  StringKit;

var
  Parts: TStringDynArray;
begin
  Parts := TStringKit.Split('red,green,blue', ',');
  Writeln(Length(Parts));
  Writeln(TStringKit.Join(Parts, ' | '));
end.

Expected output:

3
red | green | blue

Caveat: Split is for simple delimiters, not a full CSV parser with quoted fields. TStringDynArray is declared by Types. Source program.

Validate an email#

Problem: Reject clearly malformed email-shaped input before continuing.

Recommended API: IsValidEmail.

program ValidateEmail;

{$mode objfpc}{$H+}

uses
  SysUtils,
  StringKit;

begin
  Writeln(BoolToStr(TStringKit.IsValidEmail('ada@example.com'), True));
end.

Expected output: True

Caveat: This is a practical syntax check, not confirmation that a mailbox exists or full RFC validation. Source program.

Validate IPv4 and IPv6#

Problem: Check which common IP literals are syntactically valid.

Recommended API: IsValidIPv4 and IsValidIPv6.

program ValidateIp;

{$mode objfpc}{$H+}

uses
  SysUtils,
  StringKit;

begin
  Writeln(BoolToStr(TStringKit.IsValidIPv4('192.168.0.1'), True));
  Writeln(BoolToStr(TStringKit.IsValidIPv6('2001:db8::1'), True));
end.

Expected output:

True
True

Caveat: Validation does not establish network reachability. Source program.

Encode HTML#

Problem: Render text safely as text inside an HTML context.

Recommended API: HTMLEncode.

program HtmlEncode;

{$mode objfpc}{$H+}

uses
  StringKit;

begin
  Writeln(TStringKit.HTMLEncode('<b>Hello</b>'));
end.

Expected output: <b>Hello</b>

Caveat: This encodes the five essential HTML characters. Contexts such as JavaScript and CSS need their own safe handling. Source program.

Form URL encoding and percent encoding#

Problem: Encode text for a URI component or an HTML form body.

Recommended API: PercentEncode or FormURLEncode.

program UrlEncoding;

{$mode objfpc}{$H+}

uses
  StringKit;

begin
  Writeln(TStringKit.PercentEncode('a b+c'));
  Writeln(TStringKit.FormURLEncode('a b+c'));
end.

Expected output:

a%20b%2Bc
a+b%2Bc

Caveat: Spaces differ: percent encoding uses %20, while form encoding uses +. Source program.

Decode hex safely#

Problem: Decode external hexadecimal without accepting malformed pairs.

Recommended API: TryHexDecode.

program TryHexDecode;

{$mode objfpc}{$H+}

uses
  StringKit;

var
  Decoded: string;
begin
  if TStringKit.TryHexDecode('48656C6C6F', Decoded) then
    Writeln(Decoded);
end.

Expected output: Hello

Caveat: On malformed input the function returns False and clears Decoded. Source program.

Decode Base64 safely#

Problem: Decode expected Base64 while handling malformed input explicitly.

Recommended API: TryDecode64.

program TryBase64Decode;

{$mode objfpc}{$H+}

uses
  StringKit;

var
  Decoded: string;
begin
  if TStringKit.TryDecode64('SGVsbG8=', Decoded) then
    Writeln(Decoded);
end.

Expected output: Hello

Caveat: TryDecode64 accepts whitespace in valid Base64 text; malformed input returns False and clears Decoded. Source program.

Convert Roman numerals safely#

Problem: Accept only canonical Roman numerals in the supported range.

Recommended API: TryFromRoman.

program TryRoman;

{$mode objfpc}{$H+}

uses
  StringKit;

var
  Value: Integer;
begin
  if TStringKit.TryFromRoman('MMXXVI', Value) then
    Writeln(Value);
end.

Expected output: 2026

Caveat: The strict parser accepts canonical values from 1 through 3999 and clears Value when it returns False. Source program.

Compare two strings approximately#

Problem: Treat close spellings as a match.

Recommended API: IsFuzzyMatch with TFuzzyMethod.

program FuzzyMatch;

{$mode objfpc}{$H+}

uses
  SysUtils,
  StringKit;

begin
  Writeln(BoolToStr(
    TStringKit.IsFuzzyMatch('colour', 'color', 0.75, fmLevenshtein), True));
end.

Expected output: True

Caveat: A threshold is a product decision; test it with your real data. Source program.

Calculate readability#

Problem: Get a quick English readability score for a short passage.

Recommended API: FleschReadingEase.

program Readability;

{$mode objfpc}{$H+}

uses
  StringKit;

begin
  Writeln(TStringKit.FleschReadingEase('The cat sat on the mat.'):0:2);
end.

Expected output: 100.00

Caveat: Syllables are estimated heuristically for English text. Source program.

Extract text with a regular expression#

Problem: Find each matching fragment in a string.

Recommended API: ExtractAllMatches.

program RegexExtract;

{$mode objfpc}{$H+}

uses
  Types,
  StringKit;

var
  Matches: TStringDynArray;
  Index: Integer;
begin
  Matches := TStringKit.ExtractAllMatches('Order #42, then #7', '#\d+');
  for Index := 0 to High(Matches) do
    Writeln(Matches[Index]);
end.

Expected output:

#42
#7

Caveat: Patterns use Free Pascal’s RegExpr syntax. Source program.