/// <summary>
/// Creates a uniform distribution over any string starting and ending with a non-word character,
/// with a length in given bounds.
/// Characters other than the first and the last are restricted to be non-zero probability characters
/// from a given distribution.
/// </summary>
/// <param name="minLength">The minimum allowed string length.</param>
/// <param name="maxLength">The maximum allowed string length.</param>
/// <param name="allowedChars">The distribution representing allowed characters.</param>
/// <returns>The created distribution.</returns>
public static StringDistribution WordMiddle(int minLength, int maxLength, ImmutableDiscreteChar allowedChars)
{
if (maxLength < minLength)
{
throw new ArgumentException("The maximum length cannot be less than the minimum length.");
}
if (minLength < 1)
{
throw new ArgumentException("The minimum length must be at least one.");
}
var nonWordChar = StringDistribution.Char(NonWordCharacter);
if ((minLength == 1) && (maxLength == 1))
{
return(nonWordChar);
}
// TODO: make a PartialUniform copy of allowedChars
var suffix = StringDistribution.Repeat(allowedChars, minTimes: Math.Max(minLength - 2, 0), maxTimes: maxLength - 2);
suffix.AppendInPlace(nonWordChar);
if (minLength == 1)
{
var allowedChar = allowedChars.GetMode();
var allowedSuffix = new string(Enumerable.Repeat(allowedChar, Math.Max(minLength - 2, 0)).ToArray()) + ' ';
var suffixLogProb = suffix.GetLogProb(allowedSuffix);
suffix.SetToSumLog(suffixLogProb, StringDistribution.Empty(), 0.0, suffix);
}
return(nonWordChar + suffix);
}
/// <summary>
/// Creates a uniform distribution over strings of whitespace characters (see <see cref="DiscreteChar.Whitespace"/>),
/// with length within the given bounds.
/// If <paramref name="maxLength"/> is set to <see langword="null"/>,
/// there will be no upper bound on the length, and the resulting distribution will thus be improper.
/// </summary>
/// <param name="minLength">The minimum possible string length. Defaults to 1.</param>
/// <param name="maxLength">
/// The maximum possible sequence length, or <see langword="null"/> for no upper bound on length.
/// Defaults to <see langword="null"/>.
/// </param>
/// <returns>The created distribution.</returns>
public static StringDistribution Whitespace(int minLength = 1, int?maxLength = null)
{
return(StringDistribution.Repeat(DiscreteChar.Whitespace(), minLength, maxLength));
}
/// <summary>
/// Creates a uniform distribution over strings of word characters (see <see cref="DiscreteChar.WordChar"/>),
/// with length within the given bounds.
/// If <paramref name="maxLength"/> is set to <see langword="null"/>,
/// there will be no upper bound on the length, and the resulting distribution will thus be improper.
/// </summary>
/// <param name="minLength">The minimum possible string length. Defaults to 1.</param>
/// <param name="maxLength">
/// The maximum possible sequence length, or <see langword="null"/> for no upper bound on length.
/// Defaults to <see langword="null"/>.
/// </param>
/// <returns>The created distribution.</returns>
public static StringDistribution WordChars(int minLength = 1, int?maxLength = null)
{
return(StringDistribution.Repeat(DiscreteChar.WordChar(), minLength, maxLength));
}
/// <summary>
/// Creates a uniform distribution over strings of digits, lowercase and uppercase letters, with length within the given bounds.
/// If <paramref name="maxLength"/> is set to <see langword="null"/>,
/// there will be no upper bound on the length, and the resulting distribution will thus be improper.
/// </summary>
/// <param name="minLength">The minimum possible string length. Defaults to 1.</param>
/// <param name="maxLength">
/// The maximum possible sequence length, or <see langword="null"/> for no upper bound on length.
/// Defaults to <see langword="null"/>.
/// </param>
/// <returns>The created distribution.</returns>
public static StringDistribution LettersOrDigits(int minLength = 1, int?maxLength = null)
{
return(StringDistribution.Repeat(DiscreteChar.LetterOrDigit(), minLength, maxLength));
}
/// <summary>
/// Creates a uniform distribution over strings of digits, with length within the given bounds.
/// If <paramref name="maxLength"/> is set to <see langword="null"/>,
/// there will be no upper bound on the length, and the resulting distribution will thus be improper.
/// </summary>
/// <param name="minLength">The minimum possible string length. Defaults to 1.</param>
/// <param name="maxLength">
/// The maximum possible sequence length, or <see langword="null"/> for no upper bound on length.
/// Defaults to <see langword="null"/>.
/// </param>
/// <param name="uniformity">The type of uniformity. Defaults to <see cref="DistributionKind.UniformOverValue"/></param>
/// <returns>The created distribution.</returns>
public static StringDistribution Digits(int minLength = 1, int?maxLength = null, DistributionKind uniformity = DistributionKind.UniformOverValue)
{
return(StringDistribution.Repeat(DiscreteChar.Digit(), minLength, maxLength, uniformity));
}
/// <summary>
/// Creates a uniform distribution over strings of lowercase and uppercase letters, with length in given bounds.
/// If <paramref name="maxLength"/> is set to <see langword="null"/>,
/// there will be no upper bound on the length, and the resulting distribution will thus be improper.
/// </summary>
/// <param name="minLength">The minimum possible string length. Defaults to 1.</param>
/// <param name="maxLength">
/// The maximum possible sequence length, or <see langword="null"/> for no upper bound on length.
/// Defaults to <see langword="null"/>.
/// </param>
/// <returns>The created distribution.</returns>
public static StringDistribution Letters(int minLength = 1, int?maxLength = null)
{
return(StringDistribution.Repeat(ImmutableDiscreteChar.Letter(), minLength, maxLength));
}
