private Config GetConfig(CleanStringType stringType, CultureInfo culture)
{
stringType = stringType & CleanStringType.RoleMask;
Dictionary <CleanStringType, Config> config;
if (_configs.ContainsKey(culture))
{
config = _configs[culture];
if (config.ContainsKey(stringType)) // have we got a config for _that_ role?
{
return(config[stringType]);
}
if (config.ContainsKey(CleanStringType.RoleMask)) // have we got a generic config for _all_ roles?
{
return(config[CleanStringType.RoleMask]);
}
}
else if (_configs.ContainsKey(_defaultCulture))
{
config = _configs[_defaultCulture];
if (config.ContainsKey(stringType)) // have we got a config for _that_ role?
{
return(config[stringType]);
}
if (config.ContainsKey(CleanStringType.RoleMask)) // have we got a generic config for _all_ roles?
{
return(config[CleanStringType.RoleMask]);
}
}
return(Config.NotConfigured);
}
public void CleanStringToAsciiWithCase(string input, string expected, CleanStringType caseType)
{
var output = _helper.CleanString(input, caseType | CleanStringType.Ascii);
// legacy does nothing
Assert.AreEqual(input, output);
}
// internal: we don't want ppl to retrieve a config and modify it
// (the helper uses a private clone to prevent modifications)
internal Config For(CleanStringType stringType, string culture)
{
culture = culture ?? "";
stringType = stringType & CleanStringType.RoleMask;
Dictionary <CleanStringType, Config> config;
if (_configs.ContainsKey(culture))
{
config = _configs[culture];
if (config.ContainsKey(stringType)) // have we got a config for _that_ role?
{
return(config[stringType]);
}
if (config.ContainsKey(CleanStringType.RoleMask)) // have we got a generic config for _all_ roles?
{
return(config[CleanStringType.RoleMask]);
}
}
else if (_configs.ContainsKey(DefaultCulture))
{
config = _configs[DefaultCulture];
if (config.ContainsKey(stringType)) // have we got a config for _that_ role?
{
return(config[stringType]);
}
if (config.ContainsKey(CleanStringType.RoleMask)) // have we got a generic config for _all_ roles?
{
return(config[CleanStringType.RoleMask]);
}
}
return(Config.NotConfigured);
}
public DefaultShortStringHelper WithConfig(CleanStringType stringRole,
Func <string, string> preFilter = null,
bool breakTermsOnUpper = true, bool allowLeadingDigits = false, bool allowUnderscoreInTerm = false)
{
return(WithConfig(_defaultCulture, stringRole,
preFilter, breakTermsOnUpper, allowLeadingDigits, allowUnderscoreInTerm));
}
public void LegacyConvertStringCase(string input, string expected, CleanStringType caseType)
{
// NOTE LegacyConvertStringCase has issues with a few cases
// -> ignore test cases
// also it removes symbols, etc... except the quote?
var output = _helper.LegacyConvertStringCase(input, caseType);
Assert.AreEqual(expected, output);
}
protected virtual string CleanString(string text, CleanStringType stringType, CultureInfo culture, char?separator)
{
// be safe
if (text == null)
{
throw new ArgumentNullException("text");
}
if (culture == null)
{
throw new ArgumentNullException("culture");
}
// get config
var config = GetConfig(stringType, culture);
stringType = config.StringTypeExtend(stringType);
// apply defaults
if ((stringType & CleanStringType.CaseMask) == CleanStringType.None)
{
stringType |= CleanStringType.CamelCase;
}
if ((stringType & CleanStringType.CodeMask) == CleanStringType.None)
{
stringType |= CleanStringType.Ascii;
}
// use configured unless specified
separator = separator ?? config.Separator;
// apply pre-filter
if (config.PreFilter != null)
{
text = config.PreFilter(text);
}
// apply replacements
//if (config.Replacements != null)
// text = ReplaceMany(text, config.Replacements);
// recode
var codeType = stringType & CleanStringType.CodeMask;
text = codeType == CleanStringType.Ascii
? Utf8ToAsciiConverter.ToAsciiString(text)
: RemoveSurrogatePairs(text);
// clean
text = CleanCodeString(text, stringType, separator.Value, culture, config);
// apply post-filter
if (config.PostFilter != null)
{
text = config.PostFilter(text);
}
return(text);
}
/// <summary>
/// Cleans a string in the context of a specified culture, using a specified separator and configuration.
/// </summary>
/// <param name="text">The text to clean.</param>
/// <param name="stringType">A flag indicating the target casing and encoding of the string. By default,
/// strings are cleaned up to camelCase and Ascii.</param>
/// <param name="separator">The separator.</param>
/// <param name="culture">The culture.</param>
/// <param name="config">The configuration.</param>
/// <returns>The clean string.</returns>
private string CleanString(string text, CleanStringType stringType, char separator, CultureInfo culture, HelperConfig config)
{
// be safe
if (text == null)
{
throw new ArgumentNullException("text");
}
if (culture == null)
{
throw new ArgumentNullException("culture");
}
// apply defaults
if ((stringType & CleanStringType.CaseMask) == CleanStringType.None)
{
stringType |= CleanStringType.CamelCase;
}
if ((stringType & CleanStringType.CodeMask) == CleanStringType.None)
{
stringType |= CleanStringType.Ascii;
}
var codeType = stringType & CleanStringType.CodeMask;
// apply pre-filter
if (config.PreFilter != null)
{
text = config.PreFilter(text);
}
// apply replacements
//if (config.Replacements != null)
// text = ReplaceMany(text, config.Replacements);
// recode
text = Recode(text, stringType);
// clean
switch (codeType)
{
case CleanStringType.Ascii:
// see note below - don't use CleanAsciiString
//text = CleanAsciiString(text, stringType, separator);
//break;
case CleanStringType.Utf8:
text = CleanUtf8String(text, stringType, separator, culture, config);
break;
case CleanStringType.Unicode:
throw new NotImplementedException("DefaultShortStringHelper does not handle unicode yet.");
default:
throw new ArgumentOutOfRangeException("stringType");
}
return(text);
}
public DefaultShortStringHelper WithConfig(CultureInfo culture, CleanStringType stringRole,
Func <string, string> preFilter = null,
bool breakTermsOnUpper = true, bool allowLeadingDigits = false, bool allowUnderscoreInTerm = false)
{
EnsureNotFrozen();
if (_configs.ContainsKey(culture) == false)
{
_configs[culture] = new Dictionary <CleanStringType, HelperConfig>();
}
_configs[culture][stringRole] = new HelperConfig(preFilter, breakTermsOnUpper, allowLeadingDigits, allowUnderscoreInTerm);
return(this);
}
public DefaultShortStringHelper WithConfig(CultureInfo culture, CleanStringType stringRole, Config config)
{
if (config == null)
{
throw new ArgumentNullException("config");
}
EnsureNotFrozen();
if (_configs.ContainsKey(culture) == false)
{
_configs[culture] = new Dictionary <CleanStringType, Config>();
}
_configs[culture][stringRole] = config.Clone(); // clone so it can't be changed
return(this);
}
/// <summary>
/// Filters a string to convert case, and more.
/// </summary>
/// <param name="phrase">the text to filter.</param>
/// <param name="cases">The string case type.</param>
/// <returns>The filtered text.</returns>
/// <remarks>
/// <para>This is the legacy method, so we can't really change it, although it has issues (see unit tests).</para>
/// <para>It does more than "converting the case", and also remove spaces, etc.</para>
/// </remarks>
public string LegacyConvertStringCase(string phrase, CleanStringType cases)
{
// ported from StringExtensions.ConvertCase
cases &= CleanStringType.CaseMask;
var splittedPhrase = Regex.Split(phrase, @"[^a-zA-Z0-9\']", RegexOptions.Compiled);
if (cases == CleanStringType.Unchanged)
{
return(string.Join("", splittedPhrase));
}
//var splittedPhrase = phrase.Split(' ', '-', '.');
var sb = new StringBuilder();
foreach (var splittedPhraseChars in splittedPhrase.Select(s => s.ToCharArray()))
{
if (splittedPhraseChars.Length > 0)
{
splittedPhraseChars[0] = ((new String(splittedPhraseChars[0], 1)).ToUpperInvariant().ToCharArray())[0];
}
sb.Append(new String(splittedPhraseChars));
}
var result = sb.ToString();
if (cases == CleanStringType.CamelCase)
{
if (result.Length > 1)
{
var pattern = new Regex("^([A-Z]*)([A-Z].*)$", RegexOptions.Singleline | RegexOptions.Compiled);
var match = pattern.Match(result);
if (match.Success)
{
result = match.Groups[1].Value.ToLowerInvariant() + match.Groups[2].Value;
return(result.Substring(0, 1).ToLowerInvariant() + result.Substring(1));
}
return(result);
}
return(result.ToLowerInvariant());
}
return(result);
}
public DefaultShortStringHelperConfig WithConfig(string culture, CleanStringType stringRole, Config config)
{
if (config == null)
{
throw new ArgumentNullException(nameof(config));
}
culture = culture ?? "";
if (_configs.ContainsKey(culture) == false)
{
_configs[culture] = new Dictionary <CleanStringType, Config>();
}
_configs[culture][stringRole] = config;
return(this);
}
// extends the config
public CleanStringType StringTypeExtend(CleanStringType stringType)
{
var st = StringType;
foreach (var mask in new[] { CleanStringType.CaseMask, CleanStringType.CodeMask })
{
var a = stringType & mask;
if (a == 0)
{
continue;
}
st = st & ~mask; // clear what we have
st = st | a; // set the new value
}
return(st);
}
/// <summary>
/// Returns a new string containing only characters within the specified code type.
/// </summary>
/// <param name="text">The string to filter.</param>
/// <param name="stringType">The string type.</param>
/// <returns>The filtered string.</returns>
/// <remarks>If <paramref name="stringType"/> is not <c>Unicode</c> then non-utf8 characters are
/// removed. If it is <c>Ascii</c> we try to do some intelligent replacement of accents, etc.</remarks>
public virtual string Recode(string text, CleanStringType stringType)
{
// be safe
if (text == null)
{
throw new ArgumentNullException("text");
}
var codeType = stringType & CleanStringType.CodeMask;
// unicode to utf8 or ascii: just remove the unicode chars
// utf8 to ascii: try to be clever and replace some chars
// what's the point?
if (codeType == CleanStringType.Unicode)
{
return(text);
}
return(codeType == CleanStringType.Utf8
? RemoveNonUtf8(text)
: Utf8ToAsciiConverter.ToAsciiString(text));
}
public string CleanString(string text, CleanStringType stringType, char separator)
{
return(text);
}
public string CleanString(string text, CleanStringType stringType, char separator, CultureInfo culture)
{
return(text);
}
/// <summary>
/// Cleans a string in the context of a specified culture.
/// </summary>
/// <param name="text">The text to clean.</param>
/// <param name="stringType">A flag indicating the target casing and encoding of the string. By default,
/// strings are cleaned up to camelCase and Ascii.</param>
/// <param name="culture">The culture.</param>
/// <returns>The clean string.</returns>
public string CleanString(string text, CleanStringType stringType, CultureInfo culture)
{
return(CleanString(text, stringType, culture, null));
}
// note: supports surrogate pairs in input string
internal void CopyTerm(string input, int ipos, char[] output, ref int opos, int len,
CleanStringType caseType, CultureInfo culture, bool isAcronym)
{
var term = input.Substring(ipos, len);
if (isAcronym)
{
if ((caseType == CleanStringType.CamelCase && len <= 2 && opos > 0) ||
(caseType == CleanStringType.PascalCase && len <= 2) ||
(caseType == CleanStringType.UmbracoCase))
{
caseType = CleanStringType.Unchanged;
}
}
// note: MSDN seems to imply that ToUpper or ToLower preserve the length
// of the string, but that this behavior is not guaranteed and could change.
char c;
int i;
string s;
switch (caseType)
{
//case CleanStringType.LowerCase:
//case CleanStringType.UpperCase:
case CleanStringType.Unchanged:
term.CopyTo(0, output, opos, len);
opos += len;
break;
case CleanStringType.LowerCase:
term = term.ToLower(culture);
term.CopyTo(0, output, opos, term.Length);
opos += term.Length;
break;
case CleanStringType.UpperCase:
term = term.ToUpper(culture);
term.CopyTo(0, output, opos, term.Length);
opos += term.Length;
break;
case CleanStringType.CamelCase:
c = term[0];
i = 1;
if (char.IsSurrogate(c))
{
s = term.Substring(ipos, 2);
s = opos == 0 ? s.ToLower(culture) : s.ToUpper(culture);
s.CopyTo(0, output, opos, s.Length);
opos += s.Length;
i++; // surrogate pair len is 2
}
else
{
output[opos] = opos++ == 0 ? char.ToLower(c, culture) : char.ToUpper(c, culture);
}
if (len > i)
{
term = term.Substring(i).ToLower(culture);
term.CopyTo(0, output, opos, term.Length);
opos += term.Length;
}
break;
case CleanStringType.PascalCase:
c = term[0];
i = 1;
if (char.IsSurrogate(c))
{
s = term.Substring(ipos, 2);
s = s.ToUpper(culture);
s.CopyTo(0, output, opos, s.Length);
opos += s.Length;
i++; // surrogate pair len is 2
}
else
{
output[opos++] = char.ToUpper(c, culture);
}
if (len > i)
{
term = term.Substring(i).ToLower(culture);
term.CopyTo(0, output, opos, term.Length);
opos += term.Length;
}
break;
case CleanStringType.UmbracoCase:
c = term[0];
i = 1;
if (char.IsSurrogate(c))
{
s = term.Substring(ipos, 2);
s = opos == 0 ? s : s.ToUpper(culture);
s.CopyTo(0, output, opos, s.Length);
opos += s.Length;
i++; // surrogate pair len is 2
}
else
{
output[opos] = opos++ == 0 ? c : char.ToUpper(c, culture);
}
if (len > i)
{
term = term.Substring(i);
term.CopyTo(0, output, opos, term.Length);
opos += term.Length;
}
break;
default:
throw new ArgumentOutOfRangeException("caseType");
}
}
/// <summary>
/// Cleans a string, using a specified separator.
/// </summary>
/// <param name="text">The text to clean.</param>
/// <param name="stringType">A flag indicating the target casing and encoding of the string. By default,
/// strings are cleaned up to camelCase and Ascii.</param>
/// <param name="separator">The separator.</param>
/// <returns>The clean string.</returns>
/// <remarks>The string is cleaned in the context of the default culture.</remarks>
public string CleanString(string text, CleanStringType stringType, char separator)
{
return CleanString(text, stringType, separator, _defaultCulture);
}
/// <summary>
/// Cleans a string.
/// </summary>
/// <param name="text">The text to clean.</param>
/// <param name="stringType">A flag indicating the target casing and encoding of the string. By default,
/// strings are cleaned up to camelCase and Ascii.</param>
/// <returns>The clean string.</returns>
/// <remarks>The string is cleaned in the context of the default culture.</remarks>
public string CleanString(string text, CleanStringType stringType)
{
return(CleanString(text, stringType, _defaultCulture, null));
}
public string CleanString(string text, CleanStringType stringType, char separator)
{
return "CLEAN-STRING-B::" + text;
}
/// <summary>
/// Cleans a string in the context of a specified culture, using a specified separator.
/// </summary>
/// <param name="text">The text to clean.</param>
/// <param name="stringType">A flag indicating the target casing and encoding of the string. By default,
/// strings are cleaned up to camelCase and Ascii.</param>
/// <param name="separator">The separator.</param>
/// <param name="culture">The culture.</param>
/// <returns>The clean string.</returns>
public virtual string CleanString(string text, CleanStringType stringType, char separator, CultureInfo culture)
{
var config = GetConfig(stringType & CleanStringType.RoleMask, culture);
return CleanString(text, stringType, separator, culture, config);
}
/// <summary>
/// Returns a new string containing only characters within the specified code type.
/// </summary>
/// <param name="text">The string to filter.</param>
/// <param name="stringType">The string type.</param>
/// <returns>The filtered string.</returns>
/// <remarks>If <paramref name="stringType"/> is not <c>Unicode</c> then non-utf8 characters are
/// removed. If it is <c>Ascii</c> we try to do some intelligent replacement of accents, etc.</remarks>
public virtual string Recode(string text, CleanStringType stringType)
{
// be safe
if (text == null)
throw new ArgumentNullException("text");
var codeType = stringType & CleanStringType.CodeMask;
// unicode to utf8 or ascii: just remove the unicode chars
// utf8 to ascii: try to be clever and replace some chars
// what's the point?
if (codeType == CleanStringType.Unicode)
return text;
return codeType == CleanStringType.Utf8
? RemoveNonUtf8(text)
: Utf8ToAsciiConverter.ToAsciiString(text);
}
public DefaultShortStringHelper WithConfig(CultureInfo culture, CleanStringType stringRole,
Func<string, string> preFilter = null,
bool breakTermsOnUpper = true, bool allowLeadingDigits = false, bool allowUnderscoreInTerm = false)
{
EnsureNotFrozen();
if (_configs.ContainsKey(culture) == false)
_configs[culture] = new Dictionary<CleanStringType, HelperConfig>();
_configs[culture][stringRole] = new HelperConfig(preFilter, breakTermsOnUpper, allowLeadingDigits, allowUnderscoreInTerm);
return this;
}
internal void CopyUtf8Term(string input, int ipos, char[] output, ref int opos, int len,
CleanStringType caseType, CultureInfo culture, /*Func<string, string> termFilter,*/ bool isAcronym)
{
var term = input.Substring(ipos, len);
ipos = 0;
//if (termFilter != null)
//{
// term = termFilter(term);
// len = term.Length;
//}
if (isAcronym)
{
if ((caseType == CleanStringType.CamelCase && len <= 2 && opos > 0) ||
(caseType == CleanStringType.PascalCase && len <= 2) ||
(caseType == CleanStringType.UmbracoCase))
caseType = CleanStringType.Unchanged;
}
char c;
switch (caseType)
{
//case CleanStringType.LowerCase:
//case CleanStringType.UpperCase:
case CleanStringType.Unchanged:
term.CopyTo(ipos, output, opos, len);
opos += len;
break;
case CleanStringType.LowerCase:
term.ToLower(culture).CopyTo(ipos, output, opos, len);
opos += len;
break;
case CleanStringType.UpperCase:
term.ToUpper(culture).CopyTo(ipos, output, opos, len);
opos += len;
break;
case CleanStringType.CamelCase:
c = term[ipos++];
output[opos] = opos++ == 0 ? char.ToLower(c, culture) : char.ToUpper(c, culture);
if (len > 1)
term.ToLower(culture).CopyTo(ipos, output, opos, len - 1);
opos += len - 1;
break;
case CleanStringType.PascalCase:
c = term[ipos++];
output[opos++] = char.ToUpper(c, culture);
if (len > 1)
term.ToLower(culture).CopyTo(ipos, output, opos, len - 1);
opos += len - 1;
break;
case CleanStringType.UmbracoCase:
c = term[ipos++];
output[opos] = opos++ == 0 ? c : char.ToUpper(c, culture);
if (len > 1)
term.CopyTo(ipos, output, opos, len - 1);
opos += len - 1;
break;
default:
throw new ArgumentOutOfRangeException("caseType");
}
}
internal string CleanUtf8String(string text, CleanStringType caseType, char separator, CultureInfo culture, HelperConfig config)
{
int opos = 0, ipos = 0;
var state = StateBreak;
caseType &= CleanStringType.CaseMask;
// if we apply global ToUpper or ToLower to text here
// then we cannot break words on uppercase chars
var input = text;
// it's faster to use an array than a StringBuilder
var ilen = input.Length;
var output = new char[ilen * 2]; // twice the length should be OK in all cases
//var termFilter = config.TermFilter;
for (var i = 0; i < ilen; i++)
{
var c = input[i];
var isDigit = char.IsDigit(c);
var isUpper = char.IsUpper(c); // false for digits, symbols...
var isLower = char.IsLower(c); // false for digits, symbols...
var isUnder = config.AllowUnderscoreInTerm && c == '_';
var isTerm = char.IsLetterOrDigit(c) || isUnder;
switch (state)
{
case StateBreak:
if (isTerm && (opos > 0 || (isUnder == false && (config.AllowLeadingDigits || isDigit == false))))
{
ipos = i;
if (opos > 0 && separator != char.MinValue)
output[opos++] = separator;
state = isUpper ? StateUp : StateWord;
}
break;
case StateWord:
if (isTerm == false || (config.BreakTermsOnUpper && isUpper))
{
CopyUtf8Term(input, ipos, output, ref opos, i - ipos, caseType, culture, /*termFilter,*/ false);
ipos = i;
state = isTerm ? StateUp : StateBreak;
if (state != StateBreak && separator != char.MinValue)
output[opos++] = separator;
}
break;
case StateAcronym:
if (isTerm == false || isLower || isDigit)
{
if (isLower && config.GreedyAcronyms == false)
i -= 1;
CopyUtf8Term(input, ipos, output, ref opos, i - ipos, caseType, culture, /*termFilter,*/ true);
ipos = i;
state = isTerm ? StateWord : StateBreak;
if (state != StateBreak && separator != char.MinValue)
output[opos++] = separator;
}
break;
case StateUp:
if (isTerm)
{
state = isUpper ? StateAcronym : StateWord;
}
else
{
CopyUtf8Term(input, ipos, output, ref opos, 1, caseType, culture, /*termFilter,*/ false);
state = StateBreak;
}
break;
default:
throw new Exception("Invalid state.");
}
}
switch (state)
{
case StateBreak:
break;
case StateWord:
CopyUtf8Term(input, ipos, output, ref opos, input.Length - ipos, caseType, culture, /*termFilter,*/ false);
break;
case StateAcronym:
case StateUp:
CopyUtf8Term(input, ipos, output, ref opos, input.Length - ipos, caseType, culture, /*termFilter,*/ true);
break;
default:
throw new Exception("Invalid state.");
}
return new string(output, 0, opos);
}
/// <summary>
/// Cleans a string in the context of a specified culture, using a specified separator and configuration.
/// </summary>
/// <param name="text">The text to clean.</param>
/// <param name="stringType">A flag indicating the target casing and encoding of the string. By default,
/// strings are cleaned up to camelCase and Ascii.</param>
/// <param name="separator">The separator.</param>
/// <param name="culture">The culture.</param>
/// <param name="config">The configuration.</param>
/// <returns>The clean string.</returns>
private string CleanString(string text, CleanStringType stringType, char separator, CultureInfo culture, HelperConfig config)
{
// be safe
if (text == null)
throw new ArgumentNullException("text");
if (culture == null)
throw new ArgumentNullException("culture");
// apply defaults
if ((stringType & CleanStringType.CaseMask) == CleanStringType.None)
stringType |= CleanStringType.CamelCase;
if ((stringType & CleanStringType.CodeMask) == CleanStringType.None)
stringType |= CleanStringType.Ascii;
var codeType = stringType & CleanStringType.CodeMask;
// apply pre-filter
if (config.PreFilter != null)
text = config.PreFilter(text);
// apply replacements
//if (config.Replacements != null)
// text = ReplaceMany(text, config.Replacements);
// recode
text = Recode(text, stringType);
// clean
switch (codeType)
{
case CleanStringType.Ascii:
// see note below - don't use CleanAsciiString
//text = CleanAsciiString(text, stringType, separator);
//break;
case CleanStringType.Utf8:
text = CleanUtf8String(text, stringType, separator, culture, config);
break;
case CleanStringType.Unicode:
throw new NotImplementedException("DefaultShortStringHelper does not handle unicode yet.");
default:
throw new ArgumentOutOfRangeException("stringType");
}
return text;
}
/// <summary>
/// Cleans a string.
/// </summary>
/// <param name="text">The text to clean.</param>
/// <param name="stringType">A flag indicating the target casing and encoding of the string. By default,
/// strings are cleaned up to camelCase and Ascii.</param>
/// <returns>The clean string.</returns>
/// <remarks>The string is cleaned in the context of the default culture.</remarks>
public string CleanString(string text, CleanStringType stringType)
{
return CleanString(text, stringType, char.MinValue, _defaultCulture);
}
public DefaultShortStringHelper WithConfig(CleanStringType stringRole, Config config)
{
return(WithConfig(_defaultCulture, stringRole, config));
}
public string CleanString(string text, CleanStringType stringType)
{
return "CLEAN-STRING-A::" + text;
}
/// <summary>
/// Cleans a string in the context of a specified culture.
/// </summary>
/// <param name="text">The text to clean.</param>
/// <param name="stringType">A flag indicating the target casing and encoding of the string. By default,
/// strings are cleaned up to camelCase and Ascii.</param>
/// <param name="culture">The culture.</param>
/// <returns>The clean string.</returns>
public string CleanString(string text, CleanStringType stringType, CultureInfo culture)
{
return CleanString(text, stringType, char.MinValue, culture);
}
public string CleanString(string text, CleanStringType stringType, char separator, System.Globalization.CultureInfo culture)
{
return "CLEAN-STRING-D::" + text;
}
public DefaultShortStringHelper WithConfig(CleanStringType stringRole,
Func<string, string> preFilter = null,
bool breakTermsOnUpper = true, bool allowLeadingDigits = false, bool allowUnderscoreInTerm = false)
{
return WithConfig(_defaultCulture, stringRole,
preFilter, breakTermsOnUpper, allowLeadingDigits, allowUnderscoreInTerm);
}
public void CleanStringWithTypeAndCulture(string input, string expected, string culture, CleanStringType stringType)
{
var cinfo = culture == null ? CultureInfo.InvariantCulture : new CultureInfo(culture);
var separator = (stringType & CleanStringType.Url) == CleanStringType.Url ? '-' : char.MinValue;
var output = _helper.CleanString(input, stringType, separator, cinfo);
Assert.AreEqual(expected, output);
}
// the new methods to clean a string (to alias, url segment...)
/// <summary>
/// Cleans a string.
/// </summary>
/// <param name="text">The text to clean.</param>
/// <param name="stringType">A flag indicating the target casing and encoding of the string. By default,
/// strings are cleaned up to camelCase and Ascii.</param>
/// <returns>The clean string.</returns>
/// <remarks>The string is cleaned in the context of the IShortStringHelper default culture.</remarks>
public static string ToCleanString(this string text, CleanStringType stringType)
{
return(ShortStringHelper.CleanString(text, stringType));
}
public void LegacyConvertStringCase(string input, string expected, CleanStringType caseType)
{
// NOTE LegacyConvertStringCase has issues with a few cases
// -> ignore test cases
// also it removes symbols, etc... except the quote?
var output = _helper.LegacyConvertStringCase(input, caseType);
Assert.AreEqual(expected, output);
}
public void CleanStringToAsciiWithTypeAndSeparator(string input, string expected, char separator, CleanStringType caseType)
{
var output = _helper.CleanString(input, caseType | CleanStringType.Ascii, separator);
Assert.AreEqual(expected, output);
}
/// <summary>
/// Cleans a string, using a specified separator.
/// </summary>
/// <param name="text">The text to clean.</param>
/// <param name="stringType">A flag indicating the target casing and encoding of the string. By default,
/// strings are cleaned up to camelCase and Ascii.</param>
/// <param name="separator">The separator.</param>
/// <returns>The clean string.</returns>
/// <remarks>The string is cleaned in the context of the default culture.</remarks>
public string CleanString(string text, CleanStringType stringType, char separator)
{
return(CleanString(text, stringType, _defaultCulture, separator));
}
public string CleanString(string text, CleanStringType stringType, char separator)
{
return text;
}
// here was a subtle, ascii-optimized version of the cleaning code, and I was
// very proud of it until benchmarking showed it was an order of magnitude slower
// that the utf8 version. Micro-optimizing sometimes isn't such a good idea.
// note: does NOT support surrogate pairs in text
internal string CleanCodeString(string text, CleanStringType caseType, char separator, CultureInfo culture, Config config)
{
int opos = 0, ipos = 0;
var state = StateBreak;
caseType &= CleanStringType.CaseMask;
// if we apply global ToUpper or ToLower to text here
// then we cannot break words on uppercase chars
var input = text;
// it's faster to use an array than a StringBuilder
var ilen = input.Length;
var output = new char[ilen * 2]; // twice the length should be OK in all cases
for (var i = 0; i < ilen; i++)
{
var c = input[i];
// leading as long as StateBreak and ipos still zero
var leading = state == StateBreak && ipos == 0;
var isTerm = config.IsTerm(c, leading);
//var isDigit = char.IsDigit(c);
var isUpper = char.IsUpper(c); // false for digits, symbols...
//var isLower = char.IsLower(c); // false for digits, symbols...
// what should I do with surrogates?
// no idea, really, so they are not supported at the moment
var isPair = char.IsSurrogate(c);
if (isPair)
{
throw new NotSupportedException("Surrogate pairs are not supported.");
}
switch (state)
{
// within a break
case StateBreak:
// begin a new term if char is a term char,
// and ( pos > 0 or it's also a valid leading char )
if (isTerm)
{
ipos = i;
if (opos > 0 && separator != char.MinValue)
{
output[opos++] = separator;
}
state = isUpper ? StateUp : StateWord;
}
break;
// within a term / word
case StateWord:
// end a term if char is not a term char,
// or ( it's uppercase and we break terms on uppercase)
if (isTerm == false || (config.BreakTermsOnUpper && isUpper))
{
CopyTerm(input, ipos, output, ref opos, i - ipos, caseType, culture, false);
ipos = i;
state = isTerm ? StateUp : StateBreak;
if (state != StateBreak && separator != char.MinValue)
{
output[opos++] = separator;
}
}
break;
// within a term / acronym
case StateAcronym:
// end an acronym if char is not a term char,
// or if it's not uppercase / config
if (isTerm == false || (config.CutAcronymOnNonUpper && isUpper == false))
{
// whether it's part of the acronym depends on whether we're greedy
if (isTerm && config.GreedyAcronyms == false)
{
i -= 1; // handle that char again, in another state - not part of the acronym
}
if (i - ipos > 1) // single-char can't be an acronym
{
CopyTerm(input, ipos, output, ref opos, i - ipos, caseType, culture, true);
ipos = i;
state = isTerm ? StateWord : StateBreak;
if (state != StateBreak && separator != char.MinValue)
{
output[opos++] = separator;
}
}
else if (isTerm)
{
state = StateWord;
}
}
else if (isUpper == false) // isTerm == true
{
// it's a term char and we don't cut...
// keep moving forward as a word
state = StateWord;
}
break;
// within a term / uppercase = could be a word or an acronym
case StateUp:
if (isTerm)
{
// add that char to the term and pick word or acronym
state = isUpper ? StateAcronym : StateWord;
}
else
{
// single char, copy then break
CopyTerm(input, ipos, output, ref opos, 1, caseType, culture, false);
state = StateBreak;
}
break;
default:
throw new Exception("Invalid state.");
}
}
switch (state)
{
case StateBreak:
break;
case StateWord:
CopyTerm(input, ipos, output, ref opos, input.Length - ipos, caseType, culture, false);
break;
case StateAcronym:
case StateUp:
CopyTerm(input, ipos, output, ref opos, input.Length - ipos, caseType, culture, true);
break;
default:
throw new Exception("Invalid state.");
}
return(new string(output, 0, opos));
}
public string CleanString(string text, CleanStringType stringType, char separator, CultureInfo culture)
{
return text;
}
// legacy does not implement these
public string CleanString(string text, CleanStringType stringType)
{
return(text);
}
public void CleanStringToAsciiWithType(string input, string expected, CleanStringType caseType)
{
var output = _helper.CleanString(input, caseType | CleanStringType.Ascii);
Assert.AreEqual(expected, output);
}
public string CleanString(string text, CleanStringType stringType, CultureInfo culture)
{
return(text);
}
public string CleanString(string text, CleanStringType stringType, char separator, string culture)
{
return("CLEAN-STRING-D::" + text);
}
/// <summary>
/// Cleans a string, using a specified separator.
/// </summary>
/// <param name="text">The text to clean.</param>
/// <param name="stringType">A flag indicating the target casing and encoding of the string. By default,
/// strings are cleaned up to camelCase and Ascii.</param>
/// <param name="separator">The separator.</param>
/// <returns>The clean string.</returns>
/// <remarks>The string is cleaned in the context of the IShortStringHelper default culture.</remarks>
public static string ToCleanString(this string text, CleanStringType stringType, char separator)
{
return(ShortStringHelper.CleanString(text, stringType, separator));
}
private HelperConfig GetConfig(CleanStringType stringType, CultureInfo culture)
{
Dictionary<CleanStringType, HelperConfig> config;
if (_configs.ContainsKey(culture))
{
config = _configs[culture];
if (config.ContainsKey(stringType)) // have we got a config for _that_ role?
return config[stringType];
if (config.ContainsKey(CleanStringType.RoleMask)) // have we got a generic config for _all_ roles?
return config[CleanStringType.RoleMask];
}
else if (_configs.ContainsKey(_defaultCulture))
{
config = _configs[_defaultCulture];
if (config.ContainsKey(stringType)) // have we got a config for _that_ role?
return config[stringType];
if (config.ContainsKey(CleanStringType.RoleMask)) // have we got a generic config for _all_ roles?
return config[CleanStringType.RoleMask];
}
return HelperConfig.Empty;
}
public void CleanStringToAsciiWithCaseAndSeparator(string input, string expected, char separator, CleanStringType caseType)
{
var output = _helper.CleanString(input, caseType | CleanStringType.Ascii, separator);
// legacy does nothing
Assert.AreEqual(input, output);
}
/// <summary>
/// Cleans a string in the context of a specified culture, using a specified separator.
/// </summary>
/// <param name="text">The text to clean.</param>
/// <param name="stringType">A flag indicating the target casing and encoding of the string. By default,
/// strings are cleaned up to camelCase and Ascii.</param>
/// <param name="separator">The separator.</param>
/// <param name="culture">The culture.</param>
/// <returns>The clean string.</returns>
public static string ToCleanString(this string text, CleanStringType stringType, char separator, CultureInfo culture)
{
return(ShortStringHelper.CleanString(text, stringType, separator, culture));
}
// legacy does not implement these
public string CleanString(string text, CleanStringType stringType)
{
return text;
}
public void CleanStringWithTypeAndCulture(string input, string expected, string culture, CleanStringType stringType)
{
// picks the proper config per culture
// and overrides some stringType params (ascii...)
var output = ShortStringHelper.CleanString(input, stringType, culture);
Assert.AreEqual(expected, output);
}
public string CleanString(string text, CleanStringType stringType, CultureInfo culture)
{
return text;
}
public string CleanString(string text, CleanStringType stringType)
{
return("CLEAN-STRING-A::" + text);
}
/// <summary>
/// Filters a string to convert case, and more.
/// </summary>
/// <param name="phrase">the text to filter.</param>
/// <param name="cases">The string case type.</param>
/// <returns>The filtered text.</returns>
/// <remarks>
/// <para>This is the legacy method, so we can't really change it, although it has issues (see unit tests).</para>
/// <para>It does more than "converting the case", and also remove spaces, etc.</para>
/// </remarks>
public string LegacyConvertStringCase(string phrase, CleanStringType cases)
{
// ported from StringExtensions.ConvertCase
cases &= CleanStringType.CaseMask;
var splittedPhrase = Regex.Split(phrase, @"[^a-zA-Z0-9\']", RegexOptions.Compiled);
if (cases == CleanStringType.Unchanged)
return string.Join("", splittedPhrase);
//var splittedPhrase = phrase.Split(' ', '-', '.');
var sb = new StringBuilder();
foreach (var splittedPhraseChars in splittedPhrase.Select(s => s.ToCharArray()))
{
if (splittedPhraseChars.Length > 0)
{
splittedPhraseChars[0] = ((new String(splittedPhraseChars[0], 1)).ToUpperInvariant().ToCharArray())[0];
}
sb.Append(new String(splittedPhraseChars));
}
var result = sb.ToString();
if (cases == CleanStringType.CamelCase)
{
if (result.Length > 1)
{
var pattern = new Regex("^([A-Z]*)([A-Z].*)$", RegexOptions.Singleline | RegexOptions.Compiled);
var match = pattern.Match(result);
if (match.Success)
{
result = match.Groups[1].Value.ToLowerInvariant() + match.Groups[2].Value;
return result.Substring(0, 1).ToLowerInvariant() + result.Substring(1);
}
return result;
}
return result.ToLowerInvariant();
}
return result;
}
public string CleanString(string text, CleanStringType stringType, string culture)
{
return("CLEAN-STRING-C::" + text);
}