Download latest version here

Version 1.59 - 14 May 2012

• Tweaked code so it'll compile under Clarion 5.5 (AppendA instead of Append internally.) 

Version 1.58 - 25 April 2012

• Changed Z_ equates so as not to clash with equates defined in Draw.

Version 1.57 - 17 April 2012

• Added Gzip and Gunzip methods. Requires the ZLIBWAPI.DLL (you will need to distribute this DLL with your application).

• New parameter added to Split method, to allow for lines to be clipped before adding them to the Lines Queue.

Version 1.56 - 20 February 2012

• Added: SplitEvery method, which splits the string into substrings based on the passed length and stores each substring in the lines property.

Version 1.55 - 15 February 2012

• Added new method - CleanfileName.
• Changed Append method so pClip parameter is overloaded. Now has the ability to suppress the Separator if the NewValue is blank.

Version 1.54 - 12 January 2012

• Added optional pSep parameter to Append method.
• Added optional parameters pClip and pSep to Prepend method.
• Added optional parameter pWholeWord to Instring method.

Version 1.53 - 4 January 2012

• Added: SetLine, AddLine and DeleteLine methods.

Version 1.52 - 22 November 2011

• Fixed: The Random function not correctly supporting the st:AlphaFirst flag for strings longer than a single character.
• The WordStart, WordEnd and Squeeze methods now support a Type parameter that allows the type of text (and hence the manner in which the methods operate) to be specified.

Version 1.51 - 9 November 2011

• Renamed: Internal property "length" renamed to "_length" so it will not inadvertently be used when the user means to call the .Length() method.

Version 1.50 - 16 September 2011

Required for CryptoNite 1.1.8 and up

• Added: Str returns the stored string
• Added: Str (string newValue) stores the contents of the string and returns the string.
• Added: Str (*string newValue) stores the contents of the string and returns the string.
• Added: SetSlice sets the value of a substring within the stored string.
• Added: Insert Inserts a passed string into the stored string at the specified position.
• Added: Quote adds the specified "quotes" to the stored (wraps the string in the passed quotes).
• Added: UnQuote removes quotation marks (or the specified values) from the start and end of the string.
• Added: FindChar fast brute force find a single character.
• Added: FindChars fast brute force find for a substring.
• Added: ColorToHex converts a Clarion RGB color stored in a long to a hex string (which can be used in CSS, HTML etc.).
• Added: ColorFromHex converts a string containing the hexadecimal representation of the color to a long containing a standard RGB Clarion color.
• Added: Added color conversion code to the stdemo example app.
• Added: UrlEncode method URL (percent) encodes the passed string
• Added: UrlDecode methods decodes the passed URL (percent) encoded string.
• Added: New Encoding/Decoding tab to the example. Demonstrates URL encoding and decoding and Base64 encoding and decoding.
• Fixed: A boundary index out of range error in Replace when a match is made against the whole string.

Version 1.45 - 20 August 2011

Required for CryptoNite 1.1.8 and up

• Fixed: GPF in ToUpper in some versions of C6 as a result of the Module statement within the method.
• Updated: The Replace method is now five times faster for multi-character replacement.
• Fixed: Problems with Replace when the pEnd parameter is used.
• Updated: Removed debug output from SaveFile.
• Updated: Added the auto attribute to local variables in the new ReverseBytes method.
• Added: Checking for passed lengths less than one in the SetLength method.
• Fixed: Potential GPF in WrapText when the string is too short to be wrapped.
• Fixed: Potential GPF in Replace when an invalid string to search for was passed, such as ''.
• Fixed: Potential GPF in FileNameOnly if a zero length string was passed, or if the method was called to extract the file name from the stored string and no value had been set.

Version 1.44 - 20 August 2011

Required for CryptoNite 1.1.8 and up

• Fixed: Removed debug code that was saving a temp file when calling the AnsiToUtf16 method.

Version 1.43 - 19 August 2011

Required for CryptoNite 1.1.8 and up

• Fixed: Append method causing conflicts in Clarion 6.1 9033 with the built in Append (as with C55 the AppendA method is available rather than Append).
• Added: An encoding parameter to the ToAnsi method. This allows the encoding to be specified in the call to ToAnsi, rather than setting the .encoding property of the object before calling ToAnsi.
• Added: SwitchEndian method reverse the byte order in a long.
• Added: BigEndian - converts a little endian four byte integer to a big endian.
• Added: LittleEndian - converts a big endian four byte integer to little endian.
• Fixed: Example code in the documentation for WordStart and WordEnd was using Sub rather than Slice to return the string value found.
• Updated: Calling Split with an empty string will result in an empty Lines queue.
• Fixed: Potential GPF in Split function when used with a pQuoteStart parameter.
• Fixed: Potential GPF in Split function as the result of an incorrect string slice.
• Added: ReverseByteOrder method. Reverses the byte order of the stored string (the entire string is reversed so '123456ABC' would become 'CBA654321').
• Added: New documentation and corrected example code for the Crop method.
• Updated: Documentation for SaveFile
• Updated: Documentation for LoadFile
• Updated: Documentation for Slice
• Updated: Documentation for the .encoding property.
• Updated: Documentation for ToUnicode
• Updated: Documentation for ToAnsi
• Updated: Documentation for AnsiToUtf16
• Updated: Documentation for Utf16ToAnsi
• Updated: Documentation for Utf8To16
• Updated: Documentation for Utf16To8
• Updated: Documentation for Utf8ToAnsi
• Updated: Documentation for AnsiToUtf8
• Updated: Documentation for ByteToHex
• Updated: Documentation for HexToByte
• Updated: Documentation for LongToHex

Version 1.42 - 02 August 2011

Required for CryptoNite 1.1.8 and up

• Fixed: WordStart method not allowing a startPos equal to the string length.
• Updated: Removed unused local variable from SaveFile and WordEnd
• Updated: Fixed incorrect spelling in comment
• Fixed: After method not correctly returning an empty string when the search string was located at the very end of the string.
• Updated: Changed all internal call to Len to call LengthA directly instead for slightly improved efficiency.
• Updated: The documentation for the Count method to include the new softClip parameter.
• Updated: The default value of the softClip parameter for the Count method to True. Note that this does not change the behavior of the method, the change of the new parameter default value is for clarity.

Version 1.41 - 27 July 2011

Required for CryptoNite 1.1.8 and up

• Updated: Documentation for the RemoveAttributes method.
• Updated: Documentation for the Replace method.
• Added: Documentation for the GetBytes methods.
• Added: Documentation for the SetBytes methods.
• Added: Documentation for the FromBytes methods.
• Added: Document for the ToBytes methods.
• Added: GetWord Procedure(long pWordNumber,long startPos = 1, bool htmlEnc=false), string, virtual. Returns a string containing the word at the passed position.
• Added: Clip Procedure(), virtual. Clips the string (removes trailing white space).
• Added: Trim Procedure(), virtual. Trims the string (removes white space from the start and the end of the string).
• Added: Squeeze Procedure(bool htmlEnc=false), virtual. Removes excess white space (each word in the string is separated by a single space).
• Added: ContainsADigit Procedure(), bool, virtual. Returns True if the string contains at least one digit character ('0' to '9')
• Added: IsAllDigits Procedure(), bool, virtual. Returns True if the string contains all digit characters ('0' to '9')
• Updated: Optimized the SetValue method
• Fixed: The SetBlob method was not returning a value on success
• Updated: Removed extraneous string slice in ToBlob
• Updated: Removed extraneous length assignment in Base64Encode
• Updated: Added the Auto attribute to values that did not require automatic initialization to improve performance throughout
• Updated: Removed unused local variables from methods
• Updated: Removed extraneous if statement from PathOnly
• Updated: Refactored SetLength
• Updated: Refactored Replace
• Updated: Refactored Count to optimize performance
• Added: The Count method now supports "soft" clipping using the softClip parameter. This allows the search to overlap the passed pEnd if the start of the string is within the bounds passed. This defaults to False (clipping is "hard" and the word must be entirely within the pStart and pEnd bounds passed to be counted).
• Updated: Refactored the LoadFile method and set the path length to the new st:MAX_PATH equate
• Updated: Refactored the SaveFile method and set the path length to the new st:MAX_PATH equate
• Updated: Removed extraneous local string variable and local assignment from FormatMessage
• Updated: Removed debug Trace statements from AnsiToUtf16, removed unused locals
• Updated: Refactored BaseToDec, optimized it and added bounds checking
• Updated: Refactored StrCpy to remove extraneous checks and slicing.
• Fixed: Lower bounds checking bug in WordStart (allowed a startPos of 0 rather than 1 to be passed)
• Updated: Refactored and optimized WordEnd.
• Fixed: WordEnd not correctly handling apostrophes because of the order that whitespace checks were being done
• Updated: Removed extraneous call to Length() from CountWords
• Fixed: Changed Length() to LengthA() in RemoveAttributes to maintain C55 compatibility.
• Updated: Documentation for the Split method
• Fixed: Missing anchors in the documentation.
  
• Many thanks to Geoff Robinson for feedback and contributions to StringTheory, including bug fixes, improvements and new methods!

Version 1.40 - 18 July 2011

Required for CryptoNite 1.1.8 and up

• Added: RemoveAttributes method added.

Version 1.39 - 07 June 2011

Required for CryptoNite 1.1.8 and up

• Fixed: Bug in the Instring when search string was found at the end of the string being searched zero would be returned as Instring was incorrectly detecting the character as out of range. This also affected methods that called Instring, such as Between and FindBetween.
• Fixed: CountWords could loop infinitely
• Added CountWords, WordStart and WordEnd to the example application.
• Fixed: CountWords, WordStart and WordEnd were not listed in the version History
• Added: Documentation for new methods.

Version 1.38 - 06 June 2011

Required for CryptoNite 1.1.8 and up

• Fixed: Bug in the Between and FindBetween methods that was searching the wrong range.
• Fixed: Potential for an infinite loop when using the pRecursive parameter of the Replace method where the string being replace was a substring of the replacement string and the pCount parameter was not used to limit the number of replacements performance.
• Fixed: The pCount parameter of the Replace method not working correctly (only a single instances was replaced when pCount was specified).
• Added: Examples of Between and FindBetween to the example application.
• Added: Between and FindBetween to the class documentation.
• Added: CountWords method - count the number of words in the string.
• Added: WordStart and WordEnd methods that return the start and end of the next word in the string given a start position.

Version 1.37 - 03 June 2011

Required for CryptoNite 1.1.8 and up

• New: Added methods to allow binary storage of a wide variety of data types. This allows data to be stored and retrieved as bytes. For example when encrypting a long or a decimal, the data can be stored as bytes for encryption and then retrieved as bytes after decryption:
  • ToBytes and FromBytes store and retrieve data using a string to and from: byte, short, ushort, long, ulong, sreal, real, decimal, cstring, string
  • SetBytes and GetBytes store the passed data in the StringTheory object as bytes and retrieve the original values for the following data types: byte, short, ushort, long, ulong, sreal, real, decimal, cstring, string
  • Generic methods to allow storage and retrieval of ANY data types
    • SetBytes Procedure (*? pVal, string pType), bool, proc, virtual
    • GetBytes Procedure (*? pVal, string pType), bool, proc, virtual
• Updated: Further optimization to the Replace method for single character searches.
• Added: pRecursive parameter to the Replace method. When set to True the replace is done in a recursive manner - so the start position is repeated to allow replacements within the newly replaced text.
• Added: FindBetween method. Identical to the Between method except that the passed pStart and pEnd parameters are updated with the start and the end positions of the string being returned.

Version 1.36 - 5 May 2011

• Update: Template changes for Clarion 8 support.

Version 1.35 - 15 April 2011

Required for CryptoNite 1.15 and up

• Fixed: The Replace method was not handling case insensitive find/replace for single character replacements.
• Changed: The Instring method pEnd parameter now specifies the end of the "slice" if the string to search. Previously pEnd was actually specifying the final start position of any string located, so a string could be found that started before pEnd, but ended after it. pStart and pEnd now specify the start and end of the string to search, and default to the start and end of the string passed if not specified.

Version 1.34 - 15 April 2011

Required for CryptoNite 1.15 and up

• Updated for SafeUpdate 2.0. This is the minimum version required for SafeUpdate 2.0.

Version 1.33 - 29 March 2011

Required for CryptoNite 1.12 and up

• Fixed: The Replace method was ignoring the pStart and pEnd parameters when doing a single character replacement.
• Added: Additional bounds checking to the Replace method for the starting position.

Version 1.32 - 19 January 2011

Required for CryptoNite 1.12 and up

• Fixed: Template error - missing #EndTab statement

Version 1.31 - 18 January 2011

Required for CryptoNite 1.04 and up

• Fixed: The Crop method not cropping the string when the pEnd parameter was omitted

Version 1.30 - 21 December 2010

Required for CryptoNite 1.04

• Fixed: GPF when converting ANSI to UTF-8.
• Important: The Ansi to Unicode conversion methods now take pointers to strings rather than cstrings. This is primarily because the strings being passed do not need to be terminated, and the null terminate would differ in length for different Unicode encodings, which can cause issues when the string is then stored as a Clarion cstring. This does not affect the use of the ToUnicode and ToAnsi methods, however will affect directly calls to the other conversion. methods including Utf8ToAnsi, AnsiToUtf8, AnsiToUtf16, Utf16ToAnsi, Utf8To16, Utf16To8, apiUtf8ToAnsi and apiAnsiToUtf8. You should adjust your code accordingly.
• Fixed: Error in the history where a section of the CryptoNite history had been included.
• Fixed: Missing history in the docs.
• Updated the example application to explicitly set the encoding when converting between UTF-8 and ANSI. When ToUnicode is called the StringTheory.encoding property is set, so when ToAnsi is called the method can determine the type of encoding, convert it and set .encoding to st:EncodeAnsi. When Unicode text is initially stored or loaded from a file the .encoding property is not set, so for example calling ToAnsi() would have no effect unless .encoding is first set to st:EncodeUtf8 or st:EncodeUtf16.

Version 1.29 - 10 December 2010

Required for CryptoNite 1.11

• Changed: Removed the automatic derivation of methods by the template.
• Fixed: Template derivation preventing the MD5 functionality from being disabled
• Fixed: Duplicate symbols with NetTalk 5.
• Fixed: GPF in the ToAnsi and ToUnicode wrapper methods.
• Added: ByteToHex Procedure (byte pByte). Converts a byte to a two character string that contains the hexidecimal equivelent of the byte value. For example a byte that contains 255 would return 'ff'.
• Added: HexToString Procedure (*string hexData). This method returns a pointer to a string that contains the original value of the hex encoded input.
• Added: FromHex Procedure. This method converts a hex encoded string back to the original value.
• Added: StrCpy procedure. This performs a safe string copy from a source string to a destination string. Handles strings of any length, and truncates the data stored in the destination string if the source string contains more data than the destination can store.

Version 1.28 - 27 November 2010

• New:
  • WrapText - performs word wrapping to wrap the text to a fixed width by breaking lines as close the the specified width as possible. Lines are only broken on white space (tabs and spaces). The default wrap width is 80 characters.
  • Updated the example with the new WrapText method

• Fixed
  • DecToBase - for bases below 10 the returned string could be truncated for larger numbers (maximum return length was 10).
  • Compile error on the WrapText method, which had not been implemented.

Version 1.27 - 27 October 2010

• New Class methods:
  • ToAnsi
    Converts the current UTF-8 or UTF-16 string to ANSI
  • ToUnicode
    Converts the current ANSI string to either UTF-8 or UTF-16
    
    Additional conversion methods (encapsulated in ToAnsi and ToUnicode)
    
  • AnsiToUtf16
  • Utf16ToAnsi
  • Utf8To16
  • Utf16To8
  • Utf8ToAnsi
  • AnsiToUtf8
    
• New class properties
  • encoding - Text encoding - Ansi, UTf-8 or Utf16
  • widechars - When a string is converted to UTF-16, this is set to the number of wide chars the string contains
    
    
• Renamed methods:
  • apiUtf8ToAnsi becomes Utf8ToAnsi
  • apiAnsiToUtf8 becomes AnsiToUtf8
    
• Fixed memory leak in Utf8ToAnsi and AnsiToUtf8 (previously named apiUtf8ToAnsi and apiAnsiToUtf8)
• Updated version history for 1.26 to include descriptions of the new methods, and links to the method documentation
• Added new methods from 1.26 and 1.27 to the class documentation.
• Fixed: The "noCase" option for find and replace in the demo wasn't being passed to the Replace method.
• Fixed: Between now allows spaces to be passed as the delimiters.

Version 1.26 - 15 June 2010

• Correct 1.25 release date in the version history
• Added new methods:
  • DecToBase Procedure (long num, long base=16, long lowerCase=1), string, virtual
    Converts a decimal number to a number using the specified base (such as Hexadecimal - base 16).
  • BaseToDec Procedure (string num, long base=16), long, virtual
    Converts the passed number stored in a string to a decimal.
  • ByteToHex Procedure (byte pByte), string, virtual
    Returns a string that contains the hexadecimal equivilent of the passed byte.
  • LongToHex Procedure (long pLong), string, virtual
    Converts a long of 4 bytes to hex. Returns a string that contains the hex of each byte in big endian order (the most significant byte first).
  • StringToHex Procedure (*string binData), *string, virtual
    Returns a string that contains the hexadecimal equivilent of each byte in the passed string (which is treated as binary data).
    For example:
    'Hello World' would be encoded to: '48656c6c6f20576f726c64'
  • ToHex Procedure (), virtual
    Converts the stored string to a hex string - each bytes is converted to two characters representing the hexidecimal equivilent.
    For example:
    'Hello World' would be encoded to: '48656c6c6f20576f726c64'

• Initial Release of the stand alone StringTheory
  
• Changes from prior versions included in OddJob:
  • Fixed Between() method including the left hand delimiter in the returned string.
  • Additional handling of boundary cases in the Replace() method.
  • Fixed a bug in the Replace method, when the string being replaced is the end of the text being searched.
  
  

Version 1.24 and prior (not available as a stand alone product) - 15 June 2010

• Added bounds checking and corrected unhandled boundary cases for the following methods:
  • Replace (handles replacement from the start and end of the string, where the resultant string is empty, where the resultant string is a single character etc.)
  • PathOnly (handles the case where the path contains only a forward or back slash)
  • Split (handles if a "line" is blank)
  • ToBlob (handles the string being assigned to the blob being empty, or being a single character)
  • Verified that Between and Before are both safe.
  • Added optional clip parameter to StringTheory Append method.
  • Updated docs for the StringTheory SetValue method
  • Added handling to SetValue for passing zero length strings (the StringTheory value is disposed of and the Length methods will return zero).
  • New StringTheory methods:
    • Before - returns the portion of the string that occurs before the specified string seperator.
    • After - returns the portion of the string that occurs after the specified string seperator.
    • Between  - returns the portion of the string that occurs between the two specified string seperators.
    • Base64Encode/Base64Decode - Base64 encode and decode either a passed string, or the value stored by the StringTheory object.
    • MD5 - Create an MD5 hash of the string.
    • New string delimiter processing methods:
      • Split - splits the string into sub strings, which are stored in the .lines property. Optionally allows the "line" start and end delimeters to be specified.
      • Join - joins the all the lines (after the string has been slit using Split, and optionally wraps each line in the specified delimiters).
    • Slice - returns a string containing the data between the specified start and end characters (a "slice").
    • Crop - Crops the stored value using the pass start and end parameters.
    • Sub - returns a sub-string of the specified length, starting at the specified character.
  • Fixed the ClipLen method infinitely loop as it was calling itself rather than calling ClipLength().
  • Fixed the Base64Encode method adding extra spaces in the encoded string.
  • Added: FromBlob method, stores the contents of the passed BLOB in the StringTheory object.
  • Added ToBlob method, saves the contents of the string into the passed BLOB.
  • Fixed: the PathOnly method, which was returning a blank string.
  • Optimised: the Replace method in the case where both strings are a single character
  • Fixed: Replace to allow an empty replacement string to be specified (which removes the searched for string)
  • Fixed: FileNameFromPath method.