Home > Language Reference > Functions

Strings

The String data type in conjunction with the functions described in this section allows you to manipulate character strings. It should be remembered that all operations on strings are costly both in terms of memory usage and calculation time; their use should be reserved for cases where they are necessary. Notably, strings should not be used to store or compare numeric values or dates.

Representation of strings

Basically, strings are blocks of bytes. However, how these bytes are interpreted depends on the code page of the device your application is running on. The following table summarizes this:

HB++ also checks for the PiLoc utility, a small program you can use to display non latin scripts such as cyrillic or greek on your Palm OS device. In case this utility was found running, HB++ uses the character encoding the user specified in the PiLoc's configuration panel.

Because different devices may use different character encodings, the same block of data may represent different actual strings. For example, a string containing the two bytes &hCD and &hC1 will be interpreted as ÍÁ on a Western device and as 土 on a Far East device. This may cause troubles when beaming data from a device to another, when synchronizing data with a desktop computer, or when sending data to a web service. The Utf8 function may be used to convert a string back and forth between the current encoding and unicode.

While the device relies on a specific code page, the HB++ IDE uses unicode. This means your source code can virtually contain any character from any character set. When compiling, HB++ uses localisation information to encode with the proper code page the strings it encounters in your source code. For more information about code page selection and about writing portable applications, refer to the Localisation section in the Programmer's Guide.

You can use the Asc and Chr functions to determine respectively the code of a given character and the character corresponding to a given code. Because both Palm OS^® and HB++ have to cooperate with external functions written in C, all character strings are internally null terminated. This implies a null character (a character whose code is zero) cannot be embeded into a string.

The maximum length of a string is limited to 32767 bytes. Building a longer string results in a runtime error. The Len function determines the number of characters in a string, while the LenB function returns the number of bytes a string occupies.

Manipulating strings

The comparison of two strings is carried out using the regular relational operators. Comparison is case sensitive, that is to say it takes upper and lower case into account. The order of two given characters is that of the respective codes of the two characters.

Sometimes, it is necessary to carry out a comparison that is case insensitive, ie not differentiate between upper and lower case. You can use the function StrComp which accepts a parameter that allows you to specify the comparison mode.

Concatenation of strings can be performed using the operator &.

The majority of regular functions for manipulating character strings are available. You can refer to the detailed list below for more information.

Converting numeric values

It is often useful to convert a character string entered into a text field into a numeric value, or conversely to convert a numeric value into a character string to display on screen.

The type conversion function CStr allows the conversion of numeric values into strings. The resulting string does not contain thousands separator, and the decimal separator is the decimal point. If you want to take into account regional settings preferences, you should use the function Format.

Conversely, if you want to convert a string containing a number into a numeric value, you can use the type conversion functions CByte, CInt, CLng, CSng, and CDbl. The string should not contain thousands separators and the decimal separator must be a decimal point. The function DelocalizeNumber returns a string without thousands separators or decimal points from a string formatted using the regional settings of the system.

If the string does not contain a number, a runtime error will occur. To avoid this, you should test the validity of the string in advance using the function IsNumeric. The following example shows how to extract a numeric value from a variable of type Double entered by the user into a text field.

Private Sub Button1_Click()
  Dim s As String, d As Double

  s = DelocalizeNumber(Field1.Text)
  if IsNumeric(s) Then
    d = CDbl(s)
    MsgBox "The value entered is : " & CStr(d)
  Else
    MsgBox "The string entered is not a number"
  End If
End Sub

When you want to convert a whole number, of which the range of validity is more narrow than that of type Double, you should check for overflow. A more efficient method that avoids explicitly testing is to use error handling:

Private Sub Button1_Click()
  Dim i As Integer

  On Error Goto ERR
  i = CInt(DelocalizeNumber(Field1.Text))
  MsgBox "The value entered is : " & CStr(i)
  Exit Sub

ERR:
  MsgBox "The string entered is not a number"
End Sub

It is also possible to convert whole numbers into their binary, octal or hexadecimal values using the functions Bin, Oct, and Hex. Converting binary, octal or hexadecimal values back is possible using the type conversion functions CByte, CInt, and CLng. Refer to the relevant pages for more information.

Reference

The functions described in this chapter are listed below.