You can use the Microsoft API function CryptBinaryToStringW.

Function Boolean CryptBinaryToString ( &
blob pbBinary, &
ulong cbBinary, &
ulong dwFlags, &
Ref string pszString, &
Ref ulong pcchString) Library "Crypt32.dll" Alias For "CryptBinaryToStringW"



CONSTANT ulong CRYPT_STRING_BASE64 = 1
CONSTANT ulong CRYPT_STRING_NOCRLF = 1073741824 // 0x40000000 - not for windows XP!
uLong 	lul_buflen
Blob 	lblob_value
string 	ls_base64

lblob_value = blob ("your blob here")
ls_base64 = Space (2 + 1.5 * (Len (lblob_value) + 2))
lul_buflen = Len (ls_base64)
If Not CryptBinaryToString (lblob_value, Len (lblob_value), CRYPT_STRING_BASE64 + CRYPT_STRING_NOCRLF, ls_base64, lul_buflen) Then
   // error
End If

MessageBox ("", ls_base64)

Hi, Aymen - 

I suggest you download and check out the CryptoAPI free code example at Roland Smith's TopWizProgramming web site:

https://topwizprogramming.com/freecode_cryptoapi.html

All of the free code samples are provided in both PB 10.5 and PB 12.6 (and higher) versions. Let me know how that works out for you.

Regards, John

[Edit: Replaced function source code with improved version that uses fewer variables and memory.]

Here's a script for a PowerBuilder function I created that performs Base64 encoding that should be compatible with PB 10:

// Public Function: of_Base64Encode(String as_Input) returns String

Constant String BASE64_ENCODING_MAP = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/"

Blob    lblob_input
Byte    lbyte_char
Integer li_rc, li_six_bit_value[4]
Long    ll_accumulator, ll_index, ll_blob_length
String  ls_encoded_token, ls_output

// Check for null or empty argument string.
If IsNull(as_input) Then
   SetNull(ls_output)
   Return ls_output
End If
If Len(as_input) = 0 Then Return ""

// Place the input string into a Blob as a series of UTF-8 characters.
// The base64 encoding process encodes thres bytes (characters) of the blob at a time,
// because three bytes = 24 bits, which is evenly divisible by six (bits), producing
// an encoded token of four characters.
lblob_input = Blob(as_input,EncodingUTF8!)
ll_blob_length = Len(lblob_input)

// Encode the blob, up to three bytes (24 bits) at a time.
ls_output = ""
For ll_index = 1 To ll_blob_length Step 3
   
   // Accumulate the next three UTF-8 characters (24 bits) into a Long,
   // then separate the bit pattern into four six-bit values.
   li_rc = GetByte(lblob_input,ll_index,lbyte_char)
   
   ll_accumulator = lbyte_char * 256
   
   If ll_index+1 <= ll_blob_length Then
      li_rc = GetByte(lblob_input,ll_index+1,lbyte_char)
      ll_accumulator += lbyte_char
   End If
   
   ll_accumulator *= 256
   
   If ll_index+2 <= ll_blob_length Then
      li_rc = GetByte(lblob_input,ll_index+2,lbyte_char)
      ll_accumulator += lbyte_char
   End If
   
   // The low-order 24 bits of the Long now contain the next three UTF-8 characters.
   // Split these 24 bits into four six-bit values.
   li_six_bit_value[4] = Mod(ll_accumulator,64)
   ll_accumulator /= 64
   li_six_bit_value[3] = Mod(ll_accumulator,64)
   ll_accumulator /= 64
   li_six_bit_value[2] = Mod(ll_accumulator,64)
   li_six_bit_value[1] = ll_accumulator / 64
   
   // Map the four six-bit values into the base64 encoding map.
   // If needed, "padding" values are translated into equal signs (=).
   ls_encoded_token = Mid(BASE64_ENCODING_MAP,li_six_bit_value[1]+1,1) + &
                      Mid(BASE64_ENCODING_MAP,li_six_bit_value[2]+1,1)
   If ll_index+1 <= ll_blob_length Then
      // Two chars (16 bits) are represented by three six-bit values.
      ls_encoded_token += Mid(BASE64_ENCODING_MAP,li_six_bit_value[3]+1,1)
   Else
      // There was only one character, so append a "pad" encoding character.
      ls_encoded_token += "="
   End If
   If ll_index+2 <= ll_blob_length Then
      // Three chars (all 24 bits) are represented by all four six-bit values.
      ls_encoded_token += Mid(BASE64_ENCODING_MAP,li_six_bit_value[4]+1,1)
   Else
      // There were only one or two characters, so append a "pad" encoding character.
      ls_encoded_token += "="
   End If
   
   // Append the Base64-encoded token onto the output string, and repeat.
   ls_output += ls_encoded_token
   
Next

// Done!
Return ls_output

HTH, John

A little more information, please.

Hex in what form? Binary (in a Blob variable) or in a String ("ff018c47", for example)? How much data? Are you wanting to produce a character string that contains the base64-encoded equivalent?

hello john 

yes i want to encode utf-8 hex string to base64 string . data is about 200 character more or less .

Aymen

Have you checked Chilkat (https://www.example-code.com/powerbuilder/default.asp)?  They have lot of activeX, may be there one which you need.  Those (all) are not free though.

Happiness Always
BKR Sivaprakash

Dear Rene

if you go to https://cryptii.com/ 

the encoding of :

0115426f627320426173656d656e74205265636f726473020f3130303032353930363730303030330314323032322d30342d32355431353a33303a30305a040a323130303130302e393905093331353031352e3135

is :

ARVCb2JzIEJhc2VtZW50IFJlY29yZHMCDzEwMDAyNTkwNjcwMDAwMwMUMjAyMi0wNC0yNVQxNTozMDowMFoECjIxMDAxMDAuOTkFCTMxNTAxNS4xNQ==

while windows Crypt32.dll ( the one you posted ) gives :

MDExNTQyNmY2MjczMjA0MjYxNzA2NTZkNjU2ZTc0MjA1MjY1NjM2ZjcyNjQ3MzAyMGYzMTMwMzAzMDMyMzUzOTMwMzYzNzMwMzAzMDMwMzMwMzE0MzIzMDMyMmQzMDM0MmQ

and the  correct one is the one in dark blue ( cryptii.com )

Dear John 

what i am working with is Arabic characters , your string list is only English characters .