Sms with accented characters

Question

Hi,
I hope somebody with some SMS/SMPP experience would be able to fix the problem or
advise as what to do. OK Now we have an application which is supposed to send an
SMS to a recipient on certain occassions and this SMS might contain accented characters.
Now in SMPP apparently we have three encoding schemes namely
1) 7 bit GSM encoding
2) 8 bit binary
3) UCS-2
Now the third one is out of question since it would reduce the size of the message
to 70 characters which is not acceptable. 2nd one which is 8 bit binary with I assume
DCS value 4(or at least thats what we are using) seems to support only latin characters.
It works for latin characters i.e., iso-8859-1 which does have some accented characters.
however for other than latin we have some problems e.g hex value for small phi in
greek (iso-8859-7) is f6. If you look at the corresponding value of f6 in Latin 1(iso-8859-1)
it is �. Now if you send greek characters setting the DCS value to 4 and u send three
phi's what you would recieve on the recipient device are three �. So it seems that
the 8 bit binary supports only iso-8859-1.
Now if I try to do some encoding into GSM 7 bit. and send three phi's thought I dont
recieve � but not phi's at the recieving end. The DCS value for the purpose I am
using is 0. How am I encoding it to GSM 7 bit.
I am creating the following array
private static final char [] GSM_7Bit_CHARACTERS = {
'@', '\243', '$', '\245', '\350', '\351', '\371', '\354', '\362', '\307',
'\n', '\330', '\370', '\r', '\305', '\345', '\u0394', '_', '\u03A6', '\u0393',
'\u039B', '\u03A9', '\u03A0', '\u03A8', '\u03A3', '\u0398', '\u039E', '\033',
'\306', '\346',
'\337', '\311', ' ', '!', '"', '#', '\244', '%', '&', '\'',
'(', ')', '*', '+', ',', '-', '.', '/', '0', '1',
'2', '3', '4', '5', '6', '7', '8', '9', ':', ';',
'<', '=', '>', '?', '\241', 'A', 'B', 'C', 'D', 'E',
'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O',
'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y',
'Z', '\304', '\326', '\321', '\334', '\247', '\277', 'a', 'b', 'c',
'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm',
'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', 'w',
'x', 'y', 'z', '\344', '\366', '\361', '\374', '\340'
Storing all the values in a Map in teh same order and then for each character in the
message sending the byte value of the index. If I do a snoop in the snoop I see the
right hex values as per GSM Encoding table but on the recipient device I see garbage.
Any ideas as to how can greek and other european character set support be added to
this application?
Am I using the wrong DCS value or any body with some other ideas.

Answer 1

I don't know about the exact character sets for the 7 and 8-bit SMS, but I doubt it's any different then character sets for anything else... So, in this case, the 7-bit set will only be able to handle the first 128 chars, which is A-Za-z0-9 and some punctuation. 8-bit would go into the extended ASCII chars in the table here:
http://www.asciitable.com/
There might be some restrictions or differences.
Basically, 7-bit is ASCII. 8-bit is ASCII+Extended or ISO8895-1. The chars your talking about sending get into the other ISO8895-X sets, which I believe would be out of range. The USC-2 should give you these, however. So either you send shorter messages, or you send ASCII text.
It would also depeend on the device to be able to display those characters in the first place. Just like on any web site, if you don't have a font that has the necessary characters, it just displays as ? or a box or whatever.