FreeCalypso > hg > freecalypso-tools

comparison uptools/libcoding/gsm7_decode.c @ 802:1c599681fd60

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
pcm-sms-decode & sms-pdu-decode: revamp bad char decoding
author Mychaela Falconia <falcon@freecalypso.org>
date Thu, 25 Mar 2021 02:58:30 +0000
parents 978571e23318
children 5637794913a8
comparison
equal deleted inserted replaced
801:da724c67159d 802:1c599681fd60
2 * This library module implements the decoding of GSM7-encoded data 2 * This library module implements the decoding of GSM7-encoded data
3 * into ASCII, ISO 8859-1 or UTF-8. 3 * into ASCII, ISO 8859-1 or UTF-8.
4 */ 4 */
5 5
6 #include <sys/types.h> 6 #include <sys/types.h>
7 #include <stdio.h>
7 8
8 extern u_short gsm7_decode_table[128]; 9 extern u_short gsm7_decode_table[128];
9 extern u_short gsm7ext_decode_table[128]; 10 extern u_short gsm7ext_decode_table[128];
10 11
11 gsm7_to_ascii_or_ext(inbuf, inlen, outbuf, outlenp, ascii_ext, newline_ok, errp) 12 gsm7_to_ascii_or_ext(inbuf, inlen, outbuf, outlenp, ascii_ext, newline_ok, errp)
12 u_char *inbuf, *outbuf; 13 u_char *inbuf, *outbuf;
13 unsigned inlen, *outlenp, *errp; 14 unsigned inlen, *outlenp, *errp;
14 { 15 {
15 u_char *inp, *endp, *outp; 16 u_char *inp, *endp, *outp;
16 unsigned errcnt = 0;
17 unsigned gsm, uni; 17 unsigned gsm, uni;
18 int is_ext;
18 19
19 inp = inbuf; 20 inp = inbuf;
20 endp = inbuf + inlen; 21 endp = inbuf + inlen;
21 outp = outbuf; 22 outp = outbuf;
22 while (inp < endp) { 23 while (inp < endp) {
23 gsm = *inp++; 24 gsm = *inp++;
24 if (gsm == 0x1B && inp < endp) 25 if (gsm == 0x1B && inp < endp && *inp != 0x1B && *inp != '\n'
25 uni = gsm7ext_decode_table[*inp++]; 26 && *inp != '\r') {
26 else 27 gsm = *inp++;
28 uni = gsm7ext_decode_table[gsm];
29 if (uni == '\\') {
30 *outp++ = '\\';
31 *outp++ = '\\';
32 continue;
33 }
34 is_ext = 1;
35 } else {
36 switch (gsm) {
37 case 0x1B:
38 *outp++ = '\\';
39 *outp++ = 'e';
40 continue;
41 case '\n':
42 if (newline_ok)
43 *outp++ = '\n';
44 else {
45 *outp++ = '\\';
46 *outp++ = 'n';
47 }
48 continue;
49 case '\r':
50 *outp++ = '\\';
51 *outp++ = 'r';
52 continue;
53 }
27 uni = gsm7_decode_table[gsm]; 54 uni = gsm7_decode_table[gsm];
28 if (uni == '\r') { 55 is_ext = 0;
29 *outp++ = '\\'; 56 }
30 *outp++ = 'r'; 57 if (!uni || !is_decoded_char_ok(uni, ascii_ext)) {
31 errcnt++; 58 if (is_ext) {
32 } else if (uni == '\n') {
33 if (newline_ok)
34 *outp++ = '\n';
35 else {
36 *outp++ = '\\'; 59 *outp++ = '\\';
37 *outp++ = 'n'; 60 *outp++ = 'e';
38 errcnt++;
39 } 61 }
40 } else if (!uni || !is_decoded_char_ok(uni, ascii_ext)) { 62 sprintf(outp, "\\%02X", gsm);
41 *outp++ = '?'; 63 outp += 3;
42 errcnt++;
43 } else if (ascii_ext == 2) 64 } else if (ascii_ext == 2)
44 outp += emit_utf8_char(uni, outp); 65 outp += emit_utf8_char(uni, outp);
45 else 66 else
46 *outp++ = uni; 67 *outp++ = uni;
47 } 68 }
48 *outp = '\0'; 69 *outp = '\0';
49 if (outlenp) 70 if (outlenp)
50 *outlenp = outp - outbuf; 71 *outlenp = outp - outbuf;
51 if (errp) 72 if (errp)
52 *errp = errcnt; 73 *errp = 0;
53 } 74 }