encoding.c: Support the ISO 8859-2 encoding.

src/include/espeak-ng/espeak_ng.h

@@ -191,6 +191,7 @@ typedef enum
 	ESPEAKNG_ENCODING_UNKNOWN,
 	ESPEAKNG_ENCODING_US_ASCII,
 	ESPEAKNG_ENCODING_ISO_8859_1,
+	ESPEAKNG_ENCODING_ISO_8859_2,
 } espeak_ng_ENCODING;
 
 ESPEAK_NG_API espeak_ng_ENCODING

src/libespeak-ng/encoding.c

@@ -33,18 +33,25 @@ MNEM_TAB mnem_encoding[] = {
 	{ "IBM819",           ESPEAKNG_ENCODING_ISO_8859_1 },
 	{ "ISO_646.irv:1991", ESPEAKNG_ENCODING_US_ASCII },
 	{ "ISO_8859-1",       ESPEAKNG_ENCODING_ISO_8859_1 },
+	{ "ISO_8859-2",       ESPEAKNG_ENCODING_ISO_8859_2 },
 	{ "ISO646-US",        ESPEAKNG_ENCODING_US_ASCII },
 	{ "ISO-8859-1",       ESPEAKNG_ENCODING_ISO_8859_1 },
 	{ "ISO-8859-1:1987",  ESPEAKNG_ENCODING_ISO_8859_1 },
+	{ "ISO-8859-2",       ESPEAKNG_ENCODING_ISO_8859_2 },
+	{ "ISO-8859-2:1987",  ESPEAKNG_ENCODING_ISO_8859_2 },
 	{ "US-ASCII",         ESPEAKNG_ENCODING_US_ASCII },
 	{ "cp367",            ESPEAKNG_ENCODING_US_ASCII },
 	{ "cp819",            ESPEAKNG_ENCODING_ISO_8859_1 },
 	{ "csASCII",          ESPEAKNG_ENCODING_US_ASCII },
 	{ "csISOLatin1",      ESPEAKNG_ENCODING_ISO_8859_1 },
+	{ "csISOLatin2",      ESPEAKNG_ENCODING_ISO_8859_2 },
 	{ "iso-ir-6",         ESPEAKNG_ENCODING_US_ASCII },
 	{ "iso-ir-100",       ESPEAKNG_ENCODING_ISO_8859_1 },
+	{ "iso-ir-101",       ESPEAKNG_ENCODING_ISO_8859_2 },
 	{ "latin1",           ESPEAKNG_ENCODING_ISO_8859_1 },
+	{ "latin2",           ESPEAKNG_ENCODING_ISO_8859_2 },
 	{ "l1",               ESPEAKNG_ENCODING_ISO_8859_1 },
+	{ "l2",               ESPEAKNG_ENCODING_ISO_8859_2 },
 	{ "us",               ESPEAKNG_ENCODING_US_ASCII },
 	{ NULL,               ESPEAKNG_ENCODING_UNKNOWN }
 };
@@ -65,6 +72,28 @@ struct espeak_ng_TEXT_DECODER_
 	const char *end;
 
 	uint32_t (*get)(espeak_ng_TEXT_DECODER *decoder);
+	uint32_t *codepage;
+};
+
+// Reference: http://www.iana.org/go/rfc1345
+// Reference: http://www.unicode.org/Public/MAPPINGS/ISO8859/8859-2.TXT
+static const uint32_t ISO_8859_2[0x80] = {
+	0x0080, 0x0081, 0x0082, 0x0083, 0x0084, 0x0085, 0x0086, 0x0087, // a0
+	0x0088, 0x0089, 0x008a, 0x008b, 0x008c, 0x008d, 0x008e, 0x008f, // a8
+	0x0090, 0x0091, 0x0092, 0x0093, 0x0094, 0x0095, 0x0096, 0x0097, // b0
+	0x0098, 0x0099, 0x009a, 0x009b, 0x009c, 0x009d, 0x009e, 0x009f, // b8
+	0x00a0, 0x0104, 0x02d8, 0x0141, 0x00a4, 0x013d, 0x015a, 0x00a7, // a0
+	0x00a8, 0x0160, 0x015e, 0x0164, 0x0179, 0x00ad, 0x017d, 0x017b, // a8
+	0x00b0, 0x0105, 0x02db, 0x0142, 0x00b4, 0x013e, 0x015b, 0x02c7, // b0
+	0x00b8, 0x0161, 0x015f, 0x0165, 0x017a, 0x02dd, 0x017e, 0x017c, // b8
+	0x0154, 0x00c1, 0x00c2, 0x0102, 0x00c4, 0x0139, 0x0106, 0x00c7, // c0
+	0x010c, 0x00c9, 0x0118, 0x00cb, 0x011a, 0x00cd, 0x00ce, 0x010e, // c8
+	0x0110, 0x0143, 0x0147, 0x00d3, 0x00d4, 0x0150, 0x00d6, 0x00d7, // d0
+	0x0158, 0x016e, 0x00da, 0x0170, 0x00dc, 0x00dd, 0x0162, 0x00df, // d8
+	0x0155, 0x00e1, 0x00e2, 0x0103, 0x00e4, 0x013a, 0x0107, 0x00e7, // e0
+	0x010d, 0x00e9, 0x0119, 0x00eb, 0x011b, 0x00ed, 0x00ee, 0x010f, // e8
+	0x0111, 0x0144, 0x0148, 0x00f3, 0x00f4, 0x0151, 0x00f6, 0x00f7, // f0
+	0x0159, 0x016f, 0x00fa, 0x0171, 0x00fc, 0x00fd, 0x0163, 0x02d9, // f8
 };
 
 static uint32_t
@@ -82,6 +111,13 @@ text_decoder_getc_iso_8859_1(espeak_ng_TEXT_DECODER *decoder)
 	return *decoder->current++ & 0xFF;
 }
 
+static uint32_t
+text_decoder_getc_codepage(espeak_ng_TEXT_DECODER *decoder)
+{
+	uint8_t c = *decoder->current++ & 0xFF;
+	return (c >= 0x80) ? decoder->codepage[c - 0x80] : c;
+}
+
 espeak_ng_TEXT_DECODER *
 create_text_decoder(void)
 {
@@ -91,6 +127,7 @@ create_text_decoder(void)
 	decoder->current = NULL;
 	decoder->end = NULL;
 	decoder->get = NULL;
+	decoder->codepage = NULL;
 	return decoder;
 }
 
@@ -108,9 +145,15 @@ initialize_encoding(espeak_ng_TEXT_DECODER *decoder,
 	{
 	case ESPEAKNG_ENCODING_US_ASCII:
 		decoder->get = text_decoder_getc_us_ascii;
+		decoder->codepage = NULL;
 		break;
 	case ESPEAKNG_ENCODING_ISO_8859_1:
 		decoder->get = text_decoder_getc_iso_8859_1;
+		decoder->codepage = NULL;
+		break;
+	case ESPEAKNG_ENCODING_ISO_8859_2:
+		decoder->get = text_decoder_getc_codepage;
+		decoder->codepage = ISO_8859_2;
 		break;
 	default:
 		return 0;

tests/encoding.c

@@ -124,6 +124,37 @@ test_iso_8859_1_encoding()
 	destroy_text_decoder(decoder);
 }
 
+void
+test_iso_8859_2_encoding()
+{
+	printf("testing ISO-8859-2 encoding\n");
+
+	assert(espeak_ng_EncodingFromName("ISO-8859-2:1987") == ESPEAKNG_ENCODING_ISO_8859_2);
+	assert(espeak_ng_EncodingFromName("ISO-8859-2") == ESPEAKNG_ENCODING_ISO_8859_2);
+	assert(espeak_ng_EncodingFromName("ISO_8859-2") == ESPEAKNG_ENCODING_ISO_8859_2);
+	assert(espeak_ng_EncodingFromName("iso-ir-101") == ESPEAKNG_ENCODING_ISO_8859_2);
+	assert(espeak_ng_EncodingFromName("latin2") == ESPEAKNG_ENCODING_ISO_8859_2);
+	assert(espeak_ng_EncodingFromName("l2") == ESPEAKNG_ENCODING_ISO_8859_2);
+	assert(espeak_ng_EncodingFromName("csISOLatin2") == ESPEAKNG_ENCODING_ISO_8859_2);
+
+	espeak_ng_TEXT_DECODER *decoder = create_text_decoder();
+
+	assert(text_decoder_decode_string(decoder, "aG\x92\xA0\xDE", 5, ESPEAKNG_ENCODING_ISO_8859_2) == ENS_OK);
+	assert(text_decoder_eof(decoder) == 0);
+	assert(text_decoder_getc(decoder) == 'a');
+	assert(text_decoder_eof(decoder) == 0);
+	assert(text_decoder_getc(decoder) == 'G');
+	assert(text_decoder_eof(decoder) == 0);
+	assert(text_decoder_getc(decoder) == 0x92);
+	assert(text_decoder_eof(decoder) == 0);
+	assert(text_decoder_getc(decoder) == 0xA0);
+	assert(text_decoder_eof(decoder) == 0);
+	assert(text_decoder_getc(decoder) == 0x0162);
+	assert(text_decoder_eof(decoder) == 1);
+
+	destroy_text_decoder(decoder);
+}
+
 int
 main(int argc, char **argv)
 {
@@ -131,6 +162,7 @@ main(int argc, char **argv)
 	test_unknown_encoding();
 	test_us_ascii_encoding();
 	test_iso_8859_1_encoding();
+	test_iso_8859_2_encoding();
 	printf("done\n");
 
 	return EXIT_SUCCESS;