encoding.c: Support the ISCII encoding.

src/include/espeak-ng/espeak_ng.h

@@ -207,6 +207,7 @@ typedef enum
 	ESPEAKNG_ENCODING_ISO_8859_15,
 	ESPEAKNG_ENCODING_ISO_8859_16,
 	ESPEAKNG_ENCODING_KOI8_R,
+	ESPEAKNG_ENCODING_ISCII,
 } espeak_ng_ENCODING;
 
 ESPEAK_NG_API espeak_ng_ENCODING

src/libespeak-ng/encoding.c

@@ -35,6 +35,7 @@ MNEM_TAB mnem_encoding[] = {
 	{ "ELOT_928",         ESPEAKNG_ENCODING_ISO_8859_7 },
 	{ "IBM367",           ESPEAKNG_ENCODING_US_ASCII },
 	{ "IBM819",           ESPEAKNG_ENCODING_ISO_8859_1 },
+	{ "ISCII",            ESPEAKNG_ENCODING_ISCII },
 	{ "ISO_646.irv:1991", ESPEAKNG_ENCODING_US_ASCII },
 	{ "ISO_8859-1",       ESPEAKNG_ENCODING_ISO_8859_1 },
 	{ "ISO_8859-1:1987",  ESPEAKNG_ENCODING_ISO_8859_1 },
@@ -468,6 +469,26 @@ static const uint32_t KOI8_R[0x80] = {
 	0x0171, 0x00f9, 0x00fa, 0x00fb, 0x00fc, 0x0119, 0x021b, 0x00ff, // f8
 };
 
+// Reference: http://varamozhi.sourceforge.net/iscii91.pdf
+static const uint32_t ISCII[0x80] = {
+	0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, // 80
+	0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, // 88
+	0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, // 90
+	0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, // 98
+	0xfffd, 0x0901, 0x0902, 0x0903, 0x0905, 0x0906, 0x0907, 0x0908, // a0
+	0x0909, 0x090a, 0x090b, 0x090e, 0x090f, 0x0910, 0x090d, 0x0912, // a8
+	0x0913, 0x0914, 0x0911, 0x0915, 0x0916, 0x0917, 0x0918, 0x0919, // b0
+	0x091a, 0x091b, 0x091c, 0x091d, 0x091e, 0x091f, 0x0920, 0x0921, // b8
+	0x0922, 0x0923, 0x0924, 0x0925, 0x0926, 0x0927, 0x0928, 0x0929, // c0
+	0x092a, 0x092b, 0x092c, 0x092d, 0x092e, 0x092f, 0x095f, 0x0930, // c8
+	0x0931, 0x0932, 0x0933, 0x0934, 0x0935, 0x0936, 0x0937, 0x0938, // d0
+	0x0939, 0x0020, 0x093e, 0x093f, 0x0940, 0x0941, 0x0942, 0x0943, // d8
+	0x0946, 0x0947, 0x0948, 0x0945, 0x094a, 0x094b, 0x094c, 0x0949, // e0
+	0x094d, 0x093c, 0x0964, 0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, // e8
+	0x0020, 0x0030, 0x0031, 0x0032, 0x0033, 0x0034, 0x0035, 0x0036, // f0
+	0x0037, 0x0038, 0x0039, 0xfffd, 0xfffd, 0xfffd, 0xfffd, 0xfffd, // f8
+};
+
 static uint32_t
 string_decoder_getc_us_ascii(espeak_ng_TEXT_DECODER *decoder)
 {
@@ -516,6 +537,7 @@ static const encoding_t string_decoders[] = {
 	{ string_decoder_getc_codepage, ISO_8859_15 },
 	{ string_decoder_getc_codepage, ISO_8859_16 },
 	{ string_decoder_getc_codepage, KOI8_R },
+	{ string_decoder_getc_codepage, ISCII },
 };
 
 espeak_ng_TEXT_DECODER *
@@ -543,7 +565,7 @@ text_decoder_decode_string(espeak_ng_TEXT_DECODER *decoder,
                            int length,
                            espeak_ng_ENCODING encoding)
 {
-	if (encoding > ESPEAKNG_ENCODING_KOI8_R)
+	if (encoding > ESPEAKNG_ENCODING_ISCII)
 		return ENS_UNKNOWN_TEXT_ENCODING;
 
 	const encoding_t *enc = string_decoders + encoding;

tests/encoding.c

@@ -117,6 +117,31 @@ test_koi8_r_encoding()
 	destroy_text_decoder(decoder);
 }
 
+void
+test_iscii_encoding()
+{
+	printf("testing ISCII encoding\n");
+
+	assert(espeak_ng_EncodingFromName("ISCII") == ESPEAKNG_ENCODING_ISCII);
+
+	espeak_ng_TEXT_DECODER *decoder = create_text_decoder();
+
+	assert(text_decoder_decode_string(decoder, "aG\x92\xA0\xE6", 5, ESPEAKNG_ENCODING_ISCII) == 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) == 0xfffd);
+	assert(text_decoder_eof(decoder) == 0);
+	assert(text_decoder_getc(decoder) == 0xfffd);
+	assert(text_decoder_eof(decoder) == 0);
+	assert(text_decoder_getc(decoder) == 0x094c);
+	assert(text_decoder_eof(decoder) == 1);
+
+	destroy_text_decoder(decoder);
+}
+
 void
 test_iso_8859_1_encoding()
 {
@@ -582,6 +607,7 @@ main(int argc, char **argv)
 
 	test_us_ascii_encoding();
 	test_koi8_r_encoding();
+	test_iscii_encoding();
 
 	test_iso_8859_1_encoding();
 	test_iso_8859_2_encoding();