From b74f756f00bff5a69b1c3dbcd826c6010c438a66 Mon Sep 17 00:00:00 2001
From: "Reece H. Dunn" <msclrhd@gmail.com>
Date: Thu, 30 Mar 2017 22:33:22 +0100
Subject: [PATCH] encoding.c: Support the ISO-10646-UCS-2 encoding.

---
 src/include/espeak-ng/espeak_ng.h |  1 +
 src/libespeak-ng/encoding.c       | 18 +++++++++++++++++-
 tests/encoding.c                  | 27 +++++++++++++++++++++++++++
 3 files changed, 45 insertions(+), 1 deletion(-)

diff --git a/src/include/espeak-ng/espeak_ng.h b/src/include/espeak-ng/espeak_ng.h
index 6e9079be..12491cee 100644
--- a/src/include/espeak-ng/espeak_ng.h
+++ b/src/include/espeak-ng/espeak_ng.h
@@ -209,6 +209,7 @@ typedef enum
 	ESPEAKNG_ENCODING_KOI8_R,
 	ESPEAKNG_ENCODING_ISCII,
 	ESPEAKNG_ENCODING_UTF_8,
+	ESPEAKNG_ENCODING_ISO_10646_UCS_2,
 } espeak_ng_ENCODING;
 
 ESPEAK_NG_API espeak_ng_ENCODING
diff --git a/src/libespeak-ng/encoding.c b/src/libespeak-ng/encoding.c
index 2c514e02..a16332be 100644
--- a/src/libespeak-ng/encoding.c
+++ b/src/libespeak-ng/encoding.c
@@ -66,6 +66,7 @@ MNEM_TAB mnem_encoding[] = {
 	{ "ISO_8859-16",      ESPEAKNG_ENCODING_ISO_8859_16 },
 	{ "ISO_8859-16:2001", ESPEAKNG_ENCODING_ISO_8859_16 },
 	{ "ISO646-US",        ESPEAKNG_ENCODING_US_ASCII },
+	{ "ISO-10646-UCS-2",  ESPEAKNG_ENCODING_ISO_10646_UCS_2 },
 	{ "ISO-8859-1",       ESPEAKNG_ENCODING_ISO_8859_1 },
 	{ "ISO-8859-2",       ESPEAKNG_ENCODING_ISO_8859_2 },
 	{ "ISO-8859-3",       ESPEAKNG_ENCODING_ISO_8859_3 },
@@ -106,6 +107,7 @@ MNEM_TAB mnem_encoding[] = {
 	{ "csKOI8R",          ESPEAKNG_ENCODING_KOI8_R },
 	{ "csTIS620",         ESPEAKNG_ENCODING_ISO_8859_11 },
 	{ "csUTF8",           ESPEAKNG_ENCODING_UTF_8 },
+	{ "csUnicode",        ESPEAKNG_ENCODING_ISO_10646_UCS_2 },
 	{ "arabic",           ESPEAKNG_ENCODING_ISO_8859_6 },
 	{ "cyrillic",         ESPEAKNG_ENCODING_ISO_8859_5 },
 	{ "greek",            ESPEAKNG_ENCODING_ISO_8859_7 },
@@ -566,6 +568,19 @@ eof:
 	return 0xFFFD;
 }
 
+static uint32_t
+string_decoder_getc_iso_10646_ucs_2(espeak_ng_TEXT_DECODER *decoder)
+{
+	if (decoder->current + 1 >= decoder->end) {
+		decoder->current = decoder->end;
+		return 0xFFFD;
+	}
+
+	uint8_t c1 = *decoder->current++ & 0xFF;
+	uint8_t c2 = *decoder->current++ & 0xFF;
+	return c1 + (c2 << 8);
+}
+
 typedef struct
 {
 	uint32_t (*get)(espeak_ng_TEXT_DECODER *decoder);
@@ -594,6 +609,7 @@ static const encoding_t string_decoders[] = {
 	{ string_decoder_getc_codepage, KOI8_R },
 	{ string_decoder_getc_codepage, ISCII },
 	{ string_decoder_getc_utf_8, NULL },
+	{ string_decoder_getc_iso_10646_ucs_2, NULL },
 };
 
 espeak_ng_TEXT_DECODER *
@@ -621,7 +637,7 @@ text_decoder_decode_string(espeak_ng_TEXT_DECODER *decoder,
                            int length,
                            espeak_ng_ENCODING encoding)
 {
-	if (encoding > ESPEAKNG_ENCODING_UTF_8)
+	if (encoding > ESPEAKNG_ENCODING_ISO_10646_UCS_2)
 		return ENS_UNKNOWN_TEXT_ENCODING;
 
 	const encoding_t *enc = string_decoders + encoding;
diff --git a/tests/encoding.c b/tests/encoding.c
index 8ec8a503..b8140b05 100644
--- a/tests/encoding.c
+++ b/tests/encoding.c
@@ -702,6 +702,32 @@ test_utf_8_encoding()
 	destroy_text_decoder(decoder);
 }
 
+void
+test_iso_10646_ucs_2_encoding()
+{
+	printf("testing ISO-10646-UCS-2 encoding\n");
+
+	assert(espeak_ng_EncodingFromName("ISO-10646-UCS-2") == ESPEAKNG_ENCODING_ISO_10646_UCS_2);
+	assert(espeak_ng_EncodingFromName("csUnicode") == ESPEAKNG_ENCODING_ISO_10646_UCS_2);
+
+	espeak_ng_TEXT_DECODER *decoder = create_text_decoder();
+
+	assert(text_decoder_decode_string(decoder, "a\00G\00\xA0\00\x22\x21\x23\x21", 9, ESPEAKNG_ENCODING_ISO_10646_UCS_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) == 0xA0);
+	assert(text_decoder_eof(decoder) == 0);
+	assert(text_decoder_getc(decoder) == 0x2122);
+	assert(text_decoder_eof(decoder) == 0);
+	assert(text_decoder_getc(decoder) == 0xFFFD);
+	assert(text_decoder_eof(decoder) == 1);
+
+	destroy_text_decoder(decoder);
+}
+
 int
 main(int argc, char **argv)
 {
@@ -730,6 +756,7 @@ main(int argc, char **argv)
 	test_iso_8859_16_encoding();
 
 	test_utf_8_encoding();
+	test_iso_10646_ucs_2_encoding();
 
 	printf("done\n");