From: Aki Tuomi <aki.tuomi@dovecot.fi>
Date: Fri, 3 Nov 2017 07:50:40 +0000 (+0200)
Subject: lib: json-parser - Insert properly escaped unicode
X-Git-Tag: 2.3.0.rc1~496
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=32bd32dcc845cd0c00d5617aea1ffbe45522b413;p=thirdparty%2Fdovecot%2Fcore.git

lib: json-parser - Insert properly escaped unicode
---

diff --git a/src/lib/json-parser.c b/src/lib/json-parser.c
index 2dc3eacd98..09964741ce 100644
--- a/src/lib/json-parser.c
+++ b/src/lib/json-parser.c
@@ -755,7 +755,7 @@ static void json_append_escaped_char(string_t *dest, unsigned char src)
 		str_append(dest, "\\\\");
 		break;
 	default:
-		if (src < 32)
+		if (src < 0x20 || src >= 0x80)
 			str_printfa(dest, "\\u%04x", src);
 		else
 			str_append_c(dest, src);
@@ -763,6 +763,19 @@ static void json_append_escaped_char(string_t *dest, unsigned char src)
 	}
 }
 
+void json_append_escaped_ucs4(string_t *dest, unichar_t chr)
+{
+	unichar_t high,low;
+	if (chr < 0x80)
+		json_append_escaped_char(dest, (unsigned char)chr);
+	else if (chr >= UTF16_SURROGATE_BASE) {
+		uni_split_surrogate(chr, &high, &low);
+		str_printfa(dest, "\\u%04x\\u%04x", high, low);
+	} else {
+		str_printfa(dest, "\\u%04x", chr);
+	}
+}
+
 void ostream_escaped_json_format(string_t *dest, unsigned char src)
 {
 	json_append_escaped_char(dest, src);
@@ -770,14 +783,24 @@ void ostream_escaped_json_format(string_t *dest, unsigned char src)
 
 void json_append_escaped(string_t *dest, const char *src)
 {
-	for (; *src != '\0'; src++)
-		json_append_escaped_char(dest, *src);
+	json_append_escaped_data(dest, (const unsigned char*)src, strlen(src));
 }
 
 void json_append_escaped_data(string_t *dest, const unsigned char *src, size_t size)
 {
 	size_t i;
-
-	for (i = 0; i < size; i++)
-		json_append_escaped_char(dest, src[i]);
+	int bytes = 0;
+	unichar_t chr;
+
+	for (i = 0; i < size;) {
+		bytes = uni_utf8_get_char_n(src+i, size-i, &chr);
+		/* if it was valid unichar, encode + move forward by bytes */
+		if (bytes > 0) {
+			json_append_escaped_ucs4(dest, chr);
+			i += bytes;
+		/* encode as byte data */
+		} else {
+			json_append_escaped_char(dest, src[i++]);
+		}
+	}
 }
diff --git a/src/lib/json-parser.h b/src/lib/json-parser.h
index 551422dcc3..78fb448f7e 100644
--- a/src/lib/json-parser.h
+++ b/src/lib/json-parser.h
@@ -1,6 +1,8 @@
 #ifndef JSON_PARSER_H
 #define JSON_PARSER_H
 
+#include "unichar.h"
+
 enum json_type {
 	/* { key: */
 	JSON_TYPE_OBJECT_KEY,
@@ -46,6 +48,8 @@ void json_parse_skip_next(struct json_parser *parser);
 int json_parse_next_stream(struct json_parser *parser,
 			   struct istream **input_r);
 
+/* Append UCS4 to already opened JSON string. */
+void json_append_escaped_ucs4(string_t *dest, unichar_t chr);
 /* Append data to already opened JSON string. src should be valid UTF-8 data. */
 void json_append_escaped(string_t *dest, const char *src);
 /* Same as json_append_escaped(), but append non-\0 terminated input. */
diff --git a/src/lib/test-json-parser.c b/src/lib/test-json-parser.c
index 1635bb848b..eedeb8440a 100644
--- a/src/lib/test-json-parser.c
+++ b/src/lib/test-json-parser.c
@@ -266,20 +266,20 @@ static void test_json_append_escaped(void)
 	string_t *str = t_str_new(32);
 
 	test_begin("json_append_escaped()");
-	json_append_escaped(str, "\b\f\r\n\t\"\\\001\002-\xC3\xA4");
-	test_assert(strcmp(str_c(str), "\\b\\f\\r\\n\\t\\\"\\\\\\u0001\\u0002-\xC3\xA4") == 0);
+	json_append_escaped(str, "\b\f\r\n\t\"\\\001\002-\xC3\xA4\xf0\x90\x90\xb7");
+	test_assert(strcmp(str_c(str), "\\b\\f\\r\\n\\t\\\"\\\\\\u0001\\u0002-\\u00e4\\ud801\\udc37") == 0);
 	test_end();
 }
 
 static void test_json_append_escaped_data(void)
 {
 	static const unsigned char test_input[] =
-		"\b\f\r\n\t\"\\\000\001\002-\xC3\xA4";
+		"\b\f\r\n\t\"\\\000\001\002-\xC3\xA4\xf0\x90\x90\xb7";
 	string_t *str = t_str_new(32);
 
 	test_begin("json_append_escaped()");
 	json_append_escaped_data(str, test_input, sizeof(test_input)-1);
-	test_assert(strcmp(str_c(str), "\\b\\f\\r\\n\\t\\\"\\\\\\u0000\\u0001\\u0002-\xC3\xA4") == 0);
+	test_assert(strcmp(str_c(str), "\\b\\f\\r\\n\\t\\\"\\\\\\u0000\\u0001\\u0002-\\u00e4\\ud801\\udc37") == 0);
 	test_end();
 }