Use std::vector for charmap output

include/asm/charmap.hpp

@@ -4,6 +4,7 @@
 #define RGBDS_ASM_CHARMAP_H
 
 #include <stdint.h>
+#include <vector>
 
 #define DEFAULT_CHARMAP_NAME "main"
 
@@ -13,7 +14,7 @@ void charmap_Push(void);
 void charmap_Pop(void);
 void charmap_Add(char *mapping, uint8_t value);
 bool charmap_HasChar(char const *input);
-size_t charmap_Convert(char const *input, uint8_t *output);
+void charmap_Convert(char const *input, std::vector<uint8_t> &output);
-size_t charmap_ConvertNext(char const **input, uint8_t **output);
+size_t charmap_ConvertNext(char const **input, std::vector<uint8_t> *output);
 
 #endif // RGBDS_ASM_CHARMAP_H

include/util.hpp

@@ -5,12 +5,13 @@
 
 #include <stddef.h>
 #include <stdint.h>
+#include <vector>
 
 char const *printChar(int c);
 
 /*
  * @return The number of bytes read, or 0 if invalid data was found
  */
-size_t readUTF8Char(uint8_t *dest, char const *src);
+size_t readUTF8Char(std::vector<uint8_t> *dest, char const *src);
 
 #endif // RGBDS_UTIL_H

src/asm/charmap.cpp

@@ -139,17 +139,13 @@ bool charmap_HasChar(char const *input)
 	return charmap.nodes[nodeIdx].isTerminal;
 }
 
-size_t charmap_Convert(char const *input, uint8_t *output)
+void charmap_Convert(char const *input, std::vector<uint8_t> &output)
 {
-	uint8_t *start = output;
-
 	while (charmap_ConvertNext(&input, &output))
 		;
-
-	return output - start;
 }
 
-size_t charmap_ConvertNext(char const **input, uint8_t **output)
+size_t charmap_ConvertNext(char const **input, std::vector<uint8_t> *output)
 {
 	// The goal is to match the longest mapping possible.
 	// For that, advance through the trie with each character read.
@@ -181,22 +177,20 @@ size_t charmap_ConvertNext(char const **input, uint8_t **output)
 
 	if (matchIdx) { // A match was found, use it
 		if (output)
-			*(*output)++ = charmap.nodes[matchIdx].value;
+			output->push_back(charmap.nodes[matchIdx].value);
 
 		return 1;
 
 	} else if (**input) { // No match found, but there is some input left
 		int firstChar = **input;
 		// This will write the codepoint's value to `output`, little-endian
-		size_t codepointLen = readUTF8Char(output ? *output : NULL, *input);
+		size_t codepointLen = readUTF8Char(output, *input);
 
 		if (codepointLen == 0)
 			error("Input string is not valid UTF-8\n");
 
 		// OK because UTF-8 has no NUL in multi-byte chars
 		*input += codepointLen;
-		if (output)
-			*output += codepointLen;
 
 		// Warn if this character is not mapped but any others are
 		if (charmap.nodes.size() > 1)

src/asm/parser.y

@@ -46,8 +46,10 @@ static void lowerstring(char *dest, char const *src)
 	*dest = '\0';
 }
 
-static uint32_t str2int2(uint8_t *s, uint32_t length)
+static uint32_t str2int2(std::vector<uint8_t> const &s)
 {
+	uint32_t length = s.size();
+
 	if (length > 4)
 		warning(WARNING_NUMERIC_STRING_1,
 			"Treating string as a number ignores first %" PRIu32 " character%s\n",
@@ -1343,11 +1345,10 @@ constlist_8bit_entry : reloc_8bit_no_str {
 			sect_RelByte(&$1, 0);
 		}
 		| string {
-			uint8_t *output = (uint8_t *)malloc(strlen($1)); // Cannot be larger than that
+			std::vector<uint8_t> output;
-			size_t length = charmap_Convert($1, output);
 
-			sect_AbsByteGroup(output, length);
+			charmap_Convert($1, output);
-			free(output);
+			sect_AbsByteGroup(output.data(), output.size());
 		}
 ;
 
@@ -1359,11 +1360,10 @@ constlist_16bit_entry : reloc_16bit_no_str {
 			sect_RelWord(&$1, 0);
 		}
 		| string {
-			uint8_t *output = (uint8_t *)malloc(strlen($1)); // Cannot be larger than that
+			std::vector<uint8_t> output;
-			size_t length = charmap_Convert($1, output);
 
-			sect_AbsWordGroup(output, length);
+			charmap_Convert($1, output);
-			free(output);
+			sect_AbsWordGroup(output.data(), output.size());
 		}
 ;
 
@@ -1375,12 +1375,10 @@ constlist_32bit_entry : relocexpr_no_str {
 			sect_RelLong(&$1, 0);
 		}
 		| string {
-			// Charmaps cannot increase the length of a string
+			std::vector<uint8_t> output;
-			uint8_t *output = (uint8_t *)malloc(strlen($1));
-			size_t length = charmap_Convert($1, output);
 
-			sect_AbsLongGroup(output, length);
+			charmap_Convert($1, output);
-			free(output);
+			sect_AbsLongGroup(output.data(), output.size());
 		}
 ;
 
@@ -1421,13 +1419,10 @@ reloc_16bit_no_str : relocexpr_no_str {
 
 relocexpr	: relocexpr_no_str
 		| string {
-			// Charmaps cannot increase the length of a string
+			std::vector<uint8_t> output;
-			uint8_t *output = (uint8_t *)malloc(strlen($1));
-			uint32_t length = charmap_Convert($1, output);
-			uint32_t r = str2int2(output, length);
 
-			free(output);
+			charmap_Convert($1, output);
-			rpn_Number(&$$, r);
+			rpn_Number(&$$, str2int2(output));
 		}
 ;
 

src/util.cpp

@@ -3,6 +3,7 @@
 #include <ctype.h>
 #include <stdint.h>
 #include <stdio.h>
+#include <vector>
 
 #include "util.hpp"
 
@@ -50,7 +51,7 @@ char const *printChar(int c)
 	return buf;
 }
 
-size_t readUTF8Char(uint8_t *dest, char const *src)
+size_t readUTF8Char(std::vector<uint8_t> *dest, char const *src)
 {
 	uint32_t state = 0;
 	uint32_t codep;
@@ -61,7 +62,7 @@ size_t readUTF8Char(uint8_t *dest, char const *src)
 			return 0;
 
 		if (dest)
-			dest[i] = src[i];
+			dest->push_back(src[i]);
 		i++;
 
 		if (state == 0)

test/asm/empty-strings.asm

@@ -0,0 +1,6 @@
+SECTION "test", ROM0
+db ""
+dw ""
+dl ""
+assert ("") == 0
+assert SIZEOF("test") == 0