several utility enhancements

These were coded for the `split_str` refactor, but are generally useful elsewhere as well.

1. upgrade Unicode `text_iterator` for C++17
2. Use SCP_string in some utility functions where appropriate
3. Avoid copying C-strings in some utility functions where it isn't necessary
4. Add concat functions to `SCP_vector`
5. Change return type `int` to `bool` on some common parselo functions
6. Add `find_white_space` and `find_gray_space` parselo functions
7. Guard arguments in `message_log_add_segs`

Author: Goober5000

code/globalincs/vmallocator.h

@@ -32,6 +32,16 @@ class SCP_vector : public std::vector<T, std::allocator<T>>
 		return std::find(this->begin(), this->end(), item) != this->end();
 	}
 
+	void concat(SCP_vector<T>&& other)
+	{
+		insert(this->end(), std::make_move_iterator(other.begin()), std::make_move_iterator(other.end()));
+	}
+
+	void concat(const SCP_vector<T>& other)
+	{
+		insert(this->end(), other.begin(), other.end());
+	}
+
 	bool in_bounds(int idx) const
 	{
 		return (idx >= 0) && (static_cast<size_t>(idx) < this->size());

code/hud/hud.cpp

@@ -987,30 +987,28 @@ void HudGauge::renderStringAlignCenter(int x, int y, int area_width, const char
 
 void HudGauge::renderPrintf(int x, int y, float scale, bool config, const char* format, ...)
 {
-	char tmp[256] = "";
+	SCP_string tmp;
 	va_list args;
 
 	// format the text
 	va_start(args, format);
-	vsnprintf(tmp, sizeof(tmp), format, args);
+	vsprintf(tmp, format, args);
 	va_end(args);
-	tmp[sizeof(tmp)-1] = '\0';
 
-	renderString(x, y, tmp, scale, config);
+	renderString(x, y, tmp.c_str(), scale, config);
 }
 
 void HudGauge::renderPrintfWithGauge(int x, int y, int gauge_id, float scale, bool config, const char* format, ...)
 {
-	char tmp[256] = "";
+	SCP_string tmp;
 	va_list args;
 
 	// format the text
 	va_start(args, format);
-	vsnprintf(tmp, sizeof(tmp), format, args);
+	vsprintf(tmp, format, args);
 	va_end(args);
-	tmp[sizeof(tmp)-1] = '\0';
 
-	renderString(x, y, gauge_id, tmp, scale, config);
+	renderString(x, y, gauge_id, tmp.c_str(), scale, config);
 }
 
 void HudGauge::renderBitmapColor(int frame, int x, int y, float scale, bool config) const

code/hud/hudmessage.cpp

@@ -140,7 +140,7 @@ static SCP_vector<line_node> Msg_scrollback_lines;
 
 typedef struct HUD_ft {
 	int	end_time;						//	Timestamp at which this message will go away.
-	char	text[MAX_HUD_LINE_LEN];		//	Text to display.
+	char	text[MAX_HUD_LINE_BUF];		//	Text to display.
 	int	color;							//	0rgb color, 8 bit fields.
 } HUD_ft;
 
@@ -487,8 +487,6 @@ void HudGaugeMessages::render(float  /*frametime*/, bool config)
 void HUD_fixed_printf(float duration, color col, const char *format, ...)
 {
 	va_list	args;
-	char		tmp[HUD_MSG_LENGTH_MAX];
-	size_t		msg_length;
 
 	// make sure we only print these messages if we're in the correct state
 	if((Game_mode & GM_MULTIPLAYER) && (Netgame.game_state != NETGAME_STATE_IN_MISSION)){
@@ -497,22 +495,9 @@ void HUD_fixed_printf(float duration, color col, const char *format, ...)
 	}
 
 	va_start(args, format);
-	vsnprintf(tmp, sizeof(tmp), format, args);
+	vsnprintf(HUD_fixed_text[0].text, MAX_HUD_LINE_BUF, format, args);
 	va_end(args);
-	tmp[sizeof(tmp)-1] = '\0';
-
-	msg_length = strlen(tmp);
-
-	if ( !msg_length ) {
-		nprintf(("Warning", "HUD_fixed_printf ==> attempt to print a 0 length string in msg window\n"));
-		return;
-
-	} else if (msg_length > MAX_HUD_LINE_LEN - 1){
-		nprintf(("Warning", "HUD_fixed_printf ==> Following string truncated to %d chars: %s\n", MAX_HUD_LINE_LEN - 1, tmp));
-		tmp[MAX_HUD_LINE_LEN-1] = '\0';
-	}
-
-	strcpy_s(HUD_fixed_text[0].text, tmp);
+	HUD_fixed_text[0].text[MAX_HUD_LINE_BUF-1] = '\0';
 
 	if (duration == 0.0f){
 		HUD_fixed_text[0].end_time = timestamp(-1);
@@ -544,7 +529,7 @@ int HUD_source_get_team(int source)
 void HUD_printf(const char *format, ...)
 {
 	va_list args;
-	char tmp[HUD_MSG_LENGTH_MAX];
+	SCP_string tmp;
 
 	// make sure we only print these messages if we're in the correct state
 	if((Game_mode & GM_MULTIPLAYER) && (Net_player->state != NETPLAYER_STATE_IN_MISSION)){
@@ -553,9 +538,8 @@ void HUD_printf(const char *format, ...)
 	}
 
 	va_start(args, format);
-	vsnprintf(tmp, sizeof(tmp), format, args);
+	vsprintf(tmp, format, args);
 	va_end(args);
-	tmp[sizeof(tmp)-1] = '\0';
 
 	hud_sourced_print(HUD_SOURCE_COMPUTER, tmp);
 }
@@ -570,7 +554,7 @@ void HUD_printf(const char *format, ...)
 void HUD_sourced_printf(int source, const char *format, ...)
 {
 	va_list args;
-	char tmp[HUD_MSG_LENGTH_MAX];
+	SCP_string tmp;
 
 	// make sure we only print these messages if we're in the correct state
 	if((Game_mode & GM_MULTIPLAYER) && (Net_player->state != NETPLAYER_STATE_IN_MISSION)){
@@ -579,16 +563,40 @@ void HUD_sourced_printf(int source, const char *format, ...)
 	}
 
 	va_start(args, format);
-	vsnprintf(tmp, sizeof(tmp), format, args);
+	vsprintf(tmp, format, args);
 	va_end(args);
-	tmp[sizeof(tmp)-1] = '\0';
 
 	hud_sourced_print(source, tmp);
 }
 
+void hud_sourced_print(int source, const SCP_string &msg)
+{
+	if ( msg.empty() ) {
+		nprintf(("Warning", "HUD ==> attempt to print a 0 length string in msg window\n"));
+		return;
+	}
+
+	// add message to the scrollback log first
+	hud_add_msg_to_scrollback(msg.c_str(), source, Missiontime);
+
+	HUD_message_data new_msg;
+
+	new_msg.text = msg;
+	new_msg.source = source;
+	new_msg.x = 0;
+
+	HUD_msg_buffer.push_back(new_msg);
+
+	// Invoke the scripting hook
+	if (OnHudMessageReceivedHook->isActive()) {
+		OnHudMessageReceivedHook->run(scripting::hook_param_list(scripting::hook_param("Text", 's', msg.c_str()),
+			scripting::hook_param("SourceType", 'i', source)));
+	}
+}
+
 void hud_sourced_print(int source, const char *msg)
 {
-	if ( !strlen(msg) ) {
+	if ( !*msg ) {
 		nprintf(("Warning", "HUD ==> attempt to print a 0 length string in msg window\n"));
 		return;
 	}
@@ -643,21 +651,15 @@ void HUD_add_to_scrollback(const char *text, int source)
 
 void hud_add_msg_to_scrollback(const char *text, int source, int t)
 {
-	size_t msg_len = strlen(text);
-	if (msg_len == 0)
+	if (!*text)
 		return;
-	
-	Assert(msg_len < HUD_MSG_LENGTH_MAX);
-
-	char buf[HUD_MSG_LENGTH_MAX], *ptr;
-	strcpy_s(buf, text);
-	ptr = strstr(buf, NOX(": "));
 
 	int w = 0;
 
 	// determine the length of the sender's name for underlining
+	auto ptr = strstr(text, NOX(": "));
 	if (ptr) {
-		gr_get_string_size(&w, nullptr, buf, 1.0f, (ptr - buf));
+		gr_get_string_size(&w, nullptr, text, 1.0f, (ptr - text));
 	}
 
 	// create the new node for the vector

code/hud/hudmessage.h

@@ -16,7 +16,7 @@
 #include "graphics/generic.h"
 #include "hud/hud.h"
 
-#define MAX_HUD_LINE_LEN			256			// maximum number of characters for a HUD message
+#define MAX_HUD_LINE_BUF			256			// maximum size for a HUD message
 
 // If these are changed, the lua 'addMessageToScrollback' method in mission.cpp should be updated.
 #define HUD_SOURCE_COMPUTER		0
@@ -67,12 +67,12 @@ int HUD_team_get_source(int team);
 int HUD_source_get_team(int team);
 void HUD_printf(SCP_FORMAT_STRING const char *format, ...) SCP_FORMAT_STRING_ARGS(1, 2);
 void hud_sourced_print(int source, const char *msg);
+void hud_sourced_print(int source, const SCP_string &msg);
 void HUD_sourced_printf(int source, SCP_FORMAT_STRING const char *format, ...) SCP_FORMAT_STRING_ARGS(2, 3);  // send hud message from specified source
 void HUD_fixed_printf(float duration, color col, SCP_FORMAT_STRING const char *format, ...) SCP_FORMAT_STRING_ARGS(3, 4);		//	Display a single message for duration seconds.
 void HUD_init_fixed_text();			//	Clear all pending fixed text.
 
 void HUD_add_to_scrollback(const char *text, int source);
-void hud_add_line_to_scrollback(const char *text, int source, int t, int x, int y, int w);
 void hud_add_msg_to_scrollback(const char *text, int source, int t);
 
 class HudGaugeMessages: public HudGauge // HUD_MESSAGE_LINES

code/mission/missionlog.cpp

@@ -444,9 +444,9 @@ void message_log_add_seg(log_text_seg* entry, int x, int msg_color, const char*
 	entry->flags = flags;
 }
 
-void message_log_add_segs(const char* source_string, int msg_color, int flags = 0, SCP_vector<log_text_seg> *entry = nullptr, bool split_string = false)
+void message_log_add_segs(const char *source_string, int msg_color, int flags, SCP_vector<log_text_seg> *entry, bool split_string)
 {
-	if (!source_string) {
+	if (!source_string || !entry) {
 		mprintf(("Why are you passing a NULL pointer to message_log_add_segs?\n"));
 		return;
 	}

code/parse/parselo.cpp

@@ -68,25 +68,43 @@ static const SCP_unordered_map<SCP_string, SCP_string> retail_hashes = {
 
 
 //	Return true if this character is white space, else false.
-int is_white_space(char ch)
+bool is_white_space(char ch)
 {
 	return ((ch == ' ') || (ch == '\t') || (ch == EOLN) || (ch == CARRIAGE_RETURN));
 }
-int is_white_space(unicode::codepoint_t cp)
+
+//	Return true if this character is white space, else false.
+bool is_white_space(unicode::codepoint_t cp)
 {
 	return ((cp == UNICODE_CHAR(' ')) || (cp == UNICODE_CHAR('\t')) || (cp == (unicode::codepoint_t)EOLN) || (cp == (unicode::codepoint_t)CARRIAGE_RETURN));
 }
 
+//  Returns the length of the string up to but excluding any white space.  This could be the entire string if the string contains no white space.
+//	Equivalently, returns the position of the first white space character, or the string length if no white space is found.
+size_t find_white_space(const char *str)
+{
+	return strcspn(str, " \t\n\r");
+}
+
 // Returns true if this character is gray space, else false (gray space is white space except for EOLN).
-int is_gray_space(char ch)
+bool is_gray_space(char ch)
 {
 	return ((ch == ' ') || (ch == '\t'));
 }
 
-bool is_gray_space(unicode::codepoint_t cp) {
+// Returns true if this character is gray space, else false (gray space is white space except for EOLN).
+bool is_gray_space(unicode::codepoint_t cp)
+{
 	return cp == UNICODE_CHAR(' ') || cp == UNICODE_CHAR('\t');
 }
 
+//  Returns the length of the string up to but excluding any white space.  This could be the entire string if the string contains no white space.
+//	Equivalently, returns the position of the first white space character, or the string length if no white space is found.
+size_t find_gray_space(const char *str)
+{
+	return strcspn(str, " \t");
+}
+
 bool is_parenthesis(char ch)
 {
 	return ((ch == '(') || (ch == ')'));

code/parse/parselo.h

@@ -89,8 +89,9 @@ extern void consolidate_double_characters(char *str, char ch);
 char *three_dot_truncate(char *buffer, const char *source, size_t buffer_size);
 
 // white space
-extern int is_white_space(char ch);
-extern int is_white_space(unicode::codepoint_t cp);
+extern bool is_white_space(char ch);
+extern bool is_white_space(unicode::codepoint_t cp);
+extern size_t find_white_space(const char *str);
 extern void ignore_white_space(const char **pp = nullptr);
 extern void drop_trailing_white_space(char *str);
 extern void drop_leading_white_space(char *str);
@@ -102,8 +103,9 @@ extern void drop_leading_white_space(SCP_string &str);
 extern void drop_white_space(SCP_string &str);
 
 // gray space
-extern int is_gray_space(char ch);
+extern bool is_gray_space(char ch);
 extern bool is_gray_space(unicode::codepoint_t cp);
+extern size_t find_gray_space(const char *str);
 extern void ignore_gray_space(const char **pp = nullptr);
 
 // other

code/utils/unicode.h

@@ -42,7 +42,11 @@ class text_iterator {
  public:
 	explicit text_iterator(const char* current_byte, const char* range_start_byte, const char* range_end_byte = nullptr);
 
-	typedef codepoint_t value_type;
+	using difference_type = std::ptrdiff_t;
+	using value_type = codepoint_t;
+	using pointer = text_iterator*;
+	using reference = text_iterator&;
+	using iterator_category = std::bidirectional_iterator_tag;
 
 	const char* pos() const;
 

code/weapon/emp.cpp

@@ -495,11 +495,11 @@ void emp_hud_string(int x, int y, int gauge_id, const char *str, int resize_mode
 	if (!*str)
 		return;
 
-	// copy the string
-	strcpy_s(tmp, str);
-
 	// if the emp effect is not active, don't even bother messing with the text
 	if(emp_active_local()){
+		// use a copied string rather than the original
+		strcpy_s(tmp, str);
+
 		emp_maybe_reformat_text(tmp, 256, gauge_id);
 
 		// jitter the coords