separate parsing and interpreting text objects for certain objects

All objects parsing text into objects formerly parsed and evaluated
their arguments each update interval. This does only make sense when the
parsed text could change between updates, which is the case for execp
and execpi objects, but none of the others. So have them parse the text
when creating them, so each update interval only the output has to be
re-extracted.

In fact, this should give a performance boost when using any of the
affected objects, as parsing the input text is quite expensive since
we're evaluating templates.

src/conky.c

@@ -1270,6 +1270,9 @@ static void free_text_objects(struct text_object *root)
 				break;
 			case OBJ_if_empty:
 			case OBJ_if_match:
+				free_text_objects(obj->sub);
+				free(obj->sub);
+				/* fall through */
 			case OBJ_if_existing:
 			case OBJ_if_mounted:
 			case OBJ_if_running:
@@ -1528,6 +1531,8 @@ static void free_text_objects(struct text_object *root)
 #endif
 			case OBJ_scroll:
 				free(data.scroll.text);
+				free_text_objects(obj->sub);
+				free(obj->sub);
 				break;
 		}
 		free(obj);
@@ -2518,6 +2523,9 @@ static struct text_object *construct_text_object(const char *s,
 			obj->data.ifblock.s = 0;
 		} else {
 			obj->data.ifblock.s = strndup(arg, text_buffer_size);
+			obj->sub = malloc(sizeof(struct text_object));
+			extract_variable_text_internal(obj->sub,
+			                               obj->data.ifblock.s, 0);
 		}
 		obj_be_ifblock_if(ifblock_opaque, obj);
 	END OBJ(if_match, 0)
@@ -2526,6 +2534,9 @@ static struct text_object *construct_text_object(const char *s,
 			obj->data.ifblock.s = 0;
 		} else {
 			obj->data.ifblock.s = strndup(arg, text_buffer_size);
+			obj->sub = malloc(sizeof(struct text_object));
+			extract_variable_text_internal(obj->sub,
+			                               obj->data.ifblock.s, 0);
 		}
 		obj_be_ifblock_if(ifblock_opaque, obj);
 	END OBJ(if_existing, 0)
@@ -3066,6 +3077,9 @@ static struct text_object *construct_text_object(const char *s,
 				n1 += n2;
 			obj->data.scroll.text = strndup(arg + n1, text_buffer_size);
 			obj->data.scroll.start = 0;
+			obj->sub = malloc(sizeof(struct text_object));
+			extract_variable_text_internal(obj->sub,
+					obj->data.scroll.text, 0);
 		} else {
 			CRIT_ERR("scroll needs arguments: <length> [<step>] <text>");
 		}
@@ -4573,29 +4587,27 @@ static void generate_text_internal(char *p, int p_max_size,
 				new_alignc(p, obj->data.i);
 			}
 			OBJ(if_empty) {
-				struct text_object subroot;
+				char buf[max_user_text];
 				struct information *tmp_info =
 					malloc(sizeof(struct information));
 				memcpy(tmp_info, cur, sizeof(struct information));
-				parse_conky_vars(&subroot, obj->data.ifblock.s, p, tmp_info);
+				generate_text_internal(buf, max_user_text,
+				                       *obj->sub, tmp_info);
 
-				if (strlen(p) != 0) {
+				if (strlen(buf) != 0) {
 					DO_JUMP;
 				}
-				p[0] = '\0';
-				free_text_objects(&subroot);
 				free(tmp_info);
 			}
 			OBJ(if_match) {
 				char expression[max_user_text];
 				int val;
-				struct text_object subroot;
 				struct information *tmp_info;
 
 				tmp_info = malloc(sizeof(struct information));
 				memcpy(tmp_info, cur, sizeof(struct information));
-				parse_conky_vars(&subroot, obj->data.ifblock.s,
-						expression, tmp_info);
+				generate_text_internal(expression, max_user_text,
+				                       *obj->sub, tmp_info);
 				DBGP("parsed arg into '%s'", expression);
 
 				val = compare(expression);
@@ -4605,7 +4617,6 @@ static void generate_text_internal(char *p, int p_max_size,
 				} else if (!val) {
 					DO_JUMP;
 				}
-				free_text_objects(&subroot);
 				free(tmp_info);
 			}
 			OBJ(if_existing) {
@@ -5483,39 +5494,41 @@ head:
 #endif /* SMAPI */
 			OBJ(scroll) {
 				unsigned int j;
-				char *tmp;
-				struct text_object subroot;
-				parse_conky_vars(&subroot, obj->data.scroll.text, p, cur);
+				char *tmp, buf[max_user_text];
+				generate_text_internal(buf, max_user_text,
+				                       *obj->sub, cur);
 
-				if (strlen(p) <= obj->data.scroll.show) {
+				if (strlen(buf) <= obj->data.scroll.show) {
+					snprintf(p, p_max_size, "%s", buf);
 					break;
 				}
 #define LINESEPARATOR '|'
 				//place all the lines behind each other with LINESEPARATOR between them
-				for(j = 0; p[j] != 0; j++) {
-					if(p[j]=='\n') {
-						p[j]=LINESEPARATOR;
+				for(j = 0; buf[j] != 0; j++) {
+					if(buf[j]=='\n') {
+						buf[j]=LINESEPARATOR;
 					}
 				}
 				//scroll the output obj->data.scroll.start places by copying that many chars from
 				//the front of the string to tmp, scrolling the rest to the front and placing tmp
 				//at the back of the string
 				tmp = calloc(obj->data.scroll.start + 1, sizeof(char));
-				strncpy(tmp, p, obj->data.scroll.start); tmp[obj->data.scroll.start] = 0;
-				for(j = obj->data.scroll.start; p[j] != 0; j++){
-					p[j - obj->data.scroll.start] = p[j];
+				strncpy(tmp, buf, obj->data.scroll.start); tmp[obj->data.scroll.start] = 0;
+				for(j = obj->data.scroll.start; buf[j] != 0; j++){
+					buf[j - obj->data.scroll.start] = buf[j];
 				}
-				strcpy(&p[j - obj->data.scroll.start], tmp);
+				strcpy(&buf[j - obj->data.scroll.start], tmp);
 				free(tmp);
 				//only show the requested number of chars
 				if(obj->data.scroll.show < j) {
-					p[obj->data.scroll.show] = 0;
+					buf[obj->data.scroll.show] = 0;
 				}
 				//next time, scroll a place more or reset scrolling if we are at the end
 				obj->data.scroll.start += obj->data.scroll.step;
 				if(obj->data.scroll.start >= j){
 					 obj->data.scroll.start = 0;
 				}
+				snprintf(p, p_max_size, "%s", buf);
 			}
 #ifdef NVIDIA
 			OBJ(nvidia) {

src/text_object.h

@@ -354,6 +354,7 @@ enum text_object_type {
 
 struct text_object {
 	struct text_object *next, *prev;	/* doubly linked list of text objects */
+	struct text_object *sub;		/* for objects parsing text into objects */
 	union {
 		char *s;		/* some string */
 		int i;			/* some integer */