Merge unicode-props into trunk

FossilOrigin-Name: 95c406d482112373a41982bf36f3a9159fceff5ff95162cab34e97fd0cfaa73c

3 files changed, 286 insertions, 2 deletions

diff --git a/src/tools/Makefile.am b/src/tools/Makefile.am
index b1079e7..b05ab8a 100644
--- a/src/tools/Makefile.am
+++ b/src/tools/Makefile.am
@@ -1,10 +1,16 @@
-noinst_PROGRAMS += genboot
+noinst_PROGRAMS += genboot genunicode
 
 genboot_SOURCES = src/tools/genboot.c
+genunicode_SOURCES = src/tools/genunicode.c
 
 BUILT_SOURCES += $(top_builddir)/generated/boot.h
-CLEANFILES += $(top_builddir)/generated/boot.h
+BUILT_SOURCES += $(top_builddir)/generated/unicode_properties.c
 
 $(top_builddir)/generated/boot.h: genboot src/lib/boot.aplwc
 	$(MKDIR_P) $(top_builddir)/generated
 	./genboot $(top_srcdir)/src/lib/boot.aplwc > $@
+
+$(top_builddir)/generated/unicode_properties.c: genunicode src/unicode/DerivedCoreProperties.txt src/unicode/PropList.txt
+	$(MKDIR_P) $(top_builddir)/generated
+	./genunicode $(top_srcdir)/src/unicode > $@
+
diff --git a/src/tools/genboot.c b/src/tools/genboot.c
index e89b3aa..9d0a7e4 100644
--- a/src/tools/genboot.c
+++ b/src/tools/genboot.c
@@ -1,3 +1,22 @@
+/* 
+ * Copyright (C) 2026 Peter Mikkelsen <petermikkelsen10@gmail.com>
+ *
+ * This file is part of aplwc.
+ *
+ * Aplwc is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation, either version 3 of the License, or
+ * (at your option) any later version.
+ *
+ * Aplwc is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with aplwc.  If not, see <https://www.gnu.org/licenses/>.
+ */
+
 #include <stdio.h>
 #include <assert.h>
 
diff --git a/src/tools/genunicode.c b/src/tools/genunicode.c
new file mode 100644
index 0000000..76f2c60
--- /dev/null
+++ b/src/tools/genunicode.c
@@ -0,0 +1,259 @@
+/* 
+ * Copyright (C) 2026 Peter Mikkelsen <petermikkelsen10@gmail.com>
+ *
+ * This file is part of aplwc.
+ *
+ * Aplwc is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation, either version 3 of the License, or
+ * (at your option) any later version.
+ *
+ * Aplwc is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with aplwc.  If not, see <https://www.gnu.org/licenses/>.
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <stdlib.h>
+#include <assert.h>
+#include <setjmp.h>
+
+static char *files[] = {
+	"DerivedCoreProperties.txt",
+	"PropList.txt"
+};
+
+static char *properties[] = {
+	"ID_Start",
+	"ID_Continue",
+	"Pattern_Syntax"
+};
+
+struct line {
+	int from;
+	int to;
+	int valid;
+	char property[64];
+};
+
+static void emit_line(struct line *, int *);
+static int peek(FILE *);
+static int parse_line(FILE *, struct line *);
+static int parse_int(FILE *);
+static void parse_dots(FILE *);
+static void parse_sep(FILE *);
+
+static jmp_buf env;
+
+int
+main(int argc, char *argv[])
+{
+	FILE *f;
+	char *dir, *buf;
+	size_t dir_size, tmp_size, max_size, i, j, nfiles;
+	struct line line, prev_line;
+	int current, n;
+
+	assert(argc == 2);
+	dir = argv[1];
+	dir_size = strlen(dir);
+
+	if(dir[dir_size-1] == '/')
+		dir_size--;
+
+	nfiles = sizeof(files)/sizeof(files[0]);
+
+	max_size = 0;
+	for(i = 0; i < nfiles; i++){
+		tmp_size = strlen(files[i]);
+		if(tmp_size > max_size)
+			max_size = tmp_size;
+	}
+	max_size += dir_size + 2;
+	buf = malloc(max_size);
+	assert(buf != NULL);
+	memcpy(buf, dir, dir_size);
+	buf[dir_size] = '/';
+	current = -1;
+	n = 0;
+
+	printf("enum unicode_property {");
+	for(i = 0; i < sizeof(properties)/sizeof(properties[0]); i++)
+		printf("%s%s", (i == 0) ? "\n\t" : ",\n\t", properties[i]);
+	printf("\n};\n\n");
+
+	printf("struct unicode_range {\n\tuint32_t from;\n\tuint32_t to;\n};\n\n");
+
+	for(i = 0; i < sizeof(files)/sizeof(files[0]); i++){
+		memcpy(buf+dir_size+1, files[i], strlen(files[i])+1);
+		f = fopen(buf, "r");
+		assert(f != NULL);
+
+		printf("/* %s */\n\n", files[i]);
+
+		memset(&prev_line, 0, sizeof(prev_line));
+		while(parse_line(f, &line)){
+			if(!line.valid)
+				continue;
+
+			if((current == -1) || (strcmp(line.property, properties[current]) != 0)){
+				if(current != -1){
+					emit_line(&prev_line, &n);
+					printf("\n};\n");
+					printf("#define %s_count %d\n\n", properties[current], n);
+				}
+
+				prev_line.valid = 0;
+				current = -1;
+				for(j = 0; j < sizeof(properties)/sizeof(properties[0]); j++){
+					if(strcmp(line.property, properties[j]) == 0){
+						current = j;
+						n = 0;
+						printf("static struct unicode_range %s_ranges[] = {", properties[j]);
+						break;
+					}
+				}
+			}
+
+			if(current != -1){
+				if(prev_line.valid){
+					if(prev_line.to+1 == line.from)
+						line.from = prev_line.from;
+					else
+						emit_line(&prev_line, &n);
+				}
+				prev_line = line;
+			}
+		}
+		fclose(f);
+	}
+
+	printf("static void\n");
+	printf("unicode_get(enum unicode_property p, int *n, struct unicode_range **r)\n{\n");
+	printf("\tswitch(p){\n");
+	for(i = 0; i < sizeof(properties)/sizeof(properties[0]); i++){
+		printf("\tcase %s:\n", properties[i]);
+		printf("\t\t*n = %s_count;\n", properties[i]);
+		printf("\t\t*r = %s_ranges;\n", properties[i]);
+		printf("\t\tbreak;\n");
+	}
+	printf("\t}\n}\n\n");
+
+	free(buf);
+	return 0;
+}
+
+static void
+emit_line(struct line *line, int *n)
+{
+	printf("%s{0x%X, 0x%X}", (*n == 0) ? "\n\t" : ",\n\t", line->from, line->to);
+	(*n)++;
+}
+
+static int
+peek(FILE *f)
+{
+	int c;
+
+	c = fgetc(f);
+	ungetc(c, f);
+	return c;
+}
+
+static int
+parse_line(FILE *f, struct line *l)
+{
+	int c, i;
+
+	if(setjmp(env)){
+		l->valid = 0;
+		do
+			c = fgetc(f);
+		while(c != EOF && c != '\n');
+		return c != EOF;
+	}else{
+		l->valid = 1;
+	}
+
+	l->from = parse_int(f);
+	if(peek(f) == '.'){
+		parse_dots(f);
+		l->to = parse_int(f);
+	}else{
+		l->to = l->from;
+	}
+	parse_sep(f);
+
+	for(i = 0; ; i++){
+		c = fgetc(f);
+		l->property[i] = c;
+		if(c == ' ' || c == EOF){
+			if(i == 0)
+				longjmp(env, 1);
+			break;
+		}
+	}
+	l->property[i] = 0;
+
+	for(;;){
+		c = fgetc(f);
+		if(c == '\n' || c == EOF)
+			break;
+	}
+	return 1;
+}
+
+static int
+parse_int(FILE *f)
+{
+	int n, m, c;
+
+	n = 0;
+
+	for(;;){
+		c = peek(f);
+		if((c >= '0') && (c <= '9'))
+			m = c - '0';
+		else if((c >= 'A') && (c <= 'F'))
+			m = 10 + (c - 'A');
+		else if(c == '.' || c == ' ')
+			return n;
+		else
+			longjmp(env, 1);
+		fgetc(f);
+		n <<= 4;
+		n += m;
+	}
+}
+
+static void
+parse_dots(FILE *f)
+{
+	int i;
+
+	for(i = 0; i < 2; i++){
+		if(peek(f) != '.')
+			longjmp(env, 1);
+		fgetc(f);
+	}
+}
+
+static void
+parse_sep(FILE *f)
+{
+	while(peek(f) == ' ')
+		fgetc(f);
+
+	if(peek(f) == ';')
+		fgetc(f);
+	else
+		longjmp(env, 1);
+
+	while(peek(f) == ' ')
+		fgetc(f);
+}