More progress on bool.

15 files changed, 74 insertions, 24 deletions

diff --git a/ChangeLog b/ChangeLog
index 52b74a78..e248bcc7 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,15 @@
+2021-03-08         Arnold D. Robbins     <arnold@skeeve.com>
+
+	* awk.h (warn_bool, do_bool): Add function declarations.
+	* awkgram.y (tokentab): Add entry for "bool" builtin.
+	* builtin.c (warn_bool): New function.
+	(do_index, do_substr, do_toupper, do_tolower, do_match,
+	do_length): Call it.
+	(do_sub): If first arg to sub/gsub is bool, fatal error.
+	(do_bool): New function.
+	* field.c (do_split, do_patsplit): Call warn_bool.
+	* main.c (load_procinfo_bools): Removed function and call.
+
 2021-03-05         Arnold D. Robbins     <arnold@skeeve.com>
 
 	Start on a bool type for gawk.
diff --git a/awk.h b/awk.h
index ee055f49..f4fa4d4d 100644
--- a/awk.h
+++ b/awk.h
@@ -1455,6 +1455,7 @@ extern bool is_identchar(int c);
 extern NODE *make_regnode(NODETYPE type, NODE *exp);
 extern bool validate_qualified_name(char *token);
 /* builtin.c */
+extern void warn_bool(const char *func, int argnum, NODE *n);
 extern double double_to_int(double d);
 extern NODE *do_exp(int nargs);
 extern NODE *do_fflush(int nargs);
@@ -1504,6 +1505,7 @@ extern int strncasecmpmbs(const unsigned char *,
 			  const unsigned char *, size_t);
 extern int sanitize_exit_status(int status);
 extern void check_symtab_functab(NODE *dest, const char *fname, const char *msg);
+extern NODE *do_bool(int nargs);
 /* debug.c */
 extern void init_debug(void);
 extern int debug_prog(INSTRUCTION *pc);
diff --git a/awkgram.c b/awkgram.c
index 9c2700a3..40c3cbd2 100644
--- a/awkgram.c
+++ b/awkgram.c
@@ -4779,6 +4779,7 @@ static const struct token tokentab[] = {
 {"asorti",	Op_builtin,	 LEX_BUILTIN,	GAWKX|A(1)|A(2)|A(3),	do_asorti,	0},
 {"atan2",	Op_builtin,	 LEX_BUILTIN,	NOT_OLD|A(2),	do_atan2,	MPF(atan2)},
 {"bindtextdomain",	Op_builtin,	 LEX_BUILTIN,	GAWKX|A(1)|A(2),	do_bindtextdomain,	0},
+{"bool",	Op_builtin,    LEX_BUILTIN,	GAWKX|A(1),	do_bool,	0},
 {"break",	Op_K_break,	 LEX_BREAK,	0,		0,	0},
 {"case",	Op_K_case,	 LEX_CASE,	GAWKX,		0,	0},
 {"close",	Op_builtin,	 LEX_BUILTIN,	NOT_OLD|A(1)|A(2),	do_close,	0},
diff --git a/awkgram.y b/awkgram.y
index 6dfbe7e0..3fe0272b 100644
--- a/awkgram.y
+++ b/awkgram.y
@@ -2277,6 +2277,7 @@ static const struct token tokentab[] = {
 {"asorti",	Op_builtin,	 LEX_BUILTIN,	GAWKX|A(1)|A(2)|A(3),	do_asorti,	0},
 {"atan2",	Op_builtin,	 LEX_BUILTIN,	NOT_OLD|A(2),	do_atan2,	MPF(atan2)},
 {"bindtextdomain",	Op_builtin,	 LEX_BUILTIN,	GAWKX|A(1)|A(2),	do_bindtextdomain,	0},
+{"bool",	Op_builtin,    LEX_BUILTIN,	GAWKX|A(1),	do_bool,	0},
 {"break",	Op_K_break,	 LEX_BREAK,	0,		0,	0},
 {"case",	Op_K_case,	 LEX_CASE,	GAWKX,		0,	0},
 {"close",	Op_builtin,	 LEX_BUILTIN,	NOT_OLD|A(1)|A(2),	do_close,	0},
diff --git a/bool.notes b/bool.notes
index 51ea7268..df3ec69d 100644
--- a/bool.notes
+++ b/bool.notes
@@ -33,8 +33,13 @@ numbers, and false before true.
 6. These string function generate a runtime fatal error
 if given an argument / target of boolean type:
 
-	substr	match	index	gensub	gsub
-	sub	length	split	patsplit
+	gsub sub
+
+These functions merely treat the value as a string
+but issue a lint warning.
+
+	substr	match	index	gensub
+	length	split	patsplit
 	tolower	toupper
 
 7. Updates to API needed for an additional type, and the table
diff --git a/builtin.c b/builtin.c
index 67a89bd1..1158fad2 100644
--- a/builtin.c
+++ b/builtin.c
@@ -381,6 +381,8 @@ do_index(int nargs)
 
 	s1 = force_string(s1);
 	s2 = force_string(s2);
+	warn_bool("index", 1, s1);
+	warn_bool("index", 2, s2);
 
 	p1 = s1->stptr;
 	p2 = s2->stptr;
@@ -552,6 +554,7 @@ do_length(int nargs)
 	if (do_lint && (fixtype(tmp)->flags & STRING) == 0)
 		lintwarn(_("%s: received non-string argument"), "length");
 	tmp = force_string(tmp);
+	warn_bool("length", 1, tmp);
 
 	if (gawk_mb_cur_max > 1) {
 		tmp = force_wstring(tmp);
@@ -1779,6 +1782,16 @@ do_sqrt(int nargs)
 	return make_number((AWKNUM) sqrt(arg));
 }
 
+/* warn_bool --- warn that bool parameter is used as a string */
+
+void
+warn_bool(const char *func, int argnum, NODE *n)
+{
+	if (do_lint && (n->flags & BOOL) != 0)
+		lintwarn(_("%s: argument %d of type bool used as a string"),
+			func, argnum);
+}
+
 /* do_substr --- do the substr function */
 
 NODE *
@@ -1802,6 +1815,7 @@ do_substr(int nargs)
 	DEREF(t1);
 
 	t1 = POP_STRING();
+	warn_bool("substr", 1, t1);
 
 	if (nargs == 3) {
 		if (! (d_length >= 1)) {
@@ -2407,6 +2421,7 @@ do_tolower(int nargs)
 	NODE *t1, *t2;
 
 	t1 = POP_SCALAR();
+	warn_bool("tolower", 1, t1);
 	if (do_lint && (fixtype(t1)->flags & STRING) == 0)
 		lintwarn(_("%s: received non-string argument"), "tolower");
 	t1 = force_string(t1);
@@ -2438,6 +2453,7 @@ do_toupper(int nargs)
 	NODE *t1, *t2;
 
 	t1 = POP_SCALAR();
+	warn_bool("toupper", 1, t1);
 	if (do_lint && (fixtype(t1)->flags & STRING) == 0)
 		lintwarn(_("%s: received non-string argument"), "toupper");
 	t1 = force_string(t1);
@@ -2662,6 +2678,7 @@ do_match(int nargs)
 	tre = POP();
 	rp = re_update(tre);
 	t1 = POP_STRING();
+	warn_bool("mastch", 1, t1);
 
 	rstart = research(rp, t1->stptr, 0, t1->stlen, RE_NEED_START);
 	if (rstart >= 0) {	/* match succeded */
@@ -2882,6 +2899,7 @@ do_sub(int nargs, unsigned int flags)
 		rp = re_update(tmp);
 
 		target = POP_STRING();	/* original string */
+		warn_bool("gensub", 3, target);
 
 		glob_flag = POP_SCALAR();	/* value of global flag */
 		if (   (glob_flag->flags & STRING) != 0
@@ -2924,6 +2942,10 @@ do_sub(int nargs, unsigned int flags)
 		}
 	}
 
+	if ((target->flags & BOOL) != 0)
+		fatal(_("%s: target cannot be of type bool"),
+			(flags & GSUB) != 0 ? "gsub" : "sub");
+
 	global = (how_many == -1);
 
 	rep_node = POP_STRING();	/* replacement text */
@@ -4324,3 +4346,18 @@ check_symtab_functab(NODE *dest, const char *fname, const char *msg)
 	else if (dest == func_table)
 		fatal(msg, fname, "FUNCTAB");
 }
+
+/* do_bool --- create boolean values */
+
+NODE *
+do_bool(int nargs)
+{
+	NODE *tmp;
+	bool result;
+
+	tmp = POP_SCALAR();
+	result = boolval(tmp);
+	DEREF(tmp);
+
+	return make_bool_node(result);
+}
diff --git a/field.c b/field.c
index c21046b9..2bd58863 100644
--- a/field.c
+++ b/field.c
@@ -1019,6 +1019,7 @@ do_split(int nargs)
 	assoc_clear(arr);
 
 	src = TOP_STRING();
+	warn_bool("split", 1, src);
 	if (src->stlen == 0) {
 		/*
 		 * Skip the work if first arg is the null string.
@@ -1096,6 +1097,7 @@ do_patsplit(int nargs)
 			_("%s: cannot use %s as second argument"));
 
 	src = TOP_STRING();
+	warn_bool("patsplit", 1, src);
 
 	if ((sep->flags & REGEX) != 0)
 		sep = sep->typed_re;
diff --git a/main.c b/main.c
index 9833d848..76f28289 100644
--- a/main.c
+++ b/main.c
@@ -982,23 +982,6 @@ load_procinfo_argv()
 	assoc_set(PROCINFO_node, sub, argv_array);
 }
 
-/* load_procinfo_bools --- populate PROCINFO["true"] and PROCINFO["false"] */
-
-static void
-load_procinfo_bools()
-{
-	NODE *sub;
-	NODE *val;
-
-	val = make_bool_node(false);
-	sub = make_string("false", 5);
-	assoc_set(PROCINFO_node, sub, val);
-
-	val = make_bool_node(true);
-	sub = make_string("true", 4);
-	assoc_set(PROCINFO_node, sub, val);
-}
-
 /* load_procinfo --- populate the PROCINFO array */
 
 static NODE *
@@ -1085,7 +1068,6 @@ load_procinfo()
 	}
 #endif
 	load_procinfo_argv();
-	load_procinfo_bools();
 	return PROCINFO_node;
 }
 
diff --git a/test/ChangeLog b/test/ChangeLog
index 53fe5627..2604c270 100644
--- a/test/ChangeLog
+++ b/test/ChangeLog
@@ -1,3 +1,8 @@
+2021-03-08         Arnold D. Robbins     <arnold@skeeve.com>
+
+	* dumpvars.ok, functab5.ok, id.ok, intest.awk, symtab11.ok,
+	symtab8.ok: Updated after code changes.
+
 2021-02-13         Arnold D. Robbins     <arnold@skeeve.com>
 
 	* Makefile.am (EXTRA_DIST): argcasfile, new test.
diff --git a/test/dumpvars.ok b/test/dumpvars.ok
index 85d1c859..7caecd35 100644
--- a/test/dumpvars.ok
+++ b/test/dumpvars.ok
@@ -9,7 +9,7 @@ FILENAME: "-"
 FNR: 3
 FPAT: "[^[:space:]]+"
 FS: " "
-FUNCTAB: array, 41 elements
+FUNCTAB: array, 42 elements
 IGNORECASE: 0
 LINT: 0
 NF: 1
diff --git a/test/functab5.ok b/test/functab5.ok
index 9ac4295d..ef110989 100644
--- a/test/functab5.ok
+++ b/test/functab5.ok
@@ -3,6 +3,7 @@ asort'
 asorti'
 atan2'
 bindtextdomain'
+bool'
 chdir'
 close'
 compl'
diff --git a/test/id.ok b/test/id.ok
index ab4df74a..f990546f 100644
--- a/test/id.ok
+++ b/test/id.ok
@@ -34,6 +34,7 @@ asort -> builtin
 asorti -> builtin
 atan2 -> builtin
 bindtextdomain -> builtin
+bool -> builtin
 close -> builtin
 compl -> builtin
 cos -> builtin
diff --git a/test/intest.awk b/test/intest.awk
index f030d07a..18e0cc4d 100644
--- a/test/intest.awk
+++ b/test/intest.awk
@@ -1,4 +1,4 @@
 BEGIN {
-	bool = ((b = 1) in c);
-	print bool, b	# gawk-3.0.1 prints "0 "; should print "0 1"
+	bool_result = ((b = 1) in c);
+	print bool_result, b	# gawk-3.0.1 prints "0 "; should print "0 1"
 } 
diff --git a/test/symtab11.ok b/test/symtab11.ok
index 7d4be46c..da2cfcba 100644
--- a/test/symtab11.ok
+++ b/test/symtab11.ok
@@ -37,6 +37,7 @@ BEGIN -- Functab is next
 [asorti] = asorti
 [atan2] = atan2
 [bindtextdomain] = bindtextdomain
+[bool] = bool
 [close] = close
 [compl] = compl
 [cos] = cos
diff --git a/test/symtab8.ok b/test/symtab8.ok
index da29b585..0cf40fe9 100644
--- a/test/symtab8.ok
+++ b/test/symtab8.ok
@@ -9,7 +9,7 @@ FIELDWIDTHS: ""
 FNR: 1
 FPAT: "[^[:space:]]+"
 FS: " "
-FUNCTAB: array, 41 elements
+FUNCTAB: array, 42 elements
 IGNORECASE: 0
 LINT: 0
 NF: 1