Start on a bool type for gawk.

8 files changed, 136 insertions, 23 deletions

diff --git a/ChangeLog b/ChangeLog
index 89767c0f..52b74a78 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,19 @@
+2021-03-05         Arnold D. Robbins     <arnold@skeeve.com>
+
+	Start on a bool type for gawk.
+
+	* awk.h (BOOL): New flag value.
+	(make_bool_node): Add declaration of new function.
+	(bool_val): Check for BOOL along with NUMBER.
+	* builtin.c (do_typeof): Add support for BOOL.
+	* eval.c (flags2str): Ditto.
+	* gawkapi.h (awk_val_type): Add AWK_BOOL.
+	(awk_value_t): Add awk_bool_t element to the union and macro for it.
+	(struct gawk_api): Update the table of request/return values.
+	* main.c (load_procinfo_bools): New function.
+	(load_procinfo): Call it.
+	* node.c (make_bool_node): New function.
+
 2021-02-13         Arnold D. Robbins     <arnold@skeeve.com>
 
 	* io.c (nextfile): Use the value of ARGC directly in the for
diff --git a/awk.h b/awk.h
index c9eec663..ee055f49 100644
--- a/awk.h
+++ b/awk.h
@@ -463,6 +463,7 @@ typedef struct exp_node {
 		XARRAY		= 0x10000,
 		NUMCONSTSTR	= 0x20000,	/* have string value for numeric constant */
 		REGEX           = 0x40000,	/* this is a typed regex */
+		BOOL            = 0x80000,	/* this is a boolean value */
 	} flags;
 	long valref;
 } NODE;
@@ -1714,6 +1715,7 @@ extern NODE *r_force_number(NODE *n);
 extern NODE *r_format_val(const char *format, int index, NODE *s);
 extern NODE *r_dupnode(NODE *n);
 extern NODE *make_str_node(const char *s, size_t len, int flags);
+extern NODE *make_bool_node(bool value);
 extern NODE *make_typed_regex(const char *re, size_t len);
 extern void *more_blocks(int id);
 extern int parse_escape(const char **string_ptr);
@@ -1995,7 +1997,7 @@ static inline bool
 boolval(NODE *t)
 {
 	(void) fixtype(t);
-	if ((t->flags & NUMBER) != 0)
+	if ((t->flags & (BOOL|NUMBER)) != 0)
 		return ! is_zero(t);
 	return (t->stlen > 0);
 }
diff --git a/bool.notes b/bool.notes
new file mode 100644
index 00000000..51ea7268
--- /dev/null
+++ b/bool.notes
@@ -0,0 +1,47 @@
+Fri Mar  5 13:42:44 IST 2021
+============================
+
+Design Notes for a Boolean Type in Gawk
+
+1. At first glance, two new variables named TRUE and FALSE would make sense
+to be predefined values of boolean type.  However, since it's likely that
+many awk programs already use them, instead we'll add two new entries
+to PROCINFO: variables PROCINFO["true"] and PROCINFO["false"] of type bool.
+
+Assigning from them copies the bool type, thus they are "factories" of bool
+variables.
+
+2. They have numeric values 1 and 0 respectively, and string values
+"TRUE" and "FALSE".  Thus they differ from other variables where a
+"false" value must be zero and null.
+
+This implies all of the following:
+
+	print(PROCINFO["true"]) --> "TRUE"
+	print(PROCINFO["false"]) --> "FALSE"
+	Same for %s in printf
+	Same for bool_var ""
+	printf %d gives 0/1
+
+3. Function  bool(val)  converts val to bool, returning Boolean TRUE or FALSE.
+
+4. typeof() returns "bool".
+
+5. Numeric operators treat booleans as numbers.  asort() sorts booleans before
+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
+	tolower	toupper
+
+7. Updates to API needed for an additional type, and the table
+for requested vs. returns.
+
+8. The following extensions need revising:
+
+	- JSON extension
+	- dump / read array extensions
+	- what else?
diff --git a/builtin.c b/builtin.c
index 1b612f4a..67a89bd1 100644
--- a/builtin.c
+++ b/builtin.c
@@ -4112,7 +4112,10 @@ do_typeof(int nargs)
 		}
 		break;
 	case Node_val:
-		switch (fixtype(arg)->flags & (STRING|NUMBER|USER_INPUT|REGEX)) {
+		switch (fixtype(arg)->flags & (STRING|NUMBER|USER_INPUT|REGEX|BOOL)) {
+		case BOOL:
+			res = "bool";
+			break;
 		case NUMBER:
 			res = "number";
 			break;
diff --git a/eval.c b/eval.c
index 640f939f..75c39327 100644
--- a/eval.c
+++ b/eval.c
@@ -455,6 +455,7 @@ flags2str(int flagval)
 		{ XARRAY, "XARRAY" },
 		{ NUMCONSTSTR, "NUMCONSTSTR" },
 		{ REGEX, "REGEX" },
+		{ BOOL, "BOOL" },
 		{ 0,	NULL },
 	};
 
diff --git a/gawkapi.h b/gawkapi.h
index 54130b15..3fd02df7 100644
--- a/gawkapi.h
+++ b/gawkapi.h
@@ -366,7 +366,8 @@ typedef enum {
 	AWK_STRNUM,
 	AWK_ARRAY,
 	AWK_SCALAR,		/* opaque access to a variable */
-	AWK_VALUE_COOKIE	/* for updating a previously created value */
+	AWK_VALUE_COOKIE,	/* for updating a previously created value */
+	AWK_BOOL
 } awk_valtype_t;
 
 /*
@@ -381,6 +382,7 @@ typedef struct awk_value {
 		awk_array_t	a;
 		awk_scalar_t	scl;
 		awk_value_cookie_t vc;
+		awk_bool_t      b;
 	} u;
 #define str_value	u.s
 #define strnum_value	str_value
@@ -391,6 +393,7 @@ typedef struct awk_value {
 #define array_cookie	u.a
 #define scalar_cookie	u.scl
 #define value_cookie	u.vc
+#define bool_value	u.b
 } awk_value_t;
 
 /*
@@ -567,26 +570,28 @@ typedef struct gawk_api {
 
 	                        +-------------------------------------------------------+
 	                        |                   Type of Actual Value:               |
-	                        +--------+--------+--------+--------+-------+-----------+
-	                        | String | Strnum | Number | Regex  | Array | Undefined |
-	+-----------+-----------+--------+--------+--------+--------+-------+-----------+
-	|           | String    | String | String | String | String | false | false     |
-	|           +-----------+--------+--------+--------+--------+-------+-----------+
-	|           | Strnum    | false  | Strnum | Strnum | false  | false | false     |
-	|           +-----------+--------+--------+--------+--------+-------+-----------+
-	|           | Number    | Number | Number | Number | false  | false | false     |
-	|           +-----------+--------+--------+--------+--------+-------+-----------+
-	|           | Regex     | false  | false  | false  | Regex  | false | false     |
-	|           +-----------+--------+--------+--------+--------+-------+-----------+
-	|   Type    | Array     | false  | false  | false  | false  | Array | false     |
-	| Requested +-----------+--------+--------+--------+--------+-------+-----------+
-	|           | Scalar    | Scalar | Scalar | Scalar | Scalar | false | false     |
-	|           +-----------+--------+--------+--------+--------+-------+-----------+
-	|           | Undefined | String | Strnum | Number | Regex  | Array | Undefined |
-	|           +-----------+--------+--------+--------+--------+-------+-----------+
-	|           | Value     | false  | false  | false  | false  | false | false     |
-	|           | Cookie    |        |        |        |        |       |           |
-	+-----------+-----------+--------+--------+--------+--------+-------+-----------+
+	                        +--------+--------+--------+--------+--------+-------+-----------+
+	                        | String | Strnum | Number | Regex  | Bool   | Array | Undefined |
+	+-----------+-----------+--------+--------+--------+--------+--------+-------+-----------+
+	|           | String    | String | String | String | String | String | false | false     |
+	|           +-----------+--------+--------+--------+--------+--------+-------+-----------+
+	|           | Strnum    | false  | Strnum | Strnum | false  | false  | false | false     |
+	|           +-----------+--------+--------+--------+--------+--------+-------+-----------+
+	|           | Number    | Number | Number | Number | false  | Number | false | false     |
+	|           +-----------+--------+--------+--------+--------+--------+-------+-----------+
+	|           | Regex     | false  | false  | false  | Regex  | false  | false | false     |
+	|           +-----------+--------+--------+--------+--------+--------+-------+-----------+
+	|           | Bool      | false  | false  | false  | false  | Bool   | false | false     |
+	|           +-----------+--------+--------+--------+--------+--------+-------+-----------+
+	|   Type    | Array     | false  | false  | false  | false  | false  | Array | false     |
+	| Requested +-----------+--------+--------+--------+--------+--------+-------+-----------+
+	|           | Scalar    | Scalar | Scalar | Scalar | Scalar | Scalar | false | false     |
+	|           +-----------+--------+--------+--------+--------+--------+-------+-----------+
+	|           | Undefined | String | Strnum | Number | Regex  | Bool   | Array | Undefined |
+	|           +-----------+--------+--------+--------+--------+--------+-------+-----------+
+	|           | Value     | false  | false  | false  | false  | false  | false | false     |
+	|           | Cookie    |        |        |        |        |        |       |           |
+	+-----------+-----------+--------+--------+--------+--------+--------+-------+-----------+
 	*/
 
 	/* Functions to handle parameters passed to the extension. */
diff --git a/main.c b/main.c
index 5b153bc3..9833d848 100644
--- a/main.c
+++ b/main.c
@@ -980,7 +980,23 @@ load_procinfo_argv()
 	// hook it into PROCINFO
 	sub = make_string("argv", 4);
 	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 */
@@ -1069,6 +1085,7 @@ load_procinfo()
 	}
 #endif
 	load_procinfo_argv();
+	load_procinfo_bools();
 	return PROCINFO_node;
 }
 
diff --git a/node.c b/node.c
index c22c06ab..dda90f0f 100644
--- a/node.c
+++ b/node.c
@@ -1082,3 +1082,25 @@ more_blocks(int id)
 }
 
 #endif
+
+/* make_bool_node --- make a boolean-valued node */
+
+extern NODE *
+make_bool_node(bool value)
+{
+	NODE *val;
+	const char *sval;
+	AWKNUM nval;
+
+	sval = (value ? "TRUE" : "FALSE");
+	nval = (value ? 1.0 : 0.0);
+
+	val = make_number(nval);
+	val->stptr = estrdup(sval, strlen(sval));
+	val->stlen = strlen(sval);
+	val->flags &= ~NUMBER;
+	val->flags |= NUMCUR|STRCUR|BOOL;
+	val->stfmt = STFMT_UNUSED;
+
+	return val;
+}