Modify MAYBE_NUM usage and typeof function to return "strnum" only for actual numeric strings.

12 files changed, 54 insertions, 26 deletions

diff --git a/ChangeLog b/ChangeLog
index 3284747d..7d11fd5b 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,23 @@
 2016-07-06         Andrew J. Schorr     <aschorr@telemetry-investments.com>
 
+	* awk.h: Modify comments to indicate that MAYBE_NUM will now be
+	left enabled to indicate strnum values by the NUMBER|MAYBE_NUM
+	combination, whereas STRING|MAYBE_NUM indicates a potential strnum.
+	(fixtype): Modify MAYBE_NUM test to avoid calling force_number if
+	NUMCUR is already set.
+	* builtin.c (do_typeof): Call fixtype to resolve argument type.
+	This forces parsing of numeric strings, so there's a performance
+	penalty, but we must do this to give a correct result. The meaning
+	of "strnum" changes from "potential strnum" to "actual strnum".
+	* eval.c (set_TEXTDOMAIN): Remove some dead code left over from last
+	patch.
+	* int_array.c (is_integer): When a MAYBE_NUM is converted successfully
+	to a NUMBER, leave the MAYBE_NUM flag enabled.
+	* mpfr.c (mpg_force_number): Ditto.
+	* node.c (r_force_number): Ditto.
+
+2016-07-06         Andrew J. Schorr     <aschorr@telemetry-investments.com>
+
 	* awk.h: Modify stptr comment to indicate that all strings are now
 	NUL-terminated.
 	* builtin.c (do_mktime): Remove unnecessary logic to terminate
diff --git a/awk.h b/awk.h
index 09757b9f..9e1fae10 100644
--- a/awk.h
+++ b/awk.h
@@ -406,14 +406,16 @@ typedef struct exp_node {
 	 * 	b = a + 0	# Adds NUMCUR to a, since numeric value
 	 * 			# is now available. But the type hasn't changed!
 	 *
-	 * MAYBE_NUM is the joker.  It means "this is string data, but
-	 * the user may have really wanted it to be a number. If we have
-	 * to guess, like in a comparison, turn it into a number if the string
-	 * is indeed numeric."
+	 * MAYBE_NUM is the joker.  When STRING|MAYBE_NUM is set, it means
+	 * "this is string data, but the user may have really wanted it to be a
+	 * number. If we have to guess, like in a comparison, turn it into a
+	 * number if the string is indeed numeric."
 	 * For example,    gawk -v a=42 ....
 	 * Here, `a' gets STRING|STRCUR|MAYBE_NUM and then when used where
 	 * a number is needed, it gets turned into a NUMBER and STRING
-	 * is cleared.
+	 * is cleared. In that case, we leave the MAYBE_NUM in place, so
+	 * the combination NUMBER|MAYBE_NUM means it is a strnum a.k.a. a
+	 * "numeric string".
 	 *
 	 * WSTRCUR is for efficiency. If in a multibyte locale, and we
 	 * need to do something character based (substr, length, etc.)
@@ -1865,7 +1867,7 @@ fixtype(NODE *n)
 {
 	assert(n->type == Node_val || n->type == Node_typedregex);
 	if (n->type == Node_val) {
-		if ((n->flags & MAYBE_NUM) != 0)
+		if ((n->flags & (NUMCUR|MAYBE_NUM)) == MAYBE_NUM)
 			return force_number(n);
 		if ((n->flags & INTIND) != 0)
 			return force_string(n);
diff --git a/builtin.c b/builtin.c
index a21df18e..032f0ec7 100644
--- a/builtin.c
+++ b/builtin.c
@@ -3951,14 +3951,14 @@ do_typeof(int nargs)
 		break;
 	case Node_val:
 	case Node_var:
-		switch (arg->flags & (STRING|NUMBER|MAYBE_NUM)) {
+		switch (fixtype(arg)->flags & (STRING|NUMBER|MAYBE_NUM)) {
 		case STRING:
 			res = "string";
 			break;
 		case NUMBER:
 			res = "number";
 			break;
-		case STRING|MAYBE_NUM:
+		case NUMBER|MAYBE_NUM:
 			res = "strnum";
 			break;
 		case NUMBER|STRING:
diff --git a/eval.c b/eval.c
index aaabdb3a..cfb1d1e6 100644
--- a/eval.c
+++ b/eval.c
@@ -955,12 +955,10 @@ set_LINT()
 void
 set_TEXTDOMAIN()
 {
-	int len;
 	NODE *tmp;
 
 	tmp = TEXTDOMAIN_node->var_value = force_string(TEXTDOMAIN_node->var_value);
 	TEXTDOMAIN = tmp->stptr;
-	len = tmp->stlen;
 	/*
 	 * Note: don't call textdomain(); this value is for
 	 * the awk program, not for gawk itself.
diff --git a/int_array.c b/int_array.c
index 93e96d1f..937a91cf 100644
--- a/int_array.c
+++ b/int_array.c
@@ -142,7 +142,8 @@ is_integer(NODE *symbol, NODE *subs)
 	if (len == 1 && *cp != '-') {	/* single digit */
 		subs->numbr = (long) (*cp - '0');
 		if ((subs->flags & MAYBE_NUM) != 0) {
-			subs->flags &= ~(MAYBE_NUM|STRING);
+			/* leave MAYBE_NUM set */
+			subs->flags &= ~STRING;
 			subs->flags |= NUMBER;
 		}
 		subs->flags |= (NUMCUR|NUMINT);
@@ -158,7 +159,8 @@ is_integer(NODE *symbol, NODE *subs)
 
 	subs->numbr = l;
 	if ((subs->flags & MAYBE_NUM) != 0) {
-		subs->flags &= ~(MAYBE_NUM|STRING);
+		/* leave MAYBE_NUM set */
+		subs->flags &= ~STRING;
 		subs->flags |= NUMBER;
 	}
 	subs->flags |= NUMCUR;
diff --git a/mpfr.c b/mpfr.c
index cf7a7c21..b239e4b6 100644
--- a/mpfr.c
+++ b/mpfr.c
@@ -342,7 +342,8 @@ mpg_force_number(NODE *n)
 
 	if (force_mpnum(n, (do_non_decimal_data && ! do_traditional), true)) {
 		if ((n->flags & MAYBE_NUM) != 0) {
-			n->flags &= ~(MAYBE_NUM|STRING);
+			/* leave MAYBE_NUM set to indicate a strnum */
+			n->flags &= ~STRING;
 			n->flags |= NUMBER;
 		}
 	} else
diff --git a/node.c b/node.c
index 2cd88331..37aa9463 100644
--- a/node.c
+++ b/node.c
@@ -66,9 +66,9 @@ r_force_number(NODE *n)
 		return n;
 
 	/*
-	 * We should always set NUMCUR and clear MAYBE_NUM, and we may possibly
-	 * change STRING to NUMBER if MAYBE_NUM was set and it's a good numeric
-	 * string.
+	 * We should always set NUMCUR. If MAYBE_NUM is set and it's a
+	 * numeric string, we clear STRING and enable NUMBER, but if it's not
+	 * numeric, we disable MAYBE_NUM.
 	 */
 
 	/* All the conditionals are an attempt to avoid the expensive strtod */
@@ -166,7 +166,8 @@ badnum:
 
 goodnum:
 	if ((n->flags & MAYBE_NUM) != 0) {
-		n->flags &= ~(MAYBE_NUM|STRING);
+		/* leave MAYBE_NUM enabled to indicate that this is a strnum */
+		n->flags &= ~STRING;
 		n->flags |= NUMBER;
 	}
 	return n;
diff --git a/test/ChangeLog b/test/ChangeLog
index 5897906e..6821488c 100644
--- a/test/ChangeLog
+++ b/test/ChangeLog
@@ -1,3 +1,11 @@
+2016-07-06         Andrew J. Schorr     <aschorr@telemetry-investments.com>
+
+	* forcenum.awk: We no longer need to force the strnum conversion,
+	since typeof now does this automatically.
+	* forcenum.ok: Change "number" to "strnum" for the numeric strings.
+	* rebuild.in: Change input to include a strnum.
+	* rebuild.ok: Update results.
+
 2016-07-04         Andrew J. Schorr     <aschorr@telemetry-investments.com>
 
 	* Makefile.am (arrayind3): New test.
diff --git a/test/forcenum.awk b/test/forcenum.awk
index 54c536c9..1a7ddce7 100644
--- a/test/forcenum.awk
+++ b/test/forcenum.awk
@@ -1,8 +1,6 @@
 BEGIN {
-	# first, make some strnums
+	# make some strnums
 	nf = split("|5apple|+NaN| 6|0x1az|011Q|027", f, "|")
-	for (i = 1; i <= nf; i++) {
-		x = f[i]+0	# trigger strnum conversion to number or string
+	for (i = 1; i <= nf; i++)
 		printf "[%s] -> %g (type %s)\n", f[i], f[i], typeof(f[i])
-	}
 }
diff --git a/test/forcenum.ok b/test/forcenum.ok
index c74eefc7..a379db62 100644
--- a/test/forcenum.ok
+++ b/test/forcenum.ok
@@ -1,7 +1,7 @@
 [] -> 0 (type string)
 [5apple] -> 5 (type string)
-[+NaN] -> nan (type number)
-[ 6] -> 6 (type number)
+[+NaN] -> nan (type strnum)
+[ 6] -> 6 (type strnum)
 [0x1az] -> 26 (type string)
 [011Q] -> 9 (type string)
-[027] -> 23 (type number)
+[027] -> 23 (type strnum)
diff --git a/test/rebuild.in b/test/rebuild.in
index b2901ea9..2f16a825 100644
--- a/test/rebuild.in
+++ b/test/rebuild.in
@@ -1 +1 @@
-a b
+a 6.3
diff --git a/test/rebuild.ok b/test/rebuild.ok
index 29635279..0fe72e23 100644
--- a/test/rebuild.ok
+++ b/test/rebuild.ok
@@ -1,2 +1,2 @@
-test b
+test 6.3
 strnum