Re: [bug?] word based string subscripting past the end

[dana <dana@xxxxxxx>]

On Sun 14 Jun 2026, at 12:15, Stephane Chazelas wrote:
> I'd have expected an empty output like for character
> subscripting or array subscripting:

so would i. and i can't find anything in the documentation that contradicts
that expectation

this makes it work like $arr[n] does

dana


diff --git a/Src/params.c b/Src/params.c
index e5bbe9cb9..c361527be 100644
--- a/Src/params.c
+++ b/Src/params.c
@@ -1621,15 +1621,15 @@ getarg(char **str, int *inv, Value v, int a2, zlong *w,
 	}
 	if (word && !v->scanflags) {
 	    s = t = getstrvalue(v);
+	    if (!s || !*s)
+		return 0;
+
 	    i = wordcount(s, sep, 0);
 	    if (r < 0)
 		r += i + 1;
-	    if (r < 1)
-		r = 1;
-	    if (r > i)
-		r = i;
-	    if (!s || !*s)
+	    if (r < 1 || r > i)
 		return 0;
+
 	    while ((d = findword(&s, sep)) && --r);
 	    if (!d)
 		return 0;
diff --git a/Test/D06subscript.ztst b/Test/D06subscript.ztst
index 57cdc027c..42748577d 100644
--- a/Test/D06subscript.ztst
+++ b/Test/D06subscript.ztst
@@ -308,3 +308,18 @@ F:In math, (($i)) should be the same as ((i)), see workers/47748.
 F:Regression test for inconsistency of failed (i) on zero-length string
 >2
 >1
+
+  string='foo bar baz'
+  for 1 in 0 1 2 3 4 -1 -2 -3 -4; do
+    print -r - $1 $string[(w)$1]
+  done
+0:(w)
+>0
+>1 foo
+>2 bar
+>3 baz
+>4
+>-1 baz
+>-2 bar
+>-3 foo
+>-4