summaryrefslogtreecommitdiff
path: root/regress
diff options
context:
space:
mode:
authorrillig <rillig>2005-11-19 22:55:24 +0000
committerrillig <rillig>2005-11-19 22:55:24 +0000
commit9ee12e26fbbea62b5e5a07c1be747b7bd808c1c8 (patch)
treee1dce724a9bbd109bcde13a4a0ba4607e89ef293 /regress
parenta0488d008660f327281d98128d676775f556bb23 (diff)
downloadpkgsrc-9ee12e26fbbea62b5e5a07c1be747b7bd808c1c8.tar.gz
Don't require too much from a simple awk. 6144 bytes per line must be
enough. At least for Solaris.
Diffstat (limited to 'regress')
-rw-r--r--regress/tools/files/awk-test.sh9
1 files changed, 6 insertions, 3 deletions
diff --git a/regress/tools/files/awk-test.sh b/regress/tools/files/awk-test.sh
index 07e3b3b2c28..828904ec209 100644
--- a/regress/tools/files/awk-test.sh
+++ b/regress/tools/files/awk-test.sh
@@ -1,5 +1,5 @@
#! /bin/sh
-# $NetBSD: awk-test.sh,v 1.4 2005/11/19 22:03:54 rillig Exp $
+# $NetBSD: awk-test.sh,v 1.5 2005/11/19 22:55:24 rillig Exp $
#
set -e
@@ -55,10 +55,13 @@ done
#
# Passing strings from stdin to stdout. awk should be able to handle at
-# least 2^20 characters per line.
+# least 2^12 characters per line.
+#
+# Solaris 9 /usr/bin/awk: 2559 bytes
+# Solaris 9 /usr/bin/nawk: 6144 bytes
#
line="a"
-for i in 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20; do
+for i in 0 1 2 3 4 5 6 7 8 9 10 11 12; do
test_passline "line.2^${i}" "${line}"
line="${line}${line}"
done