Get Perl interpreter from the PATH

git-svn-id: http://svn.code.sf.net/p/netpbm/code/trunk@2982 9d0c8265-081b-0410-96cb-a4ca84ce46f8

1 files changed, 24 insertions, 0 deletions

diff --git a/converter/pbm/pbmtox10bm b/converter/pbm/pbmtox10bm
index 6e1a12a2..deb3aeab 100644
--- a/converter/pbm/pbmtox10bm
+++ b/converter/pbm/pbmtox10bm
@@ -1,3 +1,27 @@
+#!/bin/sh
+
+##############################################################################
+# This is essentially a Perl program.  We exec the Perl interpreter specifying
+# this same file as the Perl program and use the -x option to cause the Perl
+# interpreter to skip down to the Perl code.  The reason we do this instead of
+# just making /usr/bin/perl the script interpreter (instead of /bin/sh) is
+# that the user may have multiple Perl interpreters and the one he wants to
+# use is properly located in the PATH.  The user's choice of Perl interpreter
+# may be crucial, such as when the user also has a PERL5LIB environment
+# variable and it selects modules that work with only a certain main
+# interpreter program.
+#
+# An alternative some people use is to have /usr/bin/env as the script
+# interpreter.  We don't do that because we think the existence and
+# compatibility of /bin/sh is more reliable.
+#
+# Note that we aren't concerned about efficiency because the user who needs
+# high efficiency can use directly the programs that this program invokes.
+#
+##############################################################################
+
+exec perl -w -x -S -- "$0" "$@"
+
 #! /usr/bin/perl
 
 #============================================================================