Latest research awk from June 29, 1996

10 files changed, 443 insertions, 105 deletions

diff --git a/usr.bin/awk/FIXES b/usr.bin/awk/FIXES
index 9ac0b21ac16..9cda6ca12b3 100644
--- a/usr.bin/awk/FIXES
+++ b/usr.bin/awk/FIXES
@@ -458,3 +458,20 @@ May 27, 1996:
 
 	thanks to jeffrey friedl for several of these.
 
+May 28, 1996:
+	fixed appalling but apparently unimportant bug in parsing octal
+	numbers in reg exprs.
+
+	explicit hex in reg exprs now limited to 2 chars: \xa, \xaa.
+
+Jun 28, 1996:
+	changed field-splitting to conform to posix definition: fields are
+	split using the value of FS at the time of input; it used to be
+	the value when the field or NF was first referred to, a much less
+	predictable definition.  thanks to arnold robbins for encouragement
+	to do the right thing.
+
+Jun 29, 1996:
+	fixed awful bug in new field splitting; didn't get all the places
+	where input was done.
+
diff --git a/usr.bin/awk/OpenBSD-PATCHES b/usr.bin/awk/OpenBSD-PATCHES
new file mode 100644
index 00000000000..b011fb71ed0
--- /dev/null
+++ b/usr.bin/awk/OpenBSD-PATCHES
@@ -0,0 +1,384 @@
+These are the changes made from Research Awk to allow it to
+build on OpenBSD with flex.  Below is the inof that came
+with the original set of diffs (from 4.4BSD).  ytab.* was
+changed to awkgram.* for make's benefit.
+
+ - todd
+
+From: Vern Paxson <vern@daffy.ee.lbl.gov>
+
+I've ported Research Awk to flex and tested it moderately.  Note that I
+didn't have time to support the nice error-message context stuff (where
+it points out exactly where on a line it thinks an error occurred), as
+the original code made a lot of assumptions regarding the internal
+buffering of a lex scanner that are no longer valid with flex.  Also, the
+sources had a function called "isnumber" which conflicted with a macro by
+the same name in <ctype.h>, so I changed its name to is_a_number.
+
+Let me know if you find more problems.
+
+		Vern
+
+--- /home/millert/tmp/awk/awk.1	Sun Jan 19 18:06:25 1997
++++ awk.1	Sun Jan 19 17:51:39 1997
+@@ -1,3 +1,4 @@
++.\"	$OpenBSD: OpenBSD-PATCHES,v 1.1 1997/01/20 19:43:16 millert Exp $
+ .de EX
+ .nf
+ .ft CW
+@@ -13,7 +14,7 @@
+ .SH NAME
+ awk \- pattern-directed scanning and processing language
+ .SH SYNOPSIS
+-.B awk
++.B awk|nawk
+ [
+ .BI \-F
+ .I fs
+--- /home/millert/tmp/awk/awklex.l	Sun Jan 19 18:06:24 1997
++++ awklex.l	Sun Jan 19 18:00:52 1997
+@@ -29,13 +29,15 @@
+    may not be preserved in other implementations of lex.
+ */
+ 
++#ifndef FLEX_SCANNER
+ #undef	input	/* defeat lex */
+ #undef	unput
++#endif /* !FLEX_SCANNER */
+ 
+ #include <stdlib.h>
+ #include <string.h>
+ #include "awk.h"
+-#include "ytab.h"
++#include "awkgram.h"
+ 
+ extern YYSTYPE	yylval;
+ extern int	infunc;
+@@ -60,6 +62,20 @@
+ char	*s;
+ Gstring	*gs = 0;	/* initialized in main() */
+ int	cflag;
++
++#ifdef FLEX_SCANNER
++static	int	my_input( YY_CHAR *buf, int max_size );
++
++#undef YY_INPUT
++#define YY_INPUT(buf,result,max_size) result = my_input(buf, max_size);
++
++#undef YY_USER_INIT
++#define YY_USER_INIT init_input_source();
++
++#define	FIRST	((yy_start - 1) / 2)
++#else /* FLEX_SCANNER */
++#define	FIRST	(yybgin - yysvec - 1)
++#endif /* FLEX_SCANNER */
+ %}
+ 
+ A	[a-zA-Z_]
+@@ -70,7 +86,7 @@
+ WS	[ \t]
+ 
+ %%
+-	switch (yybgin-yysvec-1) {	/* witchcraft */
++	switch (FIRST) {	/* witchcraft */
+ 	case 0:
+ 		BEGIN A;
+ 		break;
+@@ -116,14 +132,20 @@
+ 
+ <A>"$"{D}+	{ yylval.cp = fieldadr(atoi(yytext+1)); RET(FIELD); }
+ <A>"$NF"	{ unputstr("(NF)"); return(INDIRECT); }
+-<A>"$"{A}{B}*	{ int c, n;
+-		  c = input(); unput(c);
+-		  if (c == '(' || c == '[' || (infunc && (n=isarg(yytext+1)) >= 0)) {
+-			unputstr(yytext+1);
+-			return(INDIRECT);
++<A>"$"{A}{B}*	{
++		  int c;
++		  char *yytext_copy = strdup(yytext);
++		  c = input(); unput(c);	/* look for '(' or '[' */
++		  if (c == '(' || c == '[' ||
++		      infunc && isarg(yytext_copy+1) >= 0) {
++			  unputstr(yytext_copy+1);
++			  free(yytext_copy);
++			  return(INDIRECT);
+ 		  } else {
+-			yylval.cp = setsymtab(yytext+1, "", 0.0, STR|NUM, symtab);
+-			RET(IVAR);
++			  yylval.cp =
++				setsymtab(yytext_copy+1,"",0.0,STR|NUM,symtab);
++			  free(yytext_copy);
++			  RET(IVAR);
+ 		  }
+ 		}
+ <A>"$"		{ RET(INDIRECT); }
+@@ -173,12 +195,15 @@
+ <A>fflush	{ yylval.i = FFLUSH; RET(BLTIN); }
+ 
+ <A>{A}{B}*	{ int n, c;
++		  char *yytext_copy = strdup(yytext);
+ 		  c = input(); unput(c);	/* look for '(' */
+-		  if (c != '(' && infunc && (n=isarg(yytext)) >= 0) {
++		  if (c != '(' && infunc && (n=isarg(yytext_copy)) >= 0) {
+ 			yylval.i = n;
++			free(yytext_copy);
+ 			RET(ARG);
+ 		  } else {
+-			yylval.cp = setsymtab(yytext, "", 0.0, STR|NUM, symtab);
++			yylval.cp = setsymtab(yytext_copy, "", 0.0, STR|NUM, symtab);
++			free(yytext_copy);
+ 			if (c == '(') {
+ 				RET(CALL);
+ 			} else {
+@@ -237,6 +262,32 @@
+ 	caddreset(gs);
+ }
+ 
++#ifdef FLEX_SCANNER
++static int my_input( YY_CHAR *buf, int max_size )
++{
++	extern uschar *lexprog;
++
++	if ( lexprog ) {		/* awk '...' */
++		int num_chars = strlen( lexprog );
++		if ( num_chars > max_size )
++			{
++			num_chars = max_size;
++			strncpy( buf, lexprog, num_chars );
++			}
++		else
++			strcpy( buf, lexprog );
++		lexprog += num_chars;
++		return num_chars;
++
++	} else {			/* awk -f ... */
++		int c = pgetc();
++		if (c == EOF)
++			return 0;
++		buf[0] = c;
++		return 1;
++	}
++}
++#else /* FLEX_SCANNER */
+ /* input() and unput() are transcriptions of the standard lex
+    macros for input and output with additions for error message
+    printing.  God help us all if someone changes how lex works.
+@@ -275,7 +326,7 @@
+ 	if (--ep < ebuf)
+ 		ep = ebuf + sizeof(ebuf) - 1;
+ }
+-
++#endif /* FLEX_SCANNER */
+ 
+ void unputstr(char *s)	/* put a string back on input */
+ {
+@@ -285,6 +336,11 @@
+ 		unput(s[i]);
+ }
+ 
++int lex_input()
++{
++	return input();
++}
++
+ /* growing-string code */
+ 
+ const int CBUFLEN = 400;
+@@ -330,3 +386,20 @@
+ 	free((void *) gs->cbuf);
+ 	free((void *) gs);
+ }
++
++#ifdef FLEX_SCANNER
++void init_input_source(void)
++{
++	extern int curpfile;
++	extern char *pfile[];
++
++	if (yyin == NULL) {
++		if (pfile[curpfile] == 0)
++			return;
++		if (strcmp((char *) pfile[curpfile], "-") == 0)
++			yyin = stdin;
++		else if ((yyin = fopen((char *) pfile[curpfile], "r")) == NULL)
++			ERROR "can't open file %s", pfile[curpfile] FATAL;
++	}
++}
++#endif
+--- /home/millert/tmp/awk/b.c	Sun Jan 19 18:06:24 1997
++++ b.c	Sun Jan 19 18:00:55 1997
+@@ -31,7 +31,7 @@
+ #include <string.h>
+ #include <stdlib.h>
+ #include "awk.h"
+-#include "ytab.h"
++#include "awkgram.h"
+ 
+ #define	HAT	(NCHARS-1)	/* matches ^ in regular expr */
+ 				/* NCHARS is 2**n */
+--- /home/millert/tmp/awk/lib.c	Sun Jan 19 18:06:24 1997
++++ lib.c	Sun Jan 19 18:01:45 1997
+@@ -29,7 +29,7 @@
+ #include <errno.h>
+ #include <stdlib.h>
+ #include "awk.h"
+-#include "ytab.h"
++#include "awkgram.h"
+ 
+ FILE	*infile	= NULL;
+ char	*file	= "";
+@@ -431,7 +431,7 @@
+ 
+ 	if (beenhere++)
+ 		return;
+-	while ((c = input()) != EOF && c != '\0')
++	while ((c = lex_input()) != EOF && c != '\0')
+ 		bclass(c);
+ 	bcheck2(bracecnt, '{', '}');
+ 	bcheck2(brackcnt, '[', ']');
+@@ -479,6 +479,7 @@
+ 
+ void eprint(void)	/* try to print context around error */
+ {
++#if 0
+ 	char *p, *q;
+ 	int c;
+ 	static int been_here = 0;
+@@ -511,6 +512,7 @@
+ 		}
+ 	putc('\n', stderr);
+ 	ep = ebuf;
++#endif
+ }
+ 
+ void bclass(int c)
+--- /home/millert/tmp/awk/main.c	Sun Jan 19 18:06:24 1997
++++ main.c	Sun Jan 19 18:00:57 1997
+@@ -27,11 +27,12 @@
+ #define DEBUG
+ #include <stdio.h>
+ #include <ctype.h>
++#include <locale.h>
+ #include <stdlib.h>
+ #include <string.h>
+ #include <signal.h>
+ #include "awk.h"
+-#include "ytab.h"
++#include "awkgram.h"
+ 
+ extern	char	**environ;
+ extern	int	nfields;
+@@ -53,7 +54,12 @@
+ 	char *fs = NULL, *marg;
+ 	int temp;
+ 
+-	cmdname = argv[0];
++	setlocale(LC_ALL, "");
++
++	if ((cmdname = strrchr(argv[0], '/')) != NULL)
++		cmdname++;
++	else
++		cmdname = argv[0];
+ 	if (argc == 1) {
+ 		fprintf(stderr, "Usage: %s [-f programfile | 'program'] [-Ffieldsep] [-v var=value] [-mf n] [-mr n] [files]\n", cmdname);
+ 		exit(1);
+--- /home/millert/tmp/awk/maketab.c	Sun Jan 19 18:06:24 1997
++++ maketab.c	Sun Jan 19 18:01:08 1997
+@@ -25,14 +25,14 @@
+ /*
+  * this program makes the table to link function names
+  * and type indices that is used by execute() in run.c.
+- * it finds the indices in ytab.h, produced by yacc.
++ * it finds the indices in awkgram.h, produced by yacc.
+  */
+ 
+ #include <stdio.h>
+ #include <string.h>
+ #include <stdlib.h>
+ #include "awk.h"
+-#include "ytab.h"
++#include "awkgram.h"
+ 
+ struct xx
+ {	int token;
+@@ -120,12 +120,12 @@
+ 
+ 	printf("#include <stdio.h>\n");
+ 	printf("#include \"awk.h\"\n");
+-	printf("#include \"ytab.h\"\n\n");
++	printf("#include \"awkgram.h\"\n\n");
+ 	for (i = SIZE; --i >= 0; )
+ 		names[i] = "";
+ 
+-	if ((fp = fopen("ytab.h", "r")) == NULL) {
+-		fprintf(stderr, "maketab can't open ytab.h!\n");
++	if ((fp = fopen("awkgram.h", "r")) == NULL) {
++		fprintf(stderr, "maketab can't open awkgram.h!\n");
+ 		exit(1);
+ 	}
+ 	printf("static char *printname[%d] = {\n", SIZE);
+--- /home/millert/tmp/awk/parse.c	Sun Jan 19 18:06:24 1997
++++ parse.c	Sun Jan 19 18:01:11 1997
+@@ -27,7 +27,7 @@
+ #include <string.h>
+ #include <stdlib.h>
+ #include "awk.h"
+-#include "ytab.h"
++#include "awkgram.h"
+ 
+ Node *nodealloc(int n)
+ {
+--- /home/millert/tmp/awk/proto.h	Sun Jan 19 18:06:24 1997
++++ proto.h	Sun Jan 19 17:51:39 1997
+@@ -22,7 +22,6 @@
+ USE OR PERFORMANCE OF THIS SOFTWARE.
+ ****************************************************************/
+ 
+-extern	int	yywrap(void);
+ extern	void	setfname(Cell *);
+ extern	int	constnode(Node *);
+ extern	char	*strnode(Node *);
+@@ -31,12 +30,8 @@
+ 
+ extern	int	yylex(void);
+ extern	void	startreg(void);
+-extern	int	input(void);
+-extern	void	unput(int);
++extern	int	lex_input(void);
+ extern	void	unputstr(char *);
+-extern	int	yylook(void);
+-extern	int	yyback(int *, int);
+-extern	int	yyinput(void);
+ 
+ extern	fa	*makedfa(char *, int);
+ extern	fa	*mkdfa(char *, int);
+@@ -65,6 +60,7 @@
+ extern	void	freefa(fa *);
+ 
+ extern	int	pgetc(void);
++extern	void	init_input_source(void);
+ 
+ extern	Node	*nodealloc(int);
+ extern	Node	*exptostat(Node *);
+--- /home/millert/tmp/awk/run.c	Sun Jan 19 18:06:24 1997
++++ run.c	Sun Jan 19 18:01:15 1997
+@@ -31,7 +31,7 @@
+ #include <stdlib.h>
+ #include <time.h>
+ #include "awk.h"
+-#include "ytab.h"
++#include "awkgram.h"
+ 
+ #define tempfree(x)	if (istemp(x)) tfree(x); else
+ 
+--- /home/millert/tmp/awk/tran.c	Sun Jan 19 18:06:24 1997
++++ tran.c	Sun Jan 19 18:01:18 1997
+@@ -29,7 +29,7 @@
+ #include <string.h>
+ #include <stdlib.h>
+ #include "awk.h"
+-#include "ytab.h"
++#include "awkgram.h"
+ 
+ #define	FULLTAB	2	/* rehash when table gets this x full */
+ #define	GROWTAB 4	/* grow table by this factor */
diff --git a/usr.bin/awk/awk.1 b/usr.bin/awk/awk.1
index 54b4d3b79d8..c2328aee911 100644
--- a/usr.bin/awk/awk.1
+++ b/usr.bin/awk/awk.1
@@ -1,4 +1,4 @@
-.\"	$OpenBSD: awk.1,v 1.2 1996/07/06 05:40:03 etheisen Exp $
+.\"	$OpenBSD: awk.1,v 1.3 1997/01/20 19:43:18 millert Exp $
 .de EX
 .nf
 .ft CW
@@ -426,6 +426,7 @@ Variable names with special meanings:
 .TP
 .B CONVFMT
 conversion format used when converting numbers
+(default
 .BR "%.6g" )
 .TP
 .B FS
@@ -484,12 +485,14 @@ Thus local variables may be created by providing excess parameters in
 the function definition.
 .SH EXAMPLES
 .TP
-.B
+.EX
 length($0) > 72
+.EE
 Print lines longer than 72 characters.
 .TP
-.B
+.EX
 { print $2, $1 }
+.EE
 Print first two fields in opposite order.
 .PP
 .EX
@@ -510,8 +513,9 @@ END	{ print "sum is", s, " average is", s/NR }
 .IP
 Add up first column, print sum and average.
 .TP
-.B
+.EX
 /start/, /stop/
+.EE
 Print all lines between start/stop pairs.
 .PP
 .EX
diff --git a/usr.bin/awk/awk.h b/usr.bin/awk/awk.h
index 731f6f13cab..213e43302c8 100644
--- a/usr.bin/awk/awk.h
+++ b/usr.bin/awk/awk.h
@@ -67,6 +67,7 @@ extern int	lineno;		/* line number in awk program */
 extern int	errorflag;	/* 1 if error has occurred */
 extern int	donefld;	/* 1 if record broken into fields */
 extern int	donerec;	/* 1 if record is valid (no fld has changed */
+extern char	inputFS[];	/* FS at time of input, for field splitting */
 
 extern int	dbg;
 
diff --git a/usr.bin/awk/awklex.l b/usr.bin/awk/awklex.l
index d7ea2357e08..a179a7d074a 100644
--- a/usr.bin/awk/awklex.l
+++ b/usr.bin/awk/awklex.l
@@ -1,4 +1,4 @@
-%Start A strng sc reg comment
+%Start A str sc reg comment
 
 %{
 /****************************************************************
@@ -211,7 +211,7 @@ WS	[ \t]
 			}
 		  }
 		}
-<A>\"		{ BEGIN strng; caddreset(gs); }
+<A>\"		{ BEGIN str; caddreset(gs); }
 
 <A>"}"		{ if (--bracecnt < 0) ERROR "extra }" SYNTAX; BEGIN sc; RET(';'); }
 <A>"]"		{ if (--brackcnt < 0) ERROR "extra ]" SYNTAX; RET(']'); }
@@ -231,28 +231,28 @@ WS	[ \t]
 		  RET(REGEXPR); }
 <reg>.		{ CADD; }
 
-<strng>\"	{ BEGIN A;
+<str>\"		{ BEGIN A;
 		  cadd(gs, 0); s = tostring(gs->cbuf);
 		  cunadd(gs);
 		  cadd(gs, ' '); cadd(gs, 0);
 		  yylval.cp = setsymtab(gs->cbuf, s, 0.0, CON|STR, symtab);
 		  RET(STRING); }
-<strng>\n	{ ERROR "newline in string %.10s...", gs->cbuf SYNTAX; lineno++; BEGIN A; }
-<strng>"\\\""	{ cadd(gs, '"'); }
-<strng>"\\"n	{ cadd(gs, '\n'); }
-<strng>"\\"t	{ cadd(gs, '\t'); }
-<strng>"\\"f	{ cadd(gs, '\f'); }
-<strng>"\\"r	{ cadd(gs, '\r'); }
-<strng>"\\"b	{ cadd(gs, '\b'); }
-<strng>"\\"v	{ cadd(gs, '\v'); }	/* these ANSIisms may not be known by */
-<strng>"\\"a	{ cadd(gs, '\007'); }	/* your compiler. hence 007 for bell */
-<strng>"\\\\"	{ cadd(gs, '\\'); }
-<strng>"\\"({O}{O}{O}|{O}{O}|{O}) { int n;
+<str>\n		{ ERROR "newline in string %.10s...", gs->cbuf SYNTAX; lineno++; BEGIN A; }
+<str>"\\\""	{ cadd(gs, '"'); }
+<str>"\\"n	{ cadd(gs, '\n'); }
+<str>"\\"t	{ cadd(gs, '\t'); }
+<str>"\\"f	{ cadd(gs, '\f'); }
+<str>"\\"r	{ cadd(gs, '\r'); }
+<str>"\\"b	{ cadd(gs, '\b'); }
+<str>"\\"v	{ cadd(gs, '\v'); }	/* these ANSIisms may not be known by */
+<str>"\\"a	{ cadd(gs, '\007'); }	/* your compiler. hence 007 for bell */
+<str>"\\\\"	{ cadd(gs, '\\'); }
+<str>"\\"({O}{O}{O}|{O}{O}|{O}) { int n;
 		  sscanf(yytext+1, "%o", &n); cadd(gs, n); }
-<strng>"\\"x({H}+) { int n;	/* ANSI permits any number! */
+<str>"\\"x({H}+) { int n;	/* ANSI permits any number! */
 		  sscanf(yytext+2, "%x", &n); cadd(gs, n); }
-<strng>"\\".	{ cadd(gs, yytext[1]); }
-<strng>.	{ CADD; }
+<str>"\\".	{ cadd(gs, yytext[1]); }
+<str>.		{ CADD; }
 
 %%
 
diff --git a/usr.bin/awk/b.c b/usr.bin/awk/b.c
index 3baab2ac071..9952d94ed35 100644
--- a/usr.bin/awk/b.c
+++ b/usr.bin/awk/b.c
@@ -227,11 +227,12 @@ void freetr(Node *p)	/* free parse tree */
 /* to be seen literally;  \056 is not a metacharacter. */
 
 int hexstr(char **pp)	/* find and eval hex string at pp, return new p */
-{
+{			/* only pick up one 8-bit byte (2 chars) */
 	char *p;
 	int n = 0;
+	int i;
 
-	for (p = *pp; isxdigit(*p); p++) {
+	for (i = 0, p = *pp; i < 2 && isxdigit(*p); i++, p++) {
 		if (isdigit(*p))
 			n = 16 * n + *p - '0';
 		else if (*p >= 'a' && *p <= 'f')
@@ -243,7 +244,7 @@ int hexstr(char **pp)	/* find and eval hex string at pp, return new p */
 	return n;
 }
 
-#define isoctdigit(c) ((c) >= '0' && (c) <= '8')	/* multiple use of arg */
+#define isoctdigit(c) ((c) >= '0' && (c) <= '7')	/* multiple use of arg */
 
 int quoted(char **pp)	/* pick up next thing after a \\ */
 			/* and increment *pp */
diff --git a/usr.bin/awk/lib.c b/usr.bin/awk/lib.c
index 6f72be9133a..613e8f0c712 100644
--- a/usr.bin/awk/lib.c
+++ b/usr.bin/awk/lib.c
@@ -38,6 +38,7 @@ char	*recdata;
 char	*record;
 char	*fields;
 Cell	*fldtab;
+char	inputFS[100];	/* BUG: unchecked */
 
 #define	MAXFLD	200
 int	nfields	= MAXFLD;	/* can be set from commandline in main */
@@ -158,6 +159,7 @@ int readrec(char *buf, int bufsize, FILE *inf)	/* read one record into buf */
 	char *rr;
 	int nrr;
 
+	strcpy(inputFS, *FS);	/* for subsequent field splitting */
 	if ((sep = **RS) == 0) {
 		sep = '\n';
 		while ((c=getc(inf)) == '\n' && c != EOF)	/* skip leading \n's */
@@ -228,9 +230,9 @@ void fldbld(void)	/* create fields from current record */
 	r = recloc->sval;
 	fr = fields;
 	i = 0;	/* number of fields accumulated here */
-	if (strlen(*FS) > 1) {	/* it's a regular expression */
-		i = refldbld(r, *FS);
-	} else if ((sep = **FS) == ' ') {	/* default whitespace */
+	if (strlen(inputFS) > 1) {	/* it's a regular expression */
+		i = refldbld(r, inputFS);
+	} else if ((sep = *inputFS) == ' ') {	/* default whitespace */
 		for (i = 0; ; ) {
 			while (*r == ' ' || *r == '\t' || *r == '\n')
 				r++;
@@ -249,7 +251,7 @@ void fldbld(void)	/* create fields from current record */
 			*fr++ = 0;
 		}
 		*fr = 0;
-	} else if ((sep = **FS) == 0) {		/* new: FS="" => 1 char/field */
+	} else if ((sep = *inputFS) == 0) {		/* new: FS="" => 1 char/field */
 		for (i = 0; *r != 0; r++) {
 			char buf[2];
 			i++;
@@ -381,10 +383,10 @@ void recbld(void)	/* create $0 from $1..$NF if necessary */
 	if (r > rec + recsize - 1)
 		ERROR "built giant record `%.30s...'; try -mr n", record FATAL;
 	*r = '\0';
-	dprintf( ("in recbld FS=%o, recloc=%p\n", **FS, recloc) );
+	dprintf( ("in recbld inputFS=%s, recloc=%p\n", inputFS, recloc) );
 	recloc->tval = REC | STR | DONTFREE;
 	recloc->sval = record = rec;
-	dprintf( ("in recbld FS=%o, recloc=%p\n", **FS, recloc) );
+	dprintf( ("in recbld inputFS=%s, recloc=%p\n", inputFS, recloc) );
 	dprintf( ("recbld = |%s|\n", record) );
 	donerec = 1;
 }
diff --git a/usr.bin/awk/main.c b/usr.bin/awk/main.c
index cfd456a1b52..c9bfda2bd39 100644
--- a/usr.bin/awk/main.c
+++ b/usr.bin/awk/main.c
@@ -22,15 +22,15 @@ OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE
 USE OR PERFORMANCE OF THIS SOFTWARE.
 ****************************************************************/
 
-char	*version = "version May 27, 1996";
+char	*version = "version June 29, 1996";
 
 #define DEBUG
 #include <stdio.h>
 #include <ctype.h>
+#include <locale.h>
 #include <stdlib.h>
 #include <string.h>
 #include <signal.h>
-#include <locale.h>
 #include "awk.h"
 #include "awkgram.h"
 
diff --git a/usr.bin/awk/run.c b/usr.bin/awk/run.c
index 480173e8539..fa540f634e2 100644
--- a/usr.bin/awk/run.c
+++ b/usr.bin/awk/run.c
@@ -108,7 +108,7 @@ Cell *execute(Node *u)	/* execute a node of the parse tree */
 	for (a = u; ; a = a->nnext) {
 		curnode = a;
 		if (isvalue(a)) {
-			x = (Cell *)(a->narg[0]);
+			x = (Cell *) (a->narg[0]);
 			if ((x->tval & FLD) && !donefld)
 				fldbld();
 			else if ((x->tval & REC) && !donerec)
@@ -353,8 +353,6 @@ Cell *getline(Node **a, int n)	/* get next line from specific input */
 		x = execute(a[2]);		/* filename */
 		if ((int) a[1] == '|')	/* input pipe */
 			a[1] = (Node *) LE;	/* arbitrary flag */
-		if ((x->tval & STR) == 0)
-			x = copycell(x);
 		fp = openfile((int) a[1], getsval(x));
 		tempfree(x);
 		if (fp == NULL)
@@ -407,8 +405,6 @@ Cell *array(Node **a, int n)	/* a[0] is symtab, a[1] is list of subscripts */
 	buf[0] = 0;
 	for (np = a[1]; np; np = np->nnext) {
 		y = execute(np);	/* subscript */
-		if ((y->tval & STR) == 0)
-			y = copycell(y);
 		s = getsval(y);
 		strcat(buf, s);		/* BUG: unchecked! */
 		if (np->nnext)
@@ -448,8 +444,6 @@ Cell *adelete(Node **a, int n)	/* a[0] is symtab, a[1] is list of subscripts */
 		buf[0] = 0;
 		for (np = a[1]; np; np = np->nnext) {
 			y = execute(np);	/* subscript */
-			if ((y->tval & STR) == 0)
-				y = copycell(y);
 			s = getsval(y);
 			strcat(buf, s);
 			if (np->nnext)
@@ -481,8 +475,6 @@ Cell *intest(Node **a, int n)	/* a[0] is index (list), a[1] is symtab */
 	buf[0] = 0;
 	for (p = a[0]; p; p = p->nnext) {
 		x = execute(p);	/* expr */
-		if ((x->tval & STR) == 0)
-			x = copycell(x);
 		s = getsval(x);
 		strcat(buf, s);
 		tempfree(x);
@@ -511,15 +503,11 @@ Cell *matchop(Node **a, int n)	/* ~ and match() */
 		mode = 1;
 	}
 	x = execute(a[1]);	/* a[1] = target text */
-	if ((x->tval & STR) == 0)
-		x = copycell(x);
 	s = getsval(x);
 	if (a[0] == 0)		/* a[1] == 0: already-compiled reg expr */
 		i = (*mf)((fa *) a[2], s);
 	else {
 		y = execute(a[2]);	/* a[2] = regular expr */
-		if ((y->tval & STR) == 0)
-			y = copycell(y);
 		t = getsval(y);
 		pfa = makedfa(t, mode);
 		i = (*mf)(pfa, s);
@@ -587,10 +575,6 @@ Cell *relop(Node **a, int n)	/* a[0 < a[1], etc. */
 		j = x->fval - y->fval;
 		i = j<0? -1: (j>0? 1: 0);
 	} else {
-		if ((x->tval & STR) == 0)
-			x = copycell(x);
-		if ((y->tval & STR) == 0)
-			y = copycell(y);
 		i = strcmp(getsval(x), getsval(y));
 	}
 	tempfree(x);
@@ -649,8 +633,6 @@ Cell *indirect(Node **a, int n)	/* $( a[0] ) */
 	char *s;
 
 	x = execute(a[0]);
-	if ((x->tval & STR) == 0)
-		x = copycell(x);
 	m = getfval(x);
 	if (m == 0 && !isnumber(s = getsval(x)))	/* suspicion! */
 		ERROR "illegal field $(%s), name \"%s\"", s, x->nval FATAL;
@@ -674,8 +656,6 @@ Cell *substr(Node **a, int nnn)		/* substr(a[0], a[1], a[2]) */
 	y = execute(a[1]);
 	if (a[2] != 0)
 		z = execute(a[2]);
-	if ((x->tval & STR) == 0)
-		x = copycell(x);
 	s = getsval(x);
 	k = strlen(s) + 1;
 	if (k <= 1) {
@@ -719,12 +699,8 @@ Cell *sindex(Node **a, int nnn)		/* index(a[0], a[1]) */
 	Awkfloat v = 0.0;
 
 	x = execute(a[0]);
-	if ((x->tval & STR) == 0)
-		x = copycell(x);
 	s1 = getsval(x);
 	y = execute(a[1]);
-	if ((y->tval & STR) == 0)
-		y = copycell(y);
 	s2 = getsval(y);
 
 	z = gettemp();
@@ -805,8 +781,6 @@ int format(char *buf, int bufsize, char *s, Node *a)	/* printf-like conversions
 		if (a == NULL)
 			ERROR "not enough args in printf(%s)", os FATAL;
 		x = execute(a);
-		if ((x->tval & STR) == 0)
-			x = copycell(x);
 		a = a->nnext;
 		switch (flag) {
 		case 0:	sprintf((char *)p, "%s", fmt);	/* unknown, so dump it too */
@@ -847,8 +821,6 @@ Cell *awksprintf(Node **a, int n)		/* sprintf(a[0]) */
 
 	y = a[0]->nnext;
 	x = execute(a[0]);
-	if ((x->tval & STR) == 0)
-		x = copycell(x);
 	if (format(buf, sizeof buf, getsval(x), y) == -1)
 		ERROR "sprintf string %.30s... too long", buf FATAL;
 	tempfree(x);
@@ -868,8 +840,6 @@ Cell *awkprintf(Node **a, int n)		/* printf */
 
 	y = a[0]->nnext;
 	x = execute(a[0]);
-	if ((x->tval & STR) == 0)
-		x = copycell(x);
 	if (format(buf, sizeof buf, getsval(x), y) == -1)
 		ERROR "printf string %.30s... too long", buf FATAL;
 	tempfree(x);
@@ -894,14 +864,10 @@ Cell *arith(Node **a, int n)	/* a[0] + a[1], etc.  also -a[0] */
 	Cell *x, *y, *z;
 
 	x = execute(a[0]);
-	if ((x->tval & NUM) == 0)
-		x = copycell(x);
 	i = getfval(x);
 	tempfree(x);
 	if (n != UMINUS) {
 		y = execute(a[1]);
-		if ((y->tval & NUM) == 0)
-			y = copycell(y);
 		j = getfval(y);
 		tempfree(y);
 	}
@@ -1047,10 +1013,6 @@ Cell *cat(Node **a, int q)	/* a[0] cat a[1] */
 
 	x = execute(a[0]);
 	y = execute(a[1]);
-	if ((x->tval & STR) == 0)
-		x = copycell(x);
-	if ((y->tval & STR) == 0)
-		y = copycell(y);
 	getsval(x);
 	getsval(y);
 	n1 = strlen(x->sval);
@@ -1117,15 +1079,11 @@ Cell *split(Node **a, int nnn)	/* split(a[0], a[1], a[2]); a[3] is type */
 	int n, tempstat;
 
 	y = execute(a[0]);	/* source string */
-	if ((y->tval & STR) == 0)
-		y = copycell(y);
 	s = getsval(y);
 	if (a[2] == 0)		/* fs string */
 		fs = *FS;
 	else if ((int) a[3] == STRING) {	/* split(str,arr,"string") */
 		x = execute(a[2]);
-		if ((x->tval & STR) == 0)
-			x = copycell(x);
 		fs = getsval(x);
 	} else if ((int) a[3] == REGEXPR)
 		fs = (char*) "(regexpr)";	/* split(str,arr,/regexpr/) */
@@ -1433,8 +1391,6 @@ Cell *bltin(Node **a, int n)	/* builtin functions. a[0] is type, a[1] is arg lis
 	nextarg = a[1]->nnext;
 	switch (t) {
 	case FLENGTH:
-		if ((x->tval & STR) == 0)
-			x = copycell(x);
 		u = strlen(getsval(x)); break;
 	case FLOG:
 		u = errcheck(log(getfval(x)), "log"); break;
@@ -1461,8 +1417,6 @@ Cell *bltin(Node **a, int n)	/* builtin functions. a[0] is type, a[1] is arg lis
 		break;
 	case FSYSTEM:
 		fflush(stdout);		/* in case something is buffered already */
-		if ((x->tval & STR) == 0)
-			x = copycell(x);
 		u = (Awkfloat) system((char *)getsval(x)) / 256;   /* 256 is unix-dep */
 		break;
 	case FRAND:
@@ -1478,8 +1432,6 @@ Cell *bltin(Node **a, int n)	/* builtin functions. a[0] is type, a[1] is arg lis
 		break;
 	case FTOUPPER:
 	case FTOLOWER:
-		if ((x->tval & STR) == 0)
-			x = copycell(x);
 		strcpy(buf, getsval(x));
 		if (t == FTOUPPER) {
 			for (p = buf; *p; p++)
@@ -1495,8 +1447,6 @@ Cell *bltin(Node **a, int n)	/* builtin functions. a[0] is type, a[1] is arg lis
 		setsval(x, buf);
 		return x;
 	case FFLUSH:
-		if ((x->tval & STR) == 0)
-			x = copycell(x);
 		if ((fp = openfile(GT, getsval(x))) == NULL)
 			u = EOF;
 		else
@@ -1519,8 +1469,6 @@ Cell *bltin(Node **a, int n)	/* builtin functions. a[0] is type, a[1] is arg lis
 
 Cell *printstat(Node **a, int n)	/* print a[0] */
 {
-	extern char **OFMT, **CONVFMT;
-	char **save;
 	Node *x;
 	Cell *y;
 	FILE *fp;
@@ -1531,12 +1479,7 @@ Cell *printstat(Node **a, int n)	/* print a[0] */
 		fp = redirect((int)a[1], a[2]);
 	for (x = a[0]; x != NULL; x = x->nnext) {
 		y = execute(x);
-		if ((y->tval & STR) == 0)
-			y = copycell(y);
-		save = CONVFMT;
-		CONVFMT = OFMT;
 		fputs((char *)getsval(y), fp);
-		CONVFMT = save;
 		tempfree(y);
 		if (x->nnext == NULL)
 			fputs((char *)*ORS, fp);
@@ -1565,8 +1508,6 @@ FILE *redirect(int a, Node *b)	/* set up all i/o redirections */
 	char *fname;
 
 	x = execute(b);
-	if ((x->tval & STR) == 0)
-		x = copycell(x);
 	fname = getsval(x);
 	fp = openfile(a, fname);
 	if (fp == NULL)
@@ -1642,8 +1583,6 @@ Cell *closefile(Node **a, int n)
 
 	n = 0;
 	x = execute(a[0]);
-	if ((x->tval & STR) == 0)
-		x = copycell(x);
 	getsval(x);
 	for (i = 0; i < FOPEN_MAX; i++)
 		if (files[i].fname && strcmp(x->sval, files[i].fname) == 0) {
@@ -1691,8 +1630,6 @@ Cell *sub(Node **a, int nnn)	/* substitute command */
 	fa *pfa;
 
 	x = execute(a[3]);	/* target string */
-	if ((x->tval & STR) == 0)
-		x = copycell(x);
 	t = getsval(x);
 	if (a[0] == 0)		/* 0 => a[1] is already-compiled regexpr */
 		pfa = (fa *) a[1];	/* regular expression */
@@ -1702,8 +1639,6 @@ Cell *sub(Node **a, int nnn)	/* substitute command */
 		tempfree(y);
 	}
 	y = execute(a[2]);	/* replacement string */
-	if ((y->tval & STR) == 0)
-		y = copycell(y);
 	result = false;
 	if (pmatch(pfa, t)) {
 		pb = buf;
@@ -1749,21 +1684,15 @@ Cell *gsub(Node **a, int nnn)	/* global substitute */
 	mflag = 0;	/* if mflag == 0, can replace empty string */
 	num = 0;
 	x = execute(a[3]);	/* target string */
-	if ((x->tval & STR) == 0)
-		x = copycell(x);
 	t = getsval(x);
 	if (a[0] == 0)		/* 0 => a[1] is already-compiled regexpr */
 		pfa = (fa *) a[1];	/* regular expression */
 	else {
 		y = execute(a[1]);
-		if ((y->tval & STR) == 0)
-			y = copycell(y);
 		pfa = makedfa(getsval(y), 1);
 		tempfree(y);
 	}
 	y = execute(a[2]);	/* replacement string */
-	if ((y->tval & STR) == 0)
-		y = copycell(y);
 	if (pmatch(pfa, t)) {
 		tempstat = pfa->initstat;
 		pfa->initstat = 2;
diff --git a/usr.bin/awk/tran.c b/usr.bin/awk/tran.c
index 973e355765f..86ca9ead221 100644
--- a/usr.bin/awk/tran.c
+++ b/usr.bin/awk/tran.c
@@ -82,7 +82,7 @@ void syminit(void)	/* initialize symbol table with builtin vars */
 	ORS = &setsymtab("ORS", "\n", 0.0, STR|DONTFREE, symtab)->sval;
 	OFMT = &setsymtab("OFMT", "%.6g", 0.0, STR|DONTFREE, symtab)->sval;
 	CONVFMT = &setsymtab("CONVFMT", "%.6g", 0.0, STR|DONTFREE, symtab)->sval;
-	FILENAME = &setsymtab("FILENAME", "-", 0.0, STR|DONTFREE, symtab)->sval;
+	FILENAME = &setsymtab("FILENAME", "", 0.0, STR|DONTFREE, symtab)->sval;
 	nfloc = setsymtab("NF", "", 0.0, NUM, symtab);
 	NF = &nfloc->fval;
 	nrloc = setsymtab("NR", "", 0.0, NUM, symtab);