===================================================================
RCS file: /cvsrepo/anoncvs/cvs/src/usr.bin/pr/pr.c,v
retrieving revision 1.19
retrieving revision 1.20
diff -c -r1.19 -r1.20
*** src/usr.bin/pr/pr.c	2003/08/04 17:06:45	1.19
--- src/usr.bin/pr/pr.c	2003/11/09 20:13:57	1.20
***************
*** 1,4 ****
! /*	$OpenBSD: pr.c,v 1.19 2003/08/04 17:06:45 deraadt Exp $	*/
  
  /*-
   * Copyright (c) 1991 Keith Muller.
--- 1,4 ----
! /*	$OpenBSD: pr.c,v 1.20 2003/11/09 20:13:57 otto Exp $	*/
  
  /*-
   * Copyright (c) 1991 Keith Muller.
***************
*** 41,47 ****
  
  #ifndef lint
  /* from: static char sccsid[] = "@(#)pr.c	8.1 (Berkeley) 6/6/93"; */
! static char *rcsid = "$OpenBSD: pr.c,v 1.19 2003/08/04 17:06:45 deraadt Exp $";
  #endif /* not lint */
  
  #include <sys/types.h>
--- 41,47 ----
  
  #ifndef lint
  /* from: static char sccsid[] = "@(#)pr.c	8.1 (Berkeley) 6/6/93"; */
! static char *rcsid = "$OpenBSD: pr.c,v 1.20 2003/11/09 20:13:57 otto Exp $";
  #endif /* not lint */
  
  #include <sys/types.h>
***************
*** 76,95 ****
   * the original version didn't support form-feeds, while many of the ad-hoc
   * pr implementations out there do.  Addding this and making it work reasonably
   * in all four output modes required quite a bit of hacking and a few minor
!  * bugs were noted and fixed in the processs.  Some implementations have this
   * as the as -f, some as -F so we accept either.
   *
   * The implementation of form feeds on top of the existing I/O structure is
!  * a bit ideosyncratic.  Basically they are treated as temporary end-of-file
   * conditions and an additional level of "loop on form feed" is added to each
   * of the output modes to continue after such a transient end-of-file's. This
   * has the general benefit of making the existing header/trailer logic work
   * and provides a usable framework for rational behavior in multi-column modes.
   *
!  * The orginal "efficient" implementation of the "skip to page N" option was
   * bogus and I substituted the basic inhibit printing until page N approach.
   * This is still fairly bogus vis-a-vis numbering pages on multiple files
!  * restarting at one, but at least lets you consistantly reprint some large
   * document starting in the middle, in any of the output modes.
   *
   * Additional support for overprinting via <back-space> or <return> would
--- 76,95 ----
   * the original version didn't support form-feeds, while many of the ad-hoc
   * pr implementations out there do.  Addding this and making it work reasonably
   * in all four output modes required quite a bit of hacking and a few minor
!  * bugs were noted and fixed in the process.  Some implementations have this
   * as the as -f, some as -F so we accept either.
   *
   * The implementation of form feeds on top of the existing I/O structure is
!  * a bit idiosyncratic.  Basically they are treated as temporary end-of-file
   * conditions and an additional level of "loop on form feed" is added to each
   * of the output modes to continue after such a transient end-of-file's. This
   * has the general benefit of making the existing header/trailer logic work
   * and provides a usable framework for rational behavior in multi-column modes.
   *
!  * The original "efficient" implementation of the "skip to page N" option was
   * bogus and I substituted the basic inhibit printing until page N approach.
   * This is still fairly bogus vis-a-vis numbering pages on multiple files
!  * restarting at one, but at least lets you consistently reprint some large
   * document starting in the middle, in any of the output modes.
   *
   * Additional support for overprinting via <back-space> or <return> would
***************
*** 921,927 ****
       * do not know how many columns yet. The number of operands provide an
       * upper bound on the number of columns. We use the number of files
       * we can open successfully to set the number of columns. The operation
!      * of the merge operation (-m) in relation to unsuccesful file opens
       * is unspecified by posix.
       *
       * XXX - this seems moderately bogus, you'd think that specifying
--- 921,927 ----
       * do not know how many columns yet. The number of operands provide an
       * upper bound on the number of columns. We use the number of files
       * we can open successfully to set the number of columns. The operation
!      * of the merge operation (-m) in relation to unsuccessful file opens
       * is unspecified by posix.
       *
       * XXX - this seems moderately bogus, you'd think that specifying
***************
*** 946,952 ****
  	return(1);
  
      /*
!      * calculate page boundries based on open file count
       */
      if (nmwd) {
  	colwd = (pgwd - clcnt - nmwd)/clcnt;
--- 946,952 ----
  	return(1);
  
      /*
!      * calculate page boundaries based on open file count
       */
      if (nmwd) {
  	colwd = (pgwd - clcnt - nmwd)/clcnt;
***************
*** 1099,1105 ****
   *    buf:    buffer
   *    lim:    buffer length
   *    cnt:    line length or -1 if no line (EOF for example)
!  *    cps:    column positon 1st char in buffer (large line support)
   *    trnc:    throw away data more than lim up to \n 
   *    mor:    set if more data in line (not truncated)
   */
--- 1099,1105 ----
   *    buf:    buffer
   *    lim:    buffer length
   *    cnt:    line length or -1 if no line (EOF for example)
!  *    cps:    column position 1st char in buffer (large line support)
   *    trnc:    throw away data more than lim up to \n 
   *    mor:    set if more data in line (not truncated)
   */
***************
*** 1564,1570 ****
   * prhead() should be used carefully, we don't want to print out headers
   * for null input files or orphan headers at the end of files, and also
   * trailer processing is typically conditional on whether you've called
!  * prhead() at least once for a file and incremented pagecnt..  Exactly
   * how to determine whether to print a header is a little different in
   * the context each output mode, but we let the caller figure that out.
   */
--- 1564,1570 ----
   * prhead() should be used carefully, we don't want to print out headers
   * for null input files or orphan headers at the end of files, and also
   * trailer processing is typically conditional on whether you've called
!  * prhead() at least once for a file and incremented pagecnt.  Exactly
   * how to determine whether to print a header is a little different in
   * the context each output mode, but we let the caller figure that out.
   */
***************
*** 1612,1618 ****
   *    incomp    was a '\n' missing from last line output
   *
   * prtail() can now be invoked unconditionally, with the notion that if
!  * we haven't printed a hearder, these no need for a trailer
   */
  int
  prtail(int cnt, int incomp)
--- 1612,1618 ----
   *    incomp    was a '\n' missing from last line output
   *
   * prtail() can now be invoked unconditionally, with the notion that if
!  * we haven't printed a header, there is no need for a trailer
   */
  int
  prtail(int cnt, int incomp)