Fix 32 bit dependencies.

Fix UTF-8 overstriking.
Fix EBCDIC host support.

NEWS.VER

@@ -22,6 +22,12 @@
 
 * Default LESSCHARSET for MS-DOS versions is "dos".
 
+* Fix overstriking in UTF-8 mode.
+
+* Some fixes to eliminate 32-bit file offset dependencies.
+
+* Some fixes for EBCDIC systems.
+
 ======================================================================
 
  Major changes between "less" versions 354 and 358

ch.c

@@ -21,6 +21,8 @@
 #include <windows.h>
 #endif
 
+typedef POSITION BLOCKNUM;
+
 public int ignore_eoi;
 
 /*
@@ -32,7 +34,7 @@ public int ignore_eoi;
 #define LBUFSIZE 1024
 struct buf {
  struct buf *next, *prev; /* Must be first to match struct filestate */
- long block;
+ BLOCKNUM block;
  unsigned int datasize;
  unsigned char data[LBUFSIZE];
 };
@@ -44,13 +46,13 @@ struct buf {
 struct filestate {
  /* -- Following members must match struct buf */
  struct buf *buf_next, *buf_prev;
- long buf_block;
+ BLOCKNUM buf_block;
  /* -- End of struct buf copy */
  int file;
  int flags;
  POSITION fpos;
  int nbufs;
- long block;
+ BLOCKNUM block;
  unsigned int offset;
  POSITION fsize;
 };
@@ -125,7 +127,7 @@ fch_get()
  * If the LRU buffer has data in it, 
  * then maybe allocate a new buffer.
  */
- if (ch_buftail == END_OF_CHAIN || ch_buftail->block != (long)(-1))
+ if (ch_buftail == END_OF_CHAIN || ch_buftail->block != -1)
  {
  /*
  * There is no empty buffer to use.
@@ -318,8 +320,8 @@ sync_logfile()
 {
  register struct buf *bp;
  int warned = FALSE;
- long block;
- long nblocks;
+ BLOCKNUM block;
+ BLOCKNUM nblocks;
 
  nblocks = (ch_fpos + LBUFSIZE - 1) / LBUFSIZE;
  for (block = 0; block < nblocks; block++)
@@ -352,7 +354,7 @@ sync_logfile()
  */
  static int
 buffered(block)
- long block;
+ BLOCKNUM block;
 {
  register struct buf *bp;
 
@@ -370,7 +372,7 @@ buffered(block)
 ch_seek(pos)
  register POSITION pos;
 {
- long new_block;
+ BLOCKNUM new_block;
  POSITION len;
 
  len = ch_length();
@@ -470,12 +472,10 @@ ch_length()
 /*
  * Return the current position in the file.
  */
-#define tellpos(blk,off) ((POSITION)((((long)(blk)) * LBUFSIZE) + (off)))
-
  public POSITION
 ch_tell()
 {
- return (tellpos(ch_block, ch_offset));
+ return (ch_block * LBUFSIZE) + ch_offset;
 }
 
 /*
@@ -570,7 +570,7 @@ ch_flush()
  * Initialize all the buffers.
  */
  for (bp = ch_bufhead; bp != END_OF_CHAIN; bp = bp->next)
- bp->block = (long)(-1);
+ bp->block = -1;
 
  /*
  * Figure out the size of the file, if we can.
@@ -626,7 +626,7 @@ ch_addbuf()
  if (bp == NULL)
  return (1);
  ch_nbufs++;
- bp->block = (long)(-1);
+ bp->block = -1;
  bp->next = END_OF_CHAIN;
  bp->prev = ch_buftail;
  ch_buftail->next = bp;
@@ -693,7 +693,7 @@ ch_init(f, flags)
  thisfile = (struct filestate *) 
  calloc(1, sizeof(struct filestate));
  thisfile->buf_next = thisfile->buf_prev = END_OF_CHAIN;
- thisfile->buf_block = (long)(-1);
+ thisfile->buf_block = -1;
  thisfile->nbufs = 0;
  thisfile->flags = 0;
  thisfile->fpos = 0;

charset.c

@@ -34,6 +34,7 @@ struct charset {
  { "ascii", NULL, "8bcccbcc18b95.b" },
  { "dos", NULL, "8bcccbcc12bc5b223.b" },
  { "ebcdic", NULL, "5bc6bcc7bcc41b.9b7.9b5.b..8b6.10b6.b9.7b9.8b8.17b3.3b9.7b9.8b8.6b10.b.b.b." },
+ { "IBM-1047", NULL, "4cbcbc3b9cbccbccbb4c6bcc5b3cbbc4bc4bccbc191.b" },
  { "iso8859", NULL, "8bcccbcc18b95.33b." },
  { "koi8-r", NULL, "8bcccbcc18b95.b128." },
  { "latin1", NULL, "8bcccbcc18b95.33b." },
@@ -293,8 +294,22 @@ prchar(c)
  sprintf(buf, "%c", c);
  else if (c == ESC)
  sprintf(buf, "ESC");
- else if (c < 128 && !control_char(c ^ 0100))
- sprintf(buf, "^%c", c ^ 0100);
+#if IS_EBCDIC_HOST
+ else if (!binary_char(c) && c < 64)
+ sprintf(buf, "^%c",
+ /*
+ * This array roughly inverts CONTROL() #defined in less.h,
+ * and should be kept in sync with CONTROL() and IBM-1047.
+ */
+ "@ABC.I.?...KLMNO"
+ "PQRS.JH.XY.."
+ "\\]^_"
+ "......W[.....EFG"
+ "..V....D....TU.Z"[c]);
+#else
+ else if (c < 128 && !control_char(c ^ 0100))
+ sprintf(buf, "^%c", c ^ 0100);
+#endif
  else
  sprintf(buf, binfmt, c);
  return (buf);

filename.c

@@ -24,7 +24,6 @@
 #if MSDOS_COMPILER==DJGPPC
 #include <glob.h>
 #include <dir.h>
-#include <limits.h>
 #define _MAX_PATH PATH_MAX
 #endif
 #endif

less.h

@@ -70,6 +70,9 @@
 #if HAVE_CTYPE_H
 #include <ctype.h>
 #endif
+#if HAVE_LIMITS_H
+#include <limits.h>
+#endif
 #if HAVE_STDLIB_H
 #include <stdlib.h>
 #endif
@@ -136,13 +139,21 @@ void free();
 
 #define BAD_LSEEK ((off_t)-1)
 
+#ifndef CHAR_BIT
+#define CHAR_BIT 8
+#endif
+
+/*
+ * Upper bound on the string length of an integer converted to string.
+ * 302 / 1000 is ceil (log10 (2.0)). Subtract 1 for the sign bit;
+ * add 1 for integer division truncation; add 1 more for a minus sign.
+ */
+#define INT_STRLEN_BOUND(t) ((sizeof(t) * CHAR_BIT - 1) * 302 / 1000 + 1 + 1)
+
 /*
  * Special types and constants.
  */
 typedef off_t POSITION;
-#define PR_POSITION "%ld"
-#define MAX_PRINT_POSITION 20
-#define MAX_PRINT_INT 10
 
 #define NULL_POSITION ((POSITION)(-1))
 
@@ -283,6 +294,10 @@ struct textlist
 #define AT_INVIS (4)
 #define AT_STANDOUT (5)
 
+#if '0' == 240
+#define IS_EBCDIC_HOST 1
+#endif
+
 #if IS_EBCDIC_HOST
 /*
  * Long definition for EBCDIC.

less.nro.VER

@@ -1082,6 +1082,11 @@ Same as iso8859.
 Selects a character set appropriate for MS-DOS.
 .IP ebcdic
 Selects an EBCDIC character set.
+.IP IBM-1047
+Selects an EBCDIC character set used by OS/390 Unix Services.
+This is the EBCDIC analogue of latin1. You get similar results
+by setting either LESSCHARSET=IBM-1047 or LC_CTYPE=en_US
+in your environment.
 .IP koi8-r
 Selects a Russian character set.
 .IP next
@@ -1116,6 +1121,9 @@ to each of the possible values for LESSCHARSET:
  ebcdic 5bc6bcc7bcc41b.9b7.9b5.b..8b6.10b6.b9.7b
 .br
  \ \ \ \ \ \ 9.8b8.17b3.3b9.7b9.8b8.6b10.b.b.b.
+.br
+ IBM-1047 4cbcbc3b9cbccbccbb4c6bcc5b3cbbc4bc4bccbc
+ \ \ \ \ \ \ 191.b
 .br
  iso8859 8bcccbcc18b95.33b.
 .br

line.c

@@ -471,12 +471,37 @@ do_append(c, pos)
  * bold (if an identical character is overstruck),
  * or just deletion of the character in the buffer.
  */
- overstrike = 0;
- if ((char)c == linebuf[curr])
+ overstrike--;
+ if (utf_mode && curr > 1 && (char)c == linebuf[curr-2])
+ {
+ backc();
+ backc();
+ overstrike = 2;
+ } else if (utf_mode && curr > 0 && (char)c == linebuf[curr-1])
+ {
+ backc();
  STOREC(linebuf[curr], AT_BOLD);
- else if (c == '_')
+ overstrike = 1;
+ } else if ((char)c == linebuf[curr])
+ {
+ STOREC(linebuf[curr], AT_BOLD);
+ } else if (c == '_')
+ {
+ if (utf_mode)
+ {
+ if (curr > 0 && IS_CONT(linebuf[curr]))
+ attr[curr-1] = AT_UNDERLINE;
+ if (curr > 1 && IS_CONT(linebuf[curr-1]))
+ attr[curr-2] = AT_UNDERLINE;
+ if (curr > 2 && IS_CONT(linebuf[curr-2]))
+ attr[curr-3] = AT_UNDERLINE;
+ if (curr > 3 && IS_CONT(linebuf[curr-3]))
+ attr[curr-4] = AT_UNDERLINE;
+ if (curr > 4 && IS_CONT(linebuf[curr-4]))
+ attr[curr-5] = AT_UNDERLINE;
+ }
  STOREC(linebuf[curr], AT_UNDERLINE);
- else if (linebuf[curr] == '_')
+ } else if (linebuf[curr] == '_')
  STOREC(c, AT_UNDERLINE);
  else if (control_char(c))
  goto do_control_char;

lsystem.c

@@ -14,8 +14,8 @@
  * Necessarily very OS dependent.
  */
 
-#include <signal.h>
 #include "less.h"
+#include <signal.h>
 #include "position.h"
 
 #if MSDOS_COMPILER

option.c

@@ -116,11 +116,12 @@ scan_option(s)
  * EVERY input file.
  */
  plusoption = TRUE;
- if (*s == '+')
- every_first_cmd = save(++s);
- else
- ungetsc(s);
+ str = s;
  s = optstring(s, propt('+'));
+ if (*str == '+')
+ every_first_cmd = save(++str);
+ else
+ ungetsc(str);
  continue;
  case '0': case '1': case '2': case '3': case '4':
  case '5': case '6': case '7': case '8': case '9':

os.c

@@ -33,9 +33,6 @@
 #if HAVE_VALUES_H
 #include <values.h>
 #endif
-#if HAVE_LIMITS_H
-#include <limits.h>
-#endif
 
 #if HAVE_TIME_T
 #define time_type time_t
@@ -232,36 +229,6 @@ errno_message(filename)
  return (m);
 }
 
-/*
- * Return the largest possible number that can fit in a long.
- */
- static long
-get_maxlong()
-{
-#ifdef LONG_MAX
- return (LONG_MAX);
-#else
-#ifdef MAXLONG
- return (MAXLONG);
-#else
- long n, n2;
-
- /*
- * Keep doubling n until we overflow.
- * {{ This actually only returns the largest power of two that
- * can fit in a long, but percentage() doesn't really need
- * it any more accurate than that. }}
- */
- n2 = 128; /* Hopefully no maxlong is less than 128! */
- do {
- n = n2;
- n2 *= 2;
- } while (n2 / 2 == n);
- return (n);
-#endif
-#endif
-}
-
 /*
  * Return the ratio of two POSITIONS, as a percentage.
  * {{ Assumes a POSITION is a long int. }}
@@ -270,23 +237,28 @@ get_maxlong()
 percentage(num, den)
  POSITION num, den;
 {
- if (num <= get_maxlong() / 100)
- return ((100 * num) / den);
+ POSITION num100 = num * 100;
+
+ if (num100 / 100 == num)
+ return (num100 / den);
  else
  return (num / (den / 100));
 }
 
 /*
  * Return the specified percentage of a POSITION.
- * {{ Assumes a POSITION is a long int. }}
  */
  public POSITION
 percent_pos(pos, percent)
  POSITION pos;
  int percent;
 {
- if (pos <= get_maxlong() / 100)
- return ((percent * pos) / 100);
+ POSITION result100;
+
+ if (percent == 0)
+ return (0);
+ else if ((result100 = pos * percent) / percent == pos)
+ return (result100 / 100);
  else
  return (percent * (pos / 100));
 }

output.c

@@ -250,7 +250,7 @@ iprintnum(num, radix)
  register char *s;
  int r;
  int neg;
- char buf[10];
+ char buf[INT_STRLEN_BOUND(num)];
 
  neg = (num < 0);
  if (neg)