commit bash-20170210 snapshot

CWRU/CWRU.chlog

@@ -13115,3 +13115,53 @@ lib/readline/input.c
 	  the next character as an unsigned char, so we handle multibyte
 	  character sequences correctly.  Report and fix from Grisha Levit
 	  <grishalevit@gmail.com>
+
+				    2/6
+				    ---
+shell.c
+	- --pretty-print: new invocation option, currently undocumented, dumps
+	  a pretty-printed version of a shell script given as an argument to
+	  stdout
+	- main: call pretty_print_loop() if we're in pretty-printing mode in
+	  a non-interactive shell
+
+eval.c
+	- pretty_print_loop: new function, similar to reader_loop, reads
+	  commands and prints them by displaying the result from
+	  make_command_string()
+
+externs.h
+	- pretty_print_loop: new extern declaration
+
+				    2/9
+				    ---
+doc/{bash.1,bashref.texi}
+	- here strings: correct documentation to note they do not undergo
+	  brace expansion. Report from Conor McCarthy <mr.spuratic@gmail.com>
+
+				    2/9
+				    ---
+lib/readline/display.c
+	- expand_prompt: instead of recomputing the line break offsets for
+	  prompts that span multiple screen lines on every call to rl_redisplay,
+	  do it here. Manage a local array local_prompt_newlines; set an offset
+	  every time `physchars' hits a multiple of the screen width.  Inspired
+	  by https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=843819
+	- rl_redisplay: don't calculate multi-line prompt line break indices
+	  here; use the values calculated by expand_prompt
+	- rl_{save,restore}_prompt: save and restore local_prompt_newlines
+
+				   2/11
+				   ----
+lib/readline/mbutil.c
+	- _rl_find_prev_mbchar_internal: since we're traversing the string
+	  from the beginning, shortcut non-multibyte characters in a UTF-8
+	  locale
+
+parse.y
+	- parse_comsub: if we are parsing a here document with a quoted
+	  delimiter (LEX_QUOTEDDOC), we should not remove <backslash><newline>
+	  pairs from the body of the here document.  Need to add LEX_QUOTEDDOC
+	  to the values in tflags when calling shell_getc.  Fixes bug reported
+	  by Michael Homer <michael.homer@ecs.vuw.ac.nz>, patch was originally
+	  contributed by Geir Hauge <geir.hauge@gmail.com>

MANIFEST

@@ -899,6 +899,7 @@ tests/comsub.right	f
 tests/comsub1.sub	f
 tests/comsub2.sub	f
 tests/comsub3.sub	f
+tests/comsub4.sub	f
 tests/comsub-eof.tests	f
 tests/comsub-eof0.sub	f
 tests/comsub-eof1.sub	f

doc/bash.1

@@ -5,12 +5,12 @@
 .\"	Case Western Reserve University
 .\"	chet.ramey@case.edu
 .\"
-.\"	Last Change: Fri Feb  3 16:01:50 EST 2017
+.\"	Last Change: Thu Feb  9 15:46:36 EST 2017
 .\"
 .\" bash_builtins, strip all but Built-Ins section
 .if \n(zZ=1 .ig zZ
 .if \n(zY=1 .ig zY
-.TH BASH 1 "2017 February 3" "GNU Bash 4.4"
+.TH BASH 1 "2017 February 9" "GNU Bash 4.4"
 .\"
 .\" There's some problem with having a `@'
 .\" in a tagged paragraph with the BSD man macros.
@@ -50,8 +50,8 @@ bash \- GNU Bourne-Again SHell
 [options]
 [command_string | file]
 .SH COPYRIGHT
-.if n Bash is Copyright (C) 1989-2016 by the Free Software Foundation, Inc.
-.if t Bash is Copyright \(co 1989-2016 by the Free Software Foundation, Inc.
+.if n Bash is Copyright (C) 1989-2017 by the Free Software Foundation, Inc.
+.if t Bash is Copyright \(co 1989-2017 by the Free Software Foundation, Inc.
 .SH DESCRIPTION
 .B Bash
 is an \fBsh\fR-compatible command language interpreter that
@@ -4017,7 +4017,7 @@ A variant of here documents, the format is:
 .RE
 .PP
 The \fIword\fP undergoes
-brace expansion, tilde expansion, parameter and variable expansion,
+tilde expansion, parameter and variable expansion,
 command substitution, arithmetic expansion, and quote removal.
 Pathname expansion and word splitting are not performed.
 The result is supplied as a single string, with a newline appended,

doc/bashref.texi

@@ -2773,7 +2773,7 @@ A variant of here documents, the format is:
 @end example
 
 The @var{word} undergoes
-brace expansion, tilde expansion, parameter and variable expansion,
+tilde expansion, parameter and variable expansion,
 command substitution, arithmetic expansion, and quote removal.
 Pathname expansion and word splitting are not performed.
 The result is supplied as a single string,

doc/version.texi

@@ -2,10 +2,10 @@
 Copyright (C) 1988-2017 Free Software Foundation, Inc.
 @end ignore
 
-@set LASTCHANGE Wed Feb  1 09:09:18 EST 2017
+@set LASTCHANGE Thu Feb  9 15:46:16 EST 2017
 
 @set EDITION 4.4
 @set VERSION 4.4
 
-@set UPDATED 1 February 2017
+@set UPDATED 9 February 2017
 @set UPDATED-MONTH February 2017

eval.c

@@ -193,6 +193,47 @@ reader_loop ()
   return (last_command_exit_value);
 }
 
+/* Pretty print shell scripts */
+int
+pretty_print_loop ()
+{
+  COMMAND *current_command;
+  char *command_to_print;
+  int code;
+  int global_posix_mode, last_was_newline;
+
+  global_posix_mode = posixly_correct;
+  last_was_newline = 0;
+  while (EOF_Reached == 0)
+    {
+      code = setjmp_nosigs (top_level);
+      if (code)
+        return (EXECUTION_FAILURE);
+      if (read_command() == 0)
+	{
+	  current_command = global_command;
+	  global_command = 0;
+	  posixly_correct = 1;			/* print posix-conformant */
+	  if (current_command && (command_to_print = make_command_string (current_command)))
+	    {
+	      printf ("%s\n", command_to_print);	/* for now */
+	      last_was_newline = 0;
+	    }
+	  else if (last_was_newline == 0)
+	    {
+	       printf ("\n");
+	       last_was_newline = 1;
+	    }
+	  posixly_correct = global_posix_mode;
+	  dispose_command (current_command);
+	}
+      else
+	return (EXECUTION_FAILURE);
+    }
+    
+  return (EXECUTION_SUCCESS);
+}
+
 static sighandler
 alrm_catcher(i)
      int i;

execute_cmd.c

@@ -271,6 +271,8 @@ static int function_line_number;
    report the correct line number.  Kind of a hack. */
 static int showing_function_line;
 
+static int connection_count;
+
 /* $LINENO ($BASH_LINENO) for use by an ERR trap.  Global so parse_and_execute
    can save and restore it. */
 int line_number_for_err_trap;

externs.h

@@ -94,6 +94,7 @@ extern void get_current_user_info __P((void));
 
 /* Functions from eval.c. */
 extern int reader_loop __P((void));
+extern int pretty_print_loop __P((void));
 extern int parse_command __P((void));
 extern int read_command __P((void));
 

jobs.c

@@ -848,8 +848,15 @@ bgp_delete (pid)
 
   /* Search chain using hash to find bucket in pidstat_table */
   for (psi = *(pshash_getbucket (pid)); psi != NO_PIDSTAT; psi = bgpids.storage[psi].bucket_next)
-    if (bgpids.storage[psi].pid == pid)
-      break;
+    {
+      if (bgpids.storage[psi].pid == pid)
+	break;
+      if (psi == bgpids.storage[psi].bucket_next)	/* catch reported bug */
+	{
+	  internal_warning ("bgp_delete: BUG: psi (%d) == storage[psi].bucket_next", psi);
+	  return 0;
+	}
+    }
 
   if (psi == NO_PIDSTAT)
     return 0;		/* not found */

lib/readline/display.c

@@ -251,6 +251,8 @@ static int modmark;
 
 static char *saved_local_prompt;
 static char *saved_local_prefix;
+static int *saved_local_prompt_newlines;
+
 static int saved_last_invisible;
 static int saved_visible_length;
 static int saved_prefix_length;
@@ -303,6 +305,8 @@ prompt_modestr (lenp)
 	PMT_MULTILINE	caller indicates that this is part of a multiline prompt
 */
 
+static int *local_prompt_newlines;
+
 static char *
 expand_prompt (pmt, flags, lp, lip, niflp, vlp)
      char *pmt;
@@ -311,7 +315,7 @@ expand_prompt (pmt, flags, lp, lip, niflp, vlp)
 {
   char *r, *ret, *p, *igstart, *nprompt, *ms;
   int l, rl, last, ignoring, ninvis, invfl, invflset, ind, pind, physchars;
-  int mlen;
+  int mlen, newlines, newlines_guess, bound;
 
   /* We only expand the mode string for the last line of a multiline prompt
      (a prompt with embedded newlines). */
@@ -345,6 +349,12 @@ expand_prompt (pmt, flags, lp, lip, niflp, vlp)
   l = strlen (nprompt);			/* XXX */
   r = ret = (char *)xmalloc (l + 1);
 
+  newlines_guess = (l / _rl_screenwidth) + 1;
+  local_prompt_newlines = (int *) xrealloc (local_prompt_newlines, sizeof (int) * (newlines_guess + 1));
+  local_prompt_newlines[newlines = 0] = 0;
+  for (rl = 1; rl <= newlines_guess; rl++)
+    local_prompt_newlines[rl] = -1;
+
   rl = physchars = 0;	/* mode string now part of nprompt */
   invfl = 0;	/* invisible chars in first line of prompt */
   invflset = 0;	/* we only want to set invfl once */
@@ -404,11 +414,25 @@ expand_prompt (pmt, flags, lp, lip, niflp, vlp)
 		ninvis++;		/* invisible chars byte counter */
 	    }
 
-	  if (invflset == 0 && rl >= _rl_screenwidth)
+	  if (invflset == 0 && physchars >= _rl_screenwidth)
 	    {
 	      invfl = ninvis;
 	      invflset = 1;
 	    }
+#if 0
+	  if (physchars >= ((newlines + 1) * _rl_screenwidth) && local_prompt_newlines[newlines+1] == -1)
+	    local_prompt_newlines[++newlines] = r - ret;
+#else
+	  if (physchars >= (bound = (newlines + 1) * _rl_screenwidth) && local_prompt_newlines[newlines+1] == -1)
+	    {
+	      int new;
+	      if (physchars > bound)		/* should rarely happen */
+		new = _rl_find_prev_mbchar (r, r - ret, MB_FIND_ANY);
+	      else
+	        new = r - ret;
+	      local_prompt_newlines[++newlines] = new;
+	    }
+#endif	  
 	}
     }
 
@@ -759,18 +783,9 @@ rl_redisplay ()
 
   /* what if lpos is already >= _rl_screenwidth before we start drawing the
      contents of the command line? */
-  while (lpos >= _rl_screenwidth)
+  if (lpos >= _rl_screenwidth)
     {
-      int z, p;
-      int nocorrect, wadjust;
-
-      nocorrect = 0;
-      /* Adjust depending on the invisible characters in the line.  We use a
-	 heuristic based on experience: invisible characters nearly always
-	 appear in the first and last lines of the prompt */
-      wadjust = (newlines == 0)
-		  ? prompt_invis_chars_first_line
-		  : ((newlines == prompt_lines_estimate) ? wrap_offset : prompt_invis_chars_first_line);
+      temp = 0;
 
       /* fix from Darin Johnson <darin@acuson.com> for prompt string with
          invisible characters that is longer than the screen width.  The
@@ -779,55 +794,18 @@ rl_redisplay ()
          probably too much work for the benefit gained.  How many people have
          prompts that exceed two physical lines?
          Additional logic fix from Edward Catmur <ed@catmur.co.uk> */
-#if defined (HANDLE_MULTIBYTE)
-      if (mb_cur_max > 1 && rl_byte_oriented == 0 && prompt_multibyte_chars > 0)
+      /* first copy the linebreaks array we computed in expand_prompt */
+      while (local_prompt_newlines[newlines+1] != -1)
 	{
-	  nocorrect = 1;
-	  n0 = num;
-          temp = local_prompt_len;
-          while (num < temp)
-	    {
-	      /* This has to take invisible characters in the prompt into
-		 account. */
-	      z = _rl_col_width  (local_prompt, n0, num, 1) - wadjust;
-	      if (z > _rl_screenwidth)
-		{
-	          num = _rl_find_prev_mbchar (local_prompt, num, MB_FIND_ANY);
-	          break;
-		}
-	      else if (z == _rl_screenwidth)
-		{
-		  /* If we are in the middle or at the end of a multibyte
-		     character, we want to move to the start, then find out
-		     where it ends so we know where to insert the newline.
-		     If this isn't a multibyte character, its the same as num++ */
-		  p = _rl_find_prev_mbchar (local_prompt, num, MB_FIND_ANY);
-		  num = _rl_find_next_mbchar (local_prompt, p, 1, MB_FIND_ANY);
-		  break;
-		}
-	      num++;
-	    }
-          temp = num;
-	}
-      else
-#endif /* !HANDLE_MULTIBYTE */
-	temp = ((newlines + 1) * _rl_screenwidth);
+	  temp = local_prompt_newlines[newlines+1];
+	  inv_lbreaks[++newlines] = temp;
+	}  
 
-      /* Now account for invisible characters in the current line. */
-      /* XXX - this assumes that the invisible characters may be split, but only
-	 between the first and the last lines. */
-      if (nocorrect == 0)
-	temp += wadjust;
-
-      inv_lbreaks[++newlines] = temp;
-#if defined (HANDLE_MULTIBYTE)
-      /* lpos is a physical cursor position, so it needs to take the invisible
-	 characters into account. */
+      /* Now set lpos from the last newline */
       if (mb_cur_max > 1 && rl_byte_oriented == 0 && prompt_multibyte_chars > 0)
-	lpos -= _rl_col_width (local_prompt, n0, num, 1) - wadjust;
+        lpos = _rl_col_width (local_prompt, temp, local_prompt_len, 1) - (wrap_offset - prompt_invis_chars_first_line);
       else
-#endif
-	lpos -= _rl_screenwidth;	/* all physical cursor positions */
+        lpos -= (_rl_screenwidth * newlines);
     }
 
   prompt_last_screen_line = newlines;
@@ -2561,9 +2539,12 @@ rl_save_prompt ()
   saved_visible_length = prompt_visible_length;
   saved_invis_chars_first_line = prompt_invis_chars_first_line;
   saved_physical_chars = prompt_physical_chars;
+  saved_local_prompt_newlines = local_prompt_newlines;
 
   local_prompt = local_prompt_prefix = (char *)0;
   local_prompt_len = 0;
+  local_prompt_newlines = (int *)0;
+
   prompt_last_invisible = prompt_visible_length = prompt_prefix_length = 0;
   prompt_invis_chars_first_line = prompt_physical_chars = 0;
 }
@@ -2573,10 +2554,13 @@ rl_restore_prompt ()
 {
   FREE (local_prompt);
   FREE (local_prompt_prefix);
+  FREE (local_prompt_newlines);
 
   local_prompt = saved_local_prompt;
   local_prompt_prefix = saved_local_prefix;
   local_prompt_len = saved_local_length;
+  local_prompt_newlines = saved_local_prompt_newlines;
+
   prompt_prefix_length = saved_prefix_length;
   prompt_last_invisible = saved_last_invisible;
   prompt_visible_length = saved_visible_length;

lib/readline/mbutil.c

@@ -1,6 +1,6 @@
 /* mbutil.c -- readline multibyte character utility functions */
 
-/* Copyright (C) 2001-2015 Free Software Foundation, Inc.
+/* Copyright (C) 2001-2017 Free Software Foundation, Inc.
 
    This file is part of the GNU Readline Library (Readline), a library
    for reading lines of text with interactive input and history editing.      
@@ -173,7 +173,14 @@ _rl_find_prev_mbchar_internal (string, seed, find_non_zero)
   prev = non_zero_prev = point = 0;
   while (point < seed)
     {
-      tmp = mbrtowc (&wc, string + point, length - point, &ps);
+      if (_rl_utf8locale && UTF8_SINGLEBYTE(string[point]))
+	{
+	  tmp = 1;
+	  wc = (wchar_t) string[point];
+	  memset(&ps, 0, sizeof(mbstate_t));
+	}
+      else
+	tmp = mbrtowc (&wc, string + point, length - point, &ps);
       if (MB_INVALIDCH ((size_t)tmp))
 	{
 	  /* in this case, bytes are invalid or too short to compose

parse.y

@@ -3757,7 +3757,7 @@ parse_comsub (qc, open, close, lenp, flags)
   while (count)
     {
 comsub_readchar:
-      ch = shell_getc (qc != '\'' && (tflags & (LEX_INCOMMENT|LEX_PASSNEXT)) == 0);
+      ch = shell_getc (qc != '\'' && (tflags & (LEX_INCOMMENT|LEX_PASSNEXT|LEX_QUOTEDDOC)) == 0);
 
       if (ch == EOF)
 	{

shell.c

@@ -223,6 +223,8 @@ int dump_po_strings;		/* Dump strings in $"..." in po format */
 int wordexp_only = 0;		/* Do word expansion only */
 int protected_mode = 0;		/* No command substitution with --wordexp */
 
+int pretty_print_mode = 0;	/* pretty-print a shell script */
+
 #if defined (STRICT_POSIX)
 int posixly_correct = 1;	/* Non-zero means posix.2 superset. */
 #else
@@ -251,6 +253,7 @@ static const struct {
   { "noprofile", Int, &no_profile, (char **)0x0 },
   { "norc", Int, &no_rc, (char **)0x0 },
   { "posix", Int, &posixly_correct, (char **)0x0 },
+  { "pretty-print", Int, &pretty_print_mode, (char **)0x0 },
 #if defined (WORDEXP_OPTION)
   { "protected", Int, &protected_mode, (char **)0x0 },
 #endif
@@ -779,6 +782,14 @@ main (argc, argv, env)
 
   shell_initialized = 1;
 
+  if (pretty_print_mode && interactive_shell)
+    {
+      internal_warning (_("pretty-printing mode ignored in interactive shells"));
+      pretty_print_mode = 0;
+    }
+  if (pretty_print_mode)
+    exit_shell (pretty_print_loop ());
+
   /* Read commands until exit condition. */
   reader_loop ();
   exit_shell (last_command_exit_value);

tests/RUN-ONE-TEST

@@ -1,4 +1,4 @@
-BUILD_DIR=/usr/local/build/chet/bash/bash-current
+BUILD_DIR=/usr/local/build/bash/bash-current
 THIS_SH=$BUILD_DIR/bash
 PATH=$PATH:$BUILD_DIR
 

tests/comsub.right

@@ -28,6 +28,7 @@ ok 4
 ok 5
 ok 6
 xyz
+ok 7
 \/tmp\/foo\/bar
 /tmp/foo/bar
 /tmp/foo/bar
@@ -48,3 +49,9 @@ b
 c
 1
 2
+d \
+g
+d \
+g
+d \
+g

tests/comsub.tests

@@ -47,3 +47,4 @@ echo ${foo:-$(echo a{b,c})} >/dev/null
 ${THIS_SH} ./comsub1.sub
 ${THIS_SH} ./comsub2.sub
 ${THIS_SH} ./comsub3.sub
+${THIS_SH} ./comsub4.sub

tests/comsub4.sub

@@ -0,0 +1,27 @@
+x=$(cat <<'EOT'
+d \
+g
+EOT
+)
+
+echo "$x"
+unset x
+
+x=$( cat <<\EOT\
+4
+d \
+g
+EOT4
+)
+
+echo "$x"
+unset x
+
+x=$( cat <<\EOT
+d \
+g
+EOT
+)
+
+echo "$x"
+