unofficial mirror of guile-devel@gnu.org 
 help / color / mirror / Atom feed
* patch: (read "\x1b") => #\esc
@ 2003-11-10 22:41 Paul Jarc
  2003-11-10 22:48 ` Paul Jarc
                   ` (2 more replies)
  0 siblings, 3 replies; 11+ messages in thread
From: Paul Jarc @ 2003-11-10 22:41 UTC (permalink / raw)


[-- Attachment #1: Type: TEXT/PLAIN, Size: 453 bytes --]

This patch fixes a failure to detect EOF, and also provides "\xNN" in
strings, where NN are exactly two hexadecimal digits.  This way, your
source code can be safe to print to a tty even if the strings it
defines would not be.  If this is acceptable, I'd also like to change
print.c to use this when (write)ing strings containing control
characters.

	* read.c (scm_lreadr): detect EOF after backslash, and
	interpret \xNN hexadecimal sequences.


paul

[-- Attachment #2: read-string-hex.patch --]
[-- Type: text/x-patch, Size: 1761 bytes --]

Index: guile-core/libguile/read.c
===================================================================
RCS file: /cvsroot/guile/guile/guile-core/libguile/read.c,v
retrieving revision 1.91
diff -u -r1.91 read.c
--- guile-core/libguile/read.c	4 Jun 2003 16:36:03 -0000	1.91
+++ guile-core/libguile/read.c	10 Nov 2003 22:28:07 -0000
@@ -489,7 +489,7 @@
       while ('"' != (c = scm_getc (port)))
 	{
 	  if (c == EOF)
-	    scm_input_error (FUNC_NAME, port, "end of file in string constant", SCM_EOL);
+	    str_eof: scm_input_error (FUNC_NAME, port, "end of file in string constant", SCM_EOL);
 
 	  while (j + 2 >= SCM_STRING_LENGTH (*tok_buf))
 	    scm_grow_tok_buf (tok_buf);
@@ -497,6 +497,8 @@
 	  if (c == '\\')
 	    switch (c = scm_getc (port))
 	      {
+	      case EOF:
+		goto str_eof;
 	      case '\n':
 		continue;
 	      case '0':
@@ -520,6 +522,30 @@
 	      case 'v':
 		c = '\v';
 		break;
+	      case 'x':
+		{
+		  int a, b, a_09 = 0, b_09 = 0, a_AF = 0, b_AF = 0, a_af = 0,
+		    b_af = 0;
+		  a = scm_getc (port);
+		  if (a == EOF) goto str_eof;
+		  b = scm_getc (port);
+		  if (b == EOF) goto str_eof;
+		  if      ('0' <= a && a <= '9') a_09 = 1;
+		  else if ('A' <= a && a <= 'F') a_AF = 1;
+		  else if ('a' <= a && a <= 'f') a_af = 1;
+		  if      ('0' <= b && b <= '9') b_09 = 1;
+		  else if ('A' <= b && b <= 'F') b_AF = 1;
+		  else if ('a' <= b && b <= 'f') b_af = 1;
+		  if ((a_09 || a_AF || a_af) && (b_09 || b_AF || b_af))
+		    c = (a_09? a - '0': a_AF? a - 'A' + 10: a - 'a' + 10) * 16
+		      + (b_09? b - '0': b_AF? b - 'A' + 10: b - 'a' + 10);
+		  else
+		    {
+		      scm_ungetc (b, port);
+		      scm_ungetc (a, port);
+		    }
+		  break;
+		}
 	      }
 	  SCM_STRING_CHARS (*tok_buf)[j] = c;
 	  ++j;

[-- Attachment #3: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: patch: (read "\x1b") => #\esc
  2003-11-10 22:41 patch: (read "\x1b") => #\esc Paul Jarc
@ 2003-11-10 22:48 ` Paul Jarc
  2003-11-10 23:18 ` Kevin Ryde
  2003-11-13 20:10 ` Marius Vollmer
  2 siblings, 0 replies; 11+ messages in thread
From: Paul Jarc @ 2003-11-10 22:48 UTC (permalink / raw)


Er, the subject is misleading.  Make that:
(string-ref (with-input-from-string "\"x1b\"" read) 0) => #\eof

Well, you get the idea.


paul


_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel


^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: patch: (read "\x1b") => #\esc
  2003-11-10 22:41 patch: (read "\x1b") => #\esc Paul Jarc
  2003-11-10 22:48 ` Paul Jarc
@ 2003-11-10 23:18 ` Kevin Ryde
  2003-11-11  2:45   ` Paul Jarc
  2003-11-13 20:10 ` Marius Vollmer
  2 siblings, 1 reply; 11+ messages in thread
From: Kevin Ryde @ 2003-11-10 23:18 UTC (permalink / raw)


prj@po.cwru.edu (Paul Jarc) writes:
>
> +	      case EOF:
> +		goto str_eof;

Currently that gets noticed next time around the loop does it?

(But not expecting scm_getc to return EOF a second time is probably a
good thing.)


_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel


^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: patch: (read "\x1b") => #\esc
  2003-11-10 23:18 ` Kevin Ryde
@ 2003-11-11  2:45   ` Paul Jarc
  2003-11-11 21:48     ` Kevin Ryde
  0 siblings, 1 reply; 11+ messages in thread
From: Paul Jarc @ 2003-11-11  2:45 UTC (permalink / raw)


Kevin Ryde <user42@zip.com.au> wrote:
> prj@po.cwru.edu (Paul Jarc) writes:
>> +	      case EOF:
>> +		goto str_eof;
>
> Currently that gets noticed next time around the loop does it?

Only if read() returns EOF twice, which it need not.  E.g., run
"guile -q" on a tty, and type:
guile> (string-ref "\^D^D" 0)
#\377

(The first ^D flushes the tty buffer; the second sends a single EOF.)


paul


_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel


^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: patch: (read "\x1b") => #\esc
  2003-11-11  2:45   ` Paul Jarc
@ 2003-11-11 21:48     ` Kevin Ryde
  0 siblings, 0 replies; 11+ messages in thread
From: Kevin Ryde @ 2003-11-11 21:48 UTC (permalink / raw)


prj@po.cwru.edu (Paul Jarc) writes:
>
> Only if read() returns EOF twice, which it need not.  E.g., run
> "guile -q" on a tty,

Might like that fix in the 1.6 branch too then.


_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel


^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: patch: (read "\x1b") => #\esc
  2003-11-10 22:41 patch: (read "\x1b") => #\esc Paul Jarc
  2003-11-10 22:48 ` Paul Jarc
  2003-11-10 23:18 ` Kevin Ryde
@ 2003-11-13 20:10 ` Marius Vollmer
  2003-11-18 20:04   ` Paul Jarc
  2003-11-25 17:12   ` Paul Jarc
  2 siblings, 2 replies; 11+ messages in thread
From: Marius Vollmer @ 2003-11-13 20:10 UTC (permalink / raw)


prj@po.cwru.edu (Paul Jarc) writes:

> This patch fixes a failure to detect EOF, and also provides "\xNN" in
> strings, where NN are exactly two hexadecimal digits.

Nice!

What about this change, tho: instead of ignoring a \x sequence that
does not have two hexadecmal digits, we could signal an error.
Likewise, we could signal errors for all unrecognized \ sequences.

I have applied your patch unchanged as a start.

-- 
GPG: D5D4E405 - 2F9B BCCC 8527 692A 04E3  331E FAF8 226A D5D4 E405


_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel


^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: patch: (read "\x1b") => #\esc
  2003-11-13 20:10 ` Marius Vollmer
@ 2003-11-18 20:04   ` Paul Jarc
  2003-11-19  0:24     ` Paul Jarc
  2003-11-29 23:43     ` Marius Vollmer
  2003-11-25 17:12   ` Paul Jarc
  1 sibling, 2 replies; 11+ messages in thread
From: Paul Jarc @ 2003-11-18 20:04 UTC (permalink / raw)
  Cc: guile-devel

Marius Vollmer <mvo@zagadka.de> wrote:
> What about this change, tho: instead of ignoring a \x sequence that
> does not have two hexadecmal digits, we could signal an error.
> Likewise, we could signal errors for all unrecognized \ sequences.

Will do; I was thinking that too.  Meanwhile, here's the patch for
writing.

	* print.c (scm_iprin1): use \xNN hexadecimal sequences when
	writing control characters in strings.


paul


_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel


^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: patch: (read "\x1b") => #\esc
  2003-11-18 20:04   ` Paul Jarc
@ 2003-11-19  0:24     ` Paul Jarc
  2003-11-29 23:43     ` Marius Vollmer
  1 sibling, 0 replies; 11+ messages in thread
From: Paul Jarc @ 2003-11-19  0:24 UTC (permalink / raw)
  Cc: guile-devel

[-- Attachment #1: Type: TEXT/PLAIN, Size: 66 bytes --]

I wrote:
> Meanwhile, here's the patch for writing.

Oops.


paul

[-- Attachment #2: write-string-hex.patch --]
[-- Type: text/x-patch, Size: 1222 bytes --]

Index: guile-core/libguile/print.c
===================================================================
RCS file: /cvsroot/guile/guile/guile-core/libguile/print.c,v
retrieving revision 1.155
diff -u -r1.155 print.c
--- guile-core/libguile/print.c	11 Oct 2003 00:57:25 -0000	1.155
+++ guile-core/libguile/print.c	18 Nov 2003 19:57:56 -0000
@@ -538,16 +538,24 @@
 
 	      scm_putc ('"', port);
 	      for (i = 0; i < SCM_STRING_LENGTH (exp); ++i)
-		switch (SCM_STRING_CHARS (exp)[i])
-		  {
-		  case '"':
-		  case '\\':
-		    scm_putc ('\\', port);
-		  default:
-		    scm_putc (SCM_STRING_CHARS (exp)[i], port);
-		  }
+		{
+		  unsigned char ch = SCM_STRING_CHARS (exp)[i];
+		  if ((ch < 32 && ch != '\n') || (127 <= ch && ch < 148))
+		    {
+		      static char const hex[]="0123456789abcdef";
+		      scm_putc ('\\', port);
+		      scm_putc ('x', port);
+		      scm_putc (hex [ch / 16], port);
+		      scm_putc (hex [ch % 16], port);
+		    }
+		  else
+		    {
+		      if (ch == '"' || ch == '\\')
+			scm_putc ('\\', port);
+		      scm_putc (ch, port);
+		    }
+		}
 	      scm_putc ('"', port);
-	      break;
 	    }
 	  else
 	    scm_lfwrite (SCM_STRING_CHARS (exp), SCM_STRING_LENGTH (exp), port);

[-- Attachment #3: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: patch: (read "\x1b") => #\esc
  2003-11-13 20:10 ` Marius Vollmer
  2003-11-18 20:04   ` Paul Jarc
@ 2003-11-25 17:12   ` Paul Jarc
  2003-11-30  1:05     ` Marius Vollmer
  1 sibling, 1 reply; 11+ messages in thread
From: Paul Jarc @ 2003-11-25 17:12 UTC (permalink / raw)
  Cc: guile-devel

[-- Attachment #1: Type: TEXT/PLAIN, Size: 366 bytes --]

Marius Vollmer <mvo@zagadka.de> wrote:
> What about this change, tho: instead of ignoring a \x sequence that
> does not have two hexadecmal digits, we could signal an error.
> Likewise, we could signal errors for all unrecognized \ sequences.

Here it is.

	* read.c (scm_lreadr): Signal an error for invalid escape
	sequences in strings.  Code cleanups too.


paul

[-- Attachment #2: read-string-error.patch --]
[-- Type: text/x-patch, Size: 1775 bytes --]

Index: guile-core/libguile/read.c
===================================================================
RCS file: /cvsroot/guile/guile/guile-core/libguile/read.c,v
retrieving revision 1.92
diff -u -r1.92 read.c
--- guile-core/libguile/read.c	16 Nov 2003 21:01:57 -0000	1.92
+++ guile-core/libguile/read.c	25 Nov 2003 17:06:09 -0000
@@ -524,28 +524,27 @@
 		break;
 	      case 'x':
 		{
-		  int a, b, a_09 = 0, b_09 = 0, a_AF = 0, b_AF = 0, a_af = 0,
-		    b_af = 0;
+		  int a, b;
 		  a = scm_getc (port);
 		  if (a == EOF) goto str_eof;
 		  b = scm_getc (port);
 		  if (b == EOF) goto str_eof;
-		  if      ('0' <= a && a <= '9') a_09 = 1;
-		  else if ('A' <= a && a <= 'F') a_AF = 1;
-		  else if ('a' <= a && a <= 'f') a_af = 1;
-		  if      ('0' <= b && b <= '9') b_09 = 1;
-		  else if ('A' <= b && b <= 'F') b_AF = 1;
-		  else if ('a' <= b && b <= 'f') b_af = 1;
-		  if ((a_09 || a_AF || a_af) && (b_09 || b_AF || b_af))
-		    c = (a_09? a - '0': a_AF? a - 'A' + 10: a - 'a' + 10) * 16
-		      + (b_09? b - '0': b_AF? b - 'A' + 10: b - 'a' + 10);
-		  else
-		    {
-		      scm_ungetc (b, port);
-		      scm_ungetc (a, port);
-		    }
+		  if      ('0' <= a && a <= '9') a -= '0';
+		  else if ('A' <= a && a <= 'F') a = a - 'A' + 10;
+		  else if ('a' <= a && a <= 'f') a = a - 'a' + 10;
+		  else goto bad_escaped;
+		  if      ('0' <= b && b <= '9') b -= '0';
+		  else if ('A' <= b && b <= 'F') b = b - 'A' + 10;
+		  else if ('a' <= b && b <= 'f') b = b - 'a' + 10;
+		  else goto bad_escaped;
+		  c = a * 16 + b;
 		  break;
 		}
+	      default:
+	      bad_escaped:
+		scm_input_error(FUNC_NAME, port,
+				"illegal character in escape sequence: ~S",
+				scm_list_1 (SCM_MAKE_CHAR (c)));
 	      }
 	  SCM_STRING_CHARS (*tok_buf)[j] = c;
 	  ++j;

[-- Attachment #3: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: patch: (read "\x1b") => #\esc
  2003-11-18 20:04   ` Paul Jarc
  2003-11-19  0:24     ` Paul Jarc
@ 2003-11-29 23:43     ` Marius Vollmer
  1 sibling, 0 replies; 11+ messages in thread
From: Marius Vollmer @ 2003-11-29 23:43 UTC (permalink / raw)


prj@po.cwru.edu (Paul Jarc) writes:

> Meanwhile, here's the patch for writing.

Thanks, applied!

-- 
GPG: D5D4E405 - 2F9B BCCC 8527 692A 04E3  331E FAF8 226A D5D4 E405


_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel


^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: patch: (read "\x1b") => #\esc
  2003-11-25 17:12   ` Paul Jarc
@ 2003-11-30  1:05     ` Marius Vollmer
  0 siblings, 0 replies; 11+ messages in thread
From: Marius Vollmer @ 2003-11-30  1:05 UTC (permalink / raw)


prj@po.cwru.edu (Paul Jarc) writes:

> Here it is.
>
> 	* read.c (scm_lreadr): Signal an error for invalid escape
> 	sequences in strings.  Code cleanups too.

Applied, thanks!

-- 
GPG: D5D4E405 - 2F9B BCCC 8527 692A 04E3  331E FAF8 226A D5D4 E405


_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel


^ permalink raw reply	[flat|nested] 11+ messages in thread

end of thread, other threads:[~2003-11-30  1:05 UTC | newest]

Thread overview: 11+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-11-10 22:41 patch: (read "\x1b") => #\esc Paul Jarc
2003-11-10 22:48 ` Paul Jarc
2003-11-10 23:18 ` Kevin Ryde
2003-11-11  2:45   ` Paul Jarc
2003-11-11 21:48     ` Kevin Ryde
2003-11-13 20:10 ` Marius Vollmer
2003-11-18 20:04   ` Paul Jarc
2003-11-19  0:24     ` Paul Jarc
2003-11-29 23:43     ` Marius Vollmer
2003-11-25 17:12   ` Paul Jarc
2003-11-30  1:05     ` Marius Vollmer

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).