[9fans] xd bug

[9fans] xd bug

[Pietro Gagliardi]

Hello. Regarding my previous question, the file /n/sources/contrib/
pietro/xd.out shows a bug in xd regarding characters that can't be
printed with the %c format. There should be no space between a non-
printing and a printing character; but the two spaces screw the rest
of the line up. This shows a deficiency in the table-based
implementation. I'll do my best to fix it in a clean way. Thanks.

Re: [9fans] xd bug

[Russ Cox]

> Hello. Regarding my previous question, the file /n/sources/contrib/
> pietro/xd.out shows a bug in xd regarding characters that can't be
> printed with the %c format. There should be no space between a non-
> printing and a printing character; but the two spaces screw the rest
> of the line up. This shows a deficiency in the table-based
> implementation. I'll do my best to fix it in a clean way. Thanks.

It would help a lot if (1) you cut and pasted things into email
instead of making us go look on sources for four-line files
and (2) you told us exactly what you thought was wrong.

Your xd.out says:

% xd -c -x bad
0000000  e0  Q  R  S \n
      0  e0515253 0a000000
0000005

which looks fine to me.  If I create the same file and add some
more text just to fill things out, I get:

% xd -c -b -x a | sed 6q
0000000  e0  Q  R  S \n  L  o  r  e  m     i  p  s  u  m
      0  e0 51 52 53 0a 4c 6f 72 65 6d 20 69 70 73 75 6d
      0  e0515253 0a4c6f72 656d2069 7073756d
0000010      d  o  l  o  r     s  i  t     a  m  e  t  ,
     10  20 64 6f 6c 6f 72 20 73 69 74 20 61 6d 65 74 2c
     10  20646f6c 6f722073 69742061 6d65742c

which still looks just fine.  In particular I don't see any
alignment difference between the first line and the second line.

Russ

Re: [9fans] xd bug

[Pietro Gagliardi]

If you look more closely at the alignment of the characters to the hex
values you will see that xd thinks Q has character code 0x525 and R
has character code 3. It should look more like

% xd -c -x bad
0000000  e0Q R S  \n
      0  e0515253 0a000000
0000005

On Jul 14, 2008, at 12:43 PM, Russ Cox wrote:

>> Hello. Regarding my previous question, the file /n/sources/contrib/
>> pietro/xd.out shows a bug in xd regarding characters that can't be
>> printed with the %c format. There should be no space between a non-
>> printing and a printing character; but the two spaces screw the rest
>> of the line up. This shows a deficiency in the table-based
>> implementation. I'll do my best to fix it in a clean way. Thanks.
>
> It would help a lot if (1) you cut and pasted things into email
> instead of making us go look on sources for four-line files
> and (2) you told us exactly what you thought was wrong.
>
> Your xd.out says:
>
> % xd -c -x bad
> 0000000  e0  Q  R  S \n
>      0  e0515253 0a000000
> 0000005
>
> which looks fine to me.  If I create the same file and add some
> more text just to fill things out, I get:
>
> % xd -c -b -x a | sed 6q
> 0000000  e0  Q  R  S \n  L  o  r  e  m     i  p  s  u  m
>      0  e0 51 52 53 0a 4c 6f 72 65 6d 20 69 70 73 75 6d
>      0  e0515253 0a4c6f72 656d2069 7073756d
> 0000010      d  o  l  o  r     s  i  t     a  m  e  t  ,
>     10  20 64 6f 6c 6f 72 20 73 69 74 20 61 6d 65 74 2c
>     10  20646f6c 6f722073 69742061 6d65742c
>
> which still looks just fine.  In particular I don't see any
> alignment difference between the first line and the second line.
>
> Russ
>
>

Re: [9fans] xd bug

[erik quanstrom]

> If you look more closely at the alignment of the characters to the hex
> values you will see that xd thinks Q has character code 0x525 and R
> has character code 3. It should look more like
>
> % xd -c -x bad
> 0000000  e0Q R S  \n
>       0  e0515253 0a000000
> 0000005

-x by itself doesn't output the hex codes
for bytes, it outputs the hex codes for 4 byte
integers.  i think you're thinking of this command
line instead

; xd -c -1x bad
0000000  e0  Q  R  S \n
      0  e0 51 52 53 0a

- erik

Re: [9fans] xd bug

[Pietro Gagliardi]

On Jul 14, 2008, at 1:13 PM, erik quanstrom wrote:
> -x by itself doesn't output the hex codes
> for bytes, it outputs the hex codes for 4 byte
> integers.  i think you're thinking of this command
> line instead
>
> ; xd -c -1x bad
> 0000000  e0  Q  R  S \n
>      0  e0 51 52 53 0a


On Jul 14, 2008, at 1:22 PM, Russ Cox wrote:
> The -c and -x formats are not intended to align.
> If you want hex codes aligned with characters, use -b.

Thanks for the clarifications. -b == -1x, so both solve this
misunderstanding.

Re: [9fans] xd bug

[Russ Cox]

> If you look more closely at the alignment of the characters to the hex
> values you will see that xd thinks Q has character code 0x525 and R
> has character code 3. It should look more like
>
> % xd -c -x bad
> 0000000  e0Q R S  \n
>       0  e0515253 0a000000
> 0000005

The -c and -x formats are not intended to align.
If you want hex codes aligned with characters, use -b.

% echo ABCD | xd -c -x
0000000   A  B  C  D \n
      0  41424344 0a000000
0000005
%

Russ