question about GNU 'sort'

[Syrinx]

I'm trying to sort some Usenet headers with GNU 'sort,' and
I'm having a big problem.

I want 'sort' to consider all the characters in each record,
but the program is acting like I'm invoking it with the '-d'
option (consider only blanks and alphanumeric characters), but
I'm not.

For example if I feed sort the following example lines
(between the rows of dashes):

-------------------
(3
(1
{2
[1
[3
(2
{1
{3
[2
-------------------

I get back the following:

-------------------
(1
[1
{1
(2
[2
{2
(3
[3
{3
-------------------

Instead, I need to get this:

-------------------
(1
(2
(3
[1
[2
[3
{1
{2
{3
-------------------

I've looked over the man page, and didn't see anything to
correct this behavior, but I'm undoubtedly overlooking
something.

Can someone help, please?

And while I'm at it, does anyone know of a program like GNU
sort that offers some sort of progress indication, so that I
could send a progress report back to my main program through a
pipe or something?

Thanks!

[Syrinx]

1. turns out it was a LOCALE problem

export LC_ALL=C

And now sorting works as expected.

[RoyGBiv]

2. I'm glad you figured that out ...

'Cause I got obsessed with it, and it was driving me nuts.

I spent some time this afternoon trying to work it out, and before I knew it I had put together a script that was close to doing it when I realized it was so complicated it defeated the purpose of even having sort.

[Syrinx]

3. sorry for causing an obsession!

:)

In my quest for progress indication, I've decided that I can spawn a process (in python) with the "subprocess" module, in a subthread, and go into the 'tmp' directory and measure the sizes of the temporary files created by the sort program, as they are created.

But my data file is delimited by null characters, and python is complaining about that.

"argument 1 must be (encoded string without NULL bytes), not str"

I'm not sure I'm communicating effectively. :)

How do you send a string to subprocess.Popen including a null character?

Thanks in advance if you can answer this. :hi:

[RoyGBiv]

4. Sorry ...

I don't do Python.

I just dabble but am not an actual programmer.

[Dogmudgeon]

5. I've never run into that

Admittedly, I'm a relative n00b when it comes to Python, but I've done a fair bit of corpus (linguistic) analysis with Python and have never had a problem with null-terminated text files, no matter what end-of-line characters were used. So, for me, it handles Chr = 0, 10, and 13.

Perhaps it's another config issue like the localization problem ... ?

--p!