On June 8, 2011, Paul Bowyer wrote:
--------------------------
I tried your code and it works on the ndata9.dat that you sent along
because that file contains only LF EOL chars, but when I tried it on
ndata2.dat, which contains CRLF sequences, it chokes. Here's the marked
up code that shows where it goes astray.
------------------------------------------------------------------------------------------------------------------------------
/*cint_val(%ss) := cint(charat(string(?read\-char(%ss)),3));*/
/*This might be more efficient and it gives the same result.*/
cint_val(%ss) := cint(cunlisp(?read\-char(%ss)));
------------------------------
Hi Paul,
Let's try a new function and a new experiment. I am
attaching three one line data files:
line1w.txt is: ABCD[CR][LF] ( ie a windows text file)
line1u.txt is ABCD[LF] (ie a unix text file )
line1m.txt is ABCD[CR] (ie a mac text file )
The new function eol_chars returns a list of the
end of line chars found on the first line of any
given file.
When I run it on these three one line files I
get:
------------------------
(%i1) display2d:false$
(%i2) load(eol_chars);
(%o2) "c:/work2/eol_chars.mac"
(%i3) printfile("line1w.txt")$
ABCD
(%i4) eol_chars ("line1w.txt");
(%o4) [13,10]
(%i5) printfile ("line1u.txt")$
ABCD
(%i6) eol_chars ("line1u.txt");
(%o6) [10]
(%i7) printfile ("line1m.txt")$
ABCD
(%i8) eol_chars ("line1m.txt");
(%o8) [13]
and just as a side check on previous
work:
(%i9) eol_chars ("ndata9.dat");
(%o9) [13,10]
-----------------------------------
The code for eol_chars is (thanks for
your suggestion to use cunlisp):
eol_chars(%fname) :=
block([%s,%lch,eolL :[],%nch,%Nf : 0 ],
%s : openr (%fname),
do
( %lch : ?read\-char (%s,false),
if %lch = false then return(),
%nch : cint (cunlisp (%lch)),
if not lfreeof ([10,13],%nch) then
(%Nf : %Nf + 1,
eolL : cons (%nch,eolL),
if %Nf = 2 then return ())
else
if %Nf = 1 then return()),
close(%s),
reverse (eolL))$
-------------------------
I am concerned that you found ndata9.dat to have a unix eol char (10),
because the file I attached has windows eol chars (13,10), as confirmed
above.
Is there some automatic conversion of data and text files into
unix end of line chars going on with your system??
Ted
-------------- next part --------------
A non-text attachment was scrubbed...
Name: eol_chars.mac
Type: image/x-macpaint
Size: 889 bytes
Desc: not available
URL: <http://www.math.utexas.edu/pipermail/maxima/attachments/20110608/1388e547/attachment.bin>;
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: line1m.txt
URL: <http://www.math.utexas.edu/pipermail/maxima/attachments/20110608/1388e547/attachment.txt>;
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: line1u.txt
URL: <http://www.math.utexas.edu/pipermail/maxima/attachments/20110608/1388e547/attachment-0001.txt>;
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: line1w.txt
URL: <http://www.math.utexas.edu/pipermail/maxima/attachments/20110608/1388e547/attachment-0002.txt>;
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ndata9.dat
Type: application/octet-stream
Size: 93 bytes
Desc: not available
URL: <http://www.math.utexas.edu/pipermail/maxima/attachments/20110608/1388e547/attachment.obj>;