<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<div class="moz-cite-prefix">Le 15/10/2016 21:16, Samuel Gougeon a
écrit :<br>
</div>
<blockquote cite="mid:58028075.6000702@free.fr" type="cite">
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
<div class="moz-cite-prefix">Hello Philipp,<br>
<br>
Le 14/10/2016 00:08, Philipp Mühlmann a écrit :<br>
</div>
<blockquote
cite="mid:CACf7ODv4XbywHn_JzvKhMpS=pfFU2X8kmxu+trsSNJNVFzhoXQ@mail.gmail.com"
type="cite">
<div dir="ltr">
<div>
<div>
<div>Dear Scilab users,<br>
<br>
</div>
having a data file (*.cvs) containg following format:<br>
<br>
<br>
</div>
HEADER-Line<br>
</div>
dd.mm.yyyy, HH:MM:<a moz-do-not-send="true"
href="http://SS.MS">SS.MS</a>, value01, value02<br>
<br clear="all">
<div>
<div>
<div>
<div>
<div>dd = day<br>
</div>
<div>mm = month<br>
</div>
<div>yyyy = year<br>
<br>
</div>
<div>HH = hour<br>
</div>
<div>MM = minute<br>
</div>
<div>SS = second<br>
</div>
<div>Ms = milli second<br>
<br>
</div>
<div>ValueXY = numerical value<br>
</div>
<div><br>
</div>
<div>Is it possible to use cvsRead in such a way to
define the separator beeing ',' and ':' at the same
time?<br>
</div>
</div>
</div>
</div>
</div>
</div>
</blockquote>
.<br>
Yes and no: Yes because each character of the "separator" option
is (sometimes) considered as a<br>
separator. And no, because it sometimes fails.<br>
I was designing an example to add to the csvRead() page, when a
found this bug:<br>
<br>
<tt>--> mputl("12.08.2016, 13:54:18.45, 3.145, 548.4",
TMPDIR+"/test.csv")</tt><tt><br>
<br>
</tt><tt>--> r = "''"+csvRead("test.csv","8.",[],"string")+"''"</tt><tt><br>
</tt><tt> r = </tt><tt><br>
</tt><tt>!'12' '0' '2016, 13:54:1' '45, 3' '145, 54' '4' !</tt><tt><br>
</tt><tt>--> // it works: both "8" and "." are <b>separately</b>
considered as separators<br>
<br>
--> // now trying with "," and ":":</tt><br>
<tt>--> r = "''"+csvRead("test.csv",",:", [], "string")+"''"</tt><tt><br>
</tt><tt>--> r = "''"+csvRead("test.csv",",:", [],
"string")+"''"<br>
r = <br>
'12.08.2016, 13:54:18.45, 3.145, 548.4'<br>
--> // The row is kept as is: no splitting!<br>
--> // We could expect: [</tt><tt><tt>'12.08.2016' ' 13' '54'
'18.45' ' 3.145' ' 548.4']<br>
--> // (with leading spaces, since they are not set as
separators)<br>
--> </tt></tt><br>
<br>
There are already 10 pending reported bugs about csvRead(). One
more.<br>
</blockquote>
Report : <a class="moz-txt-link-freetext" href="http://bugzilla.scilab.org/14809">http://bugzilla.scilab.org/14809</a><br>
<br>
Each char of a separator with length > 1 is considered a specific
separator only if the whole separator pattern is met in the read
row. Otherwise, no splitting is done. Hence<tt>,<br>
</tt><tt>--> r = csvRead(TMPDIR+"test.csv","8.",[],"string") //
splits the </tt><tt><tt>"12.08.2016"</tt> row, while<br>
--></tt><tt><tt> r =
csvRead(TMPDIR+"test.csv",".8",[],"string") // does not!<br>
</tt></tt><br>
Samuel Gougeon<br>
<br>
</body>
</html>