<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
</head>
<body bgcolor="#ffffff" text="#000000">
<tt>I didn't reply to the list like I had intended to. (Thunderbird 3
is different enough from 2 that I'm not yet adjusted.)<br>
<br>
Ray<br>
<br>
</tt><br>
-------- Original Message --------
<table class="moz-email-headers-table" border="0" cellpadding="0"
cellspacing="0">
<tbody>
<tr>
<th align="RIGHT" valign="BASELINE" nowrap="nowrap">Subject: </th>
<td>Re: [cmucl-help] TRANSLATE-PATHNAME Bug?</td>
</tr>
<tr>
<th align="RIGHT" valign="BASELINE" nowrap="nowrap">Date: </th>
<td>Fri, 29 Jan 2010 12:57:34 -0500</td>
</tr>
<tr>
<th align="RIGHT" valign="BASELINE" nowrap="nowrap">From: </th>
<td>Raymond Toy <a class="moz-txt-link-rfc2396E" href="mailto:toy.raymond@gmail.com"><toy.raymond@gmail.com></a></td>
</tr>
<tr>
<th align="RIGHT" valign="BASELINE" nowrap="nowrap">To: </th>
<td>Marco Antoniotti <a class="moz-txt-link-rfc2396E" href="mailto:marcoxa@cs.nyu.edu"><marcoxa@cs.nyu.edu></a></td>
</tr>
</tbody>
</table>
<br>
<br>
<pre>On 1/29/10 11:55 AM, Marco Antoniotti wrote:
> Hello
>
> I think this is a bug (although it is 19D)
>
>
> * (translate-pathname "/Users/marcoxa/Projects/Lang/CL/HELAMBDAP/
> helambdap-pkg.lisp"
> "/Users/marcoxa/Projects/Lang/CL/**/*"
> "**/*")
>
> #P"/HELAMBDAP/helambdap-pkg.lisp"
> * (describe *)
>
> #P"/HELAMBDAP/helambdap-pkg.lisp" is a structure of type PATHNAME.
> HOST: #<LISP::UNIX-HOST>.
> DEVICE: NIL.
> DIRECTORY: (:ABSOLUTE "HELAMBDAP").
> NAME: "helambdap-pkg".
> TYPE: "lisp".
> VERSION: NIL.
> *
>
> I think the result should be a :RELATIVE pathname.
>
> Is it still like this in the current version?
>
>
Yes, it still works this way.
I can't figure out from the CLHS what this should return. However, both
clisp and clozure cl return a relative pathname. Based on the comments
for TRANSLATE-DIRECTORIES, cmucl has basically always returned a
pathname directory that is the same kind as the source.
I think I prefer your interpretation. We can always trivially get back
the old result if the TO pattern is an absolute path.
Ray
</pre>
</body>
</html>