Geografic-Information-System/general/g.tempfile/g.tempfile.html

<h2>DESCRIPTION</h2>

<em>g.tempfile</em>
is designed for shell scripts that need to use large temporary files. 
GRASS provides a mechanism for temporary files that does not depend on 
/tmp. GRASS temporary files are created in the data base with the assumption 
that there will be enough space under the data base for large files. 
GRASS periodically removes temporary files that have been left behind 
by programs that failed to remove them before terminating. 

<p>
<em>g.tempfile</em>
creates an unique file and prints the name. The user is required to provide 
a process-id which will be used as part of the name of the file. 
Most Unix shells provide a way to get the process id of the current shell. 
For /bin/sh and /bin/csh this is $$. 
It is recommended that $$ be specified as the process-id for 
<em>g.tempfile</em>.

<h2>EXAMPLE</h2>

For /bin/sh scripts the following syntax should be used: 
<div class="code"><pre>
temp1=`g.tempfile pid=$$`
temp2=`g.tempfile pid=$$`
</pre></div>
For /bin/csh scripts, the following can be used: 
<div class="code"><pre>
set temp1=`g.tempfile pid=$$`
set temp2=`g.tempfile pid=$$`
</pre></div>

<h2>NOTES</h2>

Each call to <em>g.tempfile</em>
creates a different (i.e. unique) name. 

Although GRASS does eventually get around to removing
tempfiles that have been left behind, the programmer should
make every effort to remove these files. They often get
large and take up disk space. If you write /bin/sh scripts,
learn to use the /bin/sh <em>trap</em> command. If you
write /bin/csh scripts, learn to use the /bin/csh
<em>onintr</em> command.

<h2>AUTHOR</h2>

Michael Shapiro, 
U.S. Army Construction Engineering 
Research Laboratory

<p><i>Last changed: $Date$</i>