Documentation/Source/eric5.DebugClients.Python.PyProfile.html

<!DOCTYPE html>
<html><head>
<title>eric5.DebugClients.Python.PyProfile</title>
<meta charset="UTF-8">
<style>
body {
    background: #EDECE6;
    margin: 0em 1em 10em 1em;
    color: black;
}

h1 { color: white; background: #85774A; }
h2 { color: white; background: #85774A; }
h3 { color: white; background: #9D936E; }
h4 { color: white; background: #9D936E; }
    
a { color: #BA6D36; }

</style>
</head>
<body><a NAME="top" ID="top"></a>
<h1>eric5.DebugClients.Python.PyProfile</h1>
<p>
Module defining additions to the standard Python profile.py.
</p>
<h3>Global Attributes</h3>
<table>
<tr><td>None</td></tr>
</table>
<h3>Classes</h3>
<table>
<tr>
<td><a href="#PyProfile">PyProfile</a></td>
<td>Class extending the standard Python profiler with additional methods.</td>
</tr>
</table>
<h3>Functions</h3>
<table>
<tr><td>None</td></tr>
</table>
<hr /><hr />
<a NAME="PyProfile" ID="PyProfile"></a>
<h2>PyProfile</h2>
<p>
    Class extending the standard Python profiler with additional methods.
</p><p>
    This class extends the standard Python profiler by the functionality to
    save the collected timing data in a timing cache, to restore these data
    on subsequent calls, to store a profile dump to a standard filename and
    to erase these caches.
</p>
<h3>Derived from</h3>
profile.Profile
<h3>Class Attributes</h3>
<table>
<tr><td>dispatch</td></tr>
</table>
<h3>Class Methods</h3>
<table>
<tr><td>None</td></tr>
</table>
<h3>Methods</h3>
<table>
<tr>
<td><a href="#PyProfile.__init__">PyProfile</a></td>
<td>Constructor</td>
</tr><tr>
<td><a href="#PyProfile.__restore">__restore</a></td>
<td>Private method to restore the timing data from the timing cache.</td>
</tr><tr>
<td><a href="#PyProfile.dump_stats">dump_stats</a></td>
<td>Public method to dump the statistics data.</td>
</tr><tr>
<td><a href="#PyProfile.erase">erase</a></td>
<td>Public method to erase the collected timing data.</td>
</tr><tr>
<td><a href="#PyProfile.fix_frame_filename">fix_frame_filename</a></td>
<td>Public method used to fixup the filename for a given frame.</td>
</tr><tr>
<td><a href="#PyProfile.save">save</a></td>
<td>Public method to store the collected profile data.</td>
</tr><tr>
<td><a href="#PyProfile.trace_dispatch_call">trace_dispatch_call</a></td>
<td>Public method used to trace functions calls.</td>
</tr>
</table>
<h3>Static Methods</h3>
<table>
<tr><td>None</td></tr>
</table>
<a NAME="PyProfile.__init__" ID="PyProfile.__init__"></a>
<h4>PyProfile (Constructor)</h4>
<b>PyProfile</b>(<i>basename, timer=None, bias=None</i>)
<p>
        Constructor
</p><dl>
<dt><i>basename</i></dt>
<dd>
name of the script to be profiled (string)
</dd><dt><i>timer</i></dt>
<dd>
function defining the timing calculation
</dd><dt><i>bias</i></dt>
<dd>
calibration value (float)
</dd>
</dl><a NAME="PyProfile.__restore" ID="PyProfile.__restore"></a>
<h4>PyProfile.__restore</h4>
<b>__restore</b>(<i></i>)
<p>
        Private method to restore the timing data from the timing cache.
</p><a NAME="PyProfile.dump_stats" ID="PyProfile.dump_stats"></a>
<h4>PyProfile.dump_stats</h4>
<b>dump_stats</b>(<i>file</i>)
<p>
        Public method to dump the statistics data.
</p><dl>
<dt><i>file</i></dt>
<dd>
name of the file to write to (string)
</dd>
</dl><a NAME="PyProfile.erase" ID="PyProfile.erase"></a>
<h4>PyProfile.erase</h4>
<b>erase</b>(<i></i>)
<p>
        Public method to erase the collected timing data.
</p><a NAME="PyProfile.fix_frame_filename" ID="PyProfile.fix_frame_filename"></a>
<h4>PyProfile.fix_frame_filename</h4>
<b>fix_frame_filename</b>(<i>frame</i>)
<p>
        Public method used to fixup the filename for a given frame.
</p><p>
        The logic employed here is that if a module was loaded
        from a .pyc file, then the correct .py to operate with
        should be in the same path as the .pyc. The reason this
        logic is needed is that when a .pyc file is generated, the
        filename embedded and thus what is readable in the code object
        of the frame object is the fully qualified filepath when the
        pyc is generated. If files are moved from machine to machine
        this can break debugging as the .pyc will refer to the .py
        on the original machine. Another case might be sharing
        code over a network... This logic deals with that.
</p><dl>
<dt><i>frame</i></dt>
<dd>
the frame object
</dd>
</dl><dl>
<dt>Returns:</dt>
<dd>
fixed up file name (string)
</dd>
</dl><a NAME="PyProfile.save" ID="PyProfile.save"></a>
<h4>PyProfile.save</h4>
<b>save</b>(<i></i>)
<p>
        Public method to store the collected profile data.
</p><a NAME="PyProfile.trace_dispatch_call" ID="PyProfile.trace_dispatch_call"></a>
<h4>PyProfile.trace_dispatch_call</h4>
<b>trace_dispatch_call</b>(<i>frame, t</i>)
<p>
        Public method used to trace functions calls.
</p><p>
        This is a variant of the one found in the standard Python
        profile.py calling fix_frame_filename above.
</p><dl>
<dt><i>frame</i></dt>
<dd>
reference to the call frame
</dd><dt><i>t</i></dt>
<dd>
arguments of the call
</dd>
</dl><dl>
<dt>Returns:</dt>
<dd>
flag indicating a handled call
</dd>
</dl>
<div align="right"><a href="#top">Up</a></div>
<hr />
</body></html>