Package de.lmu.ifi.dbs.elki.logging

Logging facility for controlling logging behavior of the complete framework.

See: Description


Package de.lmu.ifi.dbs.elki.logging Description

Logging facility for controlling logging behavior of the complete framework.

Logging in ELKI

Logging in ELKI is closely following the Logger approach.

However, system-wide configuration of logging does not seem appropriate, therefore ELKI uses a configuration file named

logging-cli.properties
living in the package de.lmu.ifi.dbs.elki.logging (or an appropriately named directory) for command line interface based operation.

Logging levels can be configured on a per-class or per-package level using e.g.:

de.lmu.ifi.dbs.elki.index.level = FINE

to set the logging level for the index structure package to FINE.

Logging for Developers:

Developers working in ELKI are encouraged to use the following setup to make configurable logging:

  1. Introduce one or multiple static final debug flags in their classes:

    protected static final boolean debug = true || LoggingConfiguration.DEBUG;

    After development, it should be changed to false || LoggingConfiguration.DEBUG.

  2. If the class contains 'frequent' logging code, acquire a static Logger reference:

    protected static final Logging logger = Logging.getLogger(Example.class);
  3. Wrap logging statements in appropriate level checks:

    if (logger.isVerbose()) { // compute logging message logger.verbose(expensive + message + construction); }
  4. For infrequent logging, the following static convenience function is appropriate:

    LoggingUtil.exception("Out of memory in algorithm.", exception);

    This function is expensive (it acquires a stack trace to obtain class and method references, retrieves a logger reference etc.) and thus should only be used for 'rare' logging events.

  5. In cases where many tests would occur, also consider using:

    final boolean verbose = logger.isVerbose();
    // ... for, while, anything expensive
    if (verbose) {
      logger.verbose(...);
    }
    
ELKI Version 0.7.0

Copyright © 2015 ELKI Development Team, Lehr- und Forschungseinheit für Datenbanksysteme, Ludwig-Maximilians-Universität München. License information.