Pedro Tutorials

Writing Pedro Plugins


This section is intended for the developers in early adopter groups. The goal here is to allow people to begin writing their own program modules that perform data import, data export, validation or analysis on a subtree of records beginning at a root record of type X. Here are some plugin ideas; for now most of these will mean something to proteomics scientists. In the near future this documentation will be greatly improved. Examples:

If you aren't in proteomics, don't worry. Your steps as a developer are going to be fairly straight forward:

  1. Create a Java class that implements one of these interfaces:
    • pedro.plugins.DataImportPlugin
    • pedro.plugins.DataExportPlugin
    • pedro.plugins.ValidationPlugin
    • pedro.plugins.AnalysisPlugin
  2. Create a jar file for the plugin classes you make. We're going to assume for now you know how to use the Java Jar utility. As a quick reminder, the usual way of invoking it looks something like this: "jar cvf myplugin.jar ./myplugins";
  3. rename the file so it ends with ".plugins" instead of ".jar";
  4. put it in the "model/lib" directory;
  5. Load the Pedro data capture tool with the appropriate model. Pedro loads treats all files ending in .jar or .plugins as JAR files that contain class files. It specifically looks in .plugins files for those classes that will implement one of the plugin interfaces already mentioned.

    These plugins are instantiated and held in memory. When the end-user moves from one record to the next, Pedro asks its collection of plugins can deal with Record type X as it is defined in the current record model version Y.

    As a plugin developer you can make the plugin respond in any way you want. If the plugin agrees it can do something with it, then it appears in the Services menu.

    If at least one plugin can do something the menu glows red and the status bar shows the kinds of plugins available.

  6. That's it! This can now be revisited in more detail.