geWorkbench is a Java-base genomics platform and various analysis and visualization tools developed on that platform. This repository is the core infrastructure of that platform, especially the shared data types and GUI framework. It runs as a barebones desktop application and needs to plug in other components to support appropriate features.
License
floratos-lab/geworkbench-core
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
<html> <head><title>geWorkbench Developer Guide</title> <body> <h1>geWorkbench Developer Guide</h1> This document provides a guide for developers of geWorkbench. <h2>Project Structure</h2> This section describes the directory structure of the project. The root directory structure contains the JBuilder and IDEA project and library files. It also contains this file, the <tt>java.policy</tt> file, and the <tt>build.xml</tt> ant build file. <ul> <li><tt>_all</tt> - A folder for a placeholder module required for the IDEA IDE. <li><tt>annotation</tt> - Marker annotations for various chip types. <li><tt>components</tt> - The top-level directory for geWorkbench components. <li><tt>data</tt> - Sample data files for the project. <li><tt>lib</tt> - Library files for core geWorkbench. <li><tt>plugins</tt> - Plugins for the Cytoscape package. <li><tt>src</tt> - Source code for core geWorkbench. </ul> Each component is in a subdirectory of the <tt>components</tt> directory. It contains an IDEA <tt>.iml</tt> module file and it's directory structure is as follows: <ul> <li><tt>lib</tt> - Library files depended on by this component. <li><tt>src</tt> - Source code for the component. </ul> Note that the components can depend on the core classes (those defined in <tt>./src</tt>) and the core libraries (those defined in <tt>./lib</tt>). However, components cannot depend on each other, and the core classes cannot depend on components. <h2>Creating a New Component</h2> A new component should be added as a new subdirectory of <tt>components</tt>. The subdirectory should be given a simple, descriptive, lower-case name. It should contain <tt>src</tt> and <tt>lib</tt> directories. Most importantly, the component must not depend on the classes or libraries of any other component, nor should another component depend on it. Also, the core geWorkbench classes should not depend on its classes or libraries. The <tt>build.xml</tt> will enforce this, as will the IDEA project. However, the JBuilder project currently does not enforce this constraint, so be careful when using JBuilder. <h2>Coding Style</h2> The coding standards outlined by Sun <a href="http://java.sun.com/docs/codeconv/html/CodeConvTOC.doc.html">here</a> are observed. Here are the coding standards that are particularly important to us: <ol> <li><b>Capitalization</b> - Constants must be in all-capitals, with underscores between words: <pre> public static final int MAXIMUM_FILES = 100; </pre> <b>No</b> other identifiers may contain underscores but constants. Class, interface, enum and annotation names must be capitalized, with all subsequent words also capitalized: <pre> public class SampleClass extends AnotherSampleClass { </pre> Variable and member names must have their first letter lower-case, and subsequent words capitalized: <pre> private String fieldName; private abstract void processFile(File file); </pre> Some developers differentiate between method variables and class members by prepending <code>m_</code> or just <code>_</code> to member names. This is not a standard practice, so is not observed. Identifiers in property files are all lower-case with periods separating words: <pre> maximum.files=100 </pre> Type wildcards in generics must be a single capital letter: <pre> public interface Generic<T, S extends Serializable> { </pre> <li><b>Annotations</b> - Annotations can either appear on the same line as their associated declaration, or on the line before it: <pre> @Subscribe public void receive(Integer value) { @Script(dependencies = {STATE_UNINITIALIZED, STATE_COMPLETE}, result = STATE_INITIALIZED) public void initialize() { </pre> <li><b>Variable Naming</b> - Variables should be given verbose names, even if they are only used in relatively small code blocks. Using the auto-complete functionality of modern IDEs makes compliance with this rule painless. Exceptions are simple index variables, such as those introduced in <code>for(;;)</code> statements: <pre> for (int i = 0; i < 100; i++) { </pre> <li><b>Braces</b> - All <code>if</code>, <code>while</code> and similar block structures must be surrounded by braces, even if the body of the block consists of only one statement: <pre> if (index < 100) { System.out.println("index= " + index); } </pre> <li><b>No Shortcuts</b> - Java provides some C-style shortcuts, such as the <code>++</code>, <code>+=</code> and <code>?:</code> operators. The <code>++</code>, <code>+=</code> operators can be used by themselves (for example, <code>++</code> is often used in a <code>for</code> statement) but they should not be included in other statements. The <code>?:</code> operator should only be used very sparingly, usually in statements that construct strings. Here is an example of a <b>bad</b> use of <code>++</code>: <pre> while (i < 100) { System.out.println("i= " + i++); } </pre> This is preferred: <pre> while (i < 100) { System.out.println("i= " + i); i++; } </pre> This is an acceptable use of <code>?:</code>: <pre> boolean available; ... System.out.println("The file is " + (available ? "available" : "unavailable") + "."); </pre> However, it is a safe bet to never use <code>?:</code>. <li><b>Tabs</b> - A tab-width of 4 should be used. </ol> <h2>Code Documentation</h2> Code must be documented using the <a href="http://java.sun.com/j2se/javadoc/">Javadoc</a> standard. All classes, interfaces, enums and annotations must have an introductory Javadoc explaining its purpose. All public methods must also have explanatory Javadocs. Methods that perform non-trivial operations should have normal comments (not Javadoc) that explain the code. Such comments should precede the line (or lines) of code that they describe. For example: <pre> /** * Shuts down the componentRegistry (and terminates any pending aysnchronous dispatches). */ public void shutdown() { // Iterate through all active synch models Collection<SynchModel> models = synchModels.values(); for (SynchModel synchModel : models) { // Shut down the synch model synchModel.shutdown(); } } </pre> <h2>Packages</h2> Package names should follow the inverse-domain rule. If the domain of your organization is <code>subdomain.domain.tld</code>, then the package structure should be rooted at <code>tld.domain.subdomain</code>. Capital letters or underscore characters should never appear in package names. <p> Otherwise, the organization of packages is left up to the developer. </body> </html>
About
geWorkbench is a Java-base genomics platform and various analysis and visualization tools developed on that platform. This repository is the core infrastructure of that platform, especially the shared data types and GUI framework. It runs as a barebones desktop application and needs to plug in other components to support appropriate features.
Resources
License
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published