标签:
原文:http://commons.apache.org/proper/commons-configuration/userguide/quick_start.html
Configuration information is frequently stored in properties files. Consider the following simple file that defines some properties related to accessing a database. We assume that it is stored as database.properties in the local file system:
database.host = db.acme.com database.port = 8199 database.user = admin database.password = ??? database.timeout = 60000
The easiest way to read this file is via the Configurations helper class. This class offers a bunch of convenience methods for creating configuration objects from different sources. For reading a properties file the code looks as follows:
Configurations configs = new Configurations(); try { Configuration config = configs.properties(new File("config.properties")); // access configuration properties ... } catch (ConfigurationException cex) { // Something went wrong }
The Configuration object obtained in the last step can now be used to query the values for the stored configuration properties. For this purpose, numerous get methods for different property types are available. For the properties contained in the example file the following methods can be used:
String dbHost = config.getString("database.host"); int dbPort = config.getInt("database.port"); String dbUser = config.getString("database.user"); String dbPassword = config.getString("database.password", "secret"); // provide a default long dbTimeout = config.getLong("database.timeout");
Note that the keys passed to the get methods match the keys contained in the properties file. If a key cannot be resolved, the default behavior of a configuration is to return null. (Methods that return a primitive type throw an exception because in this case there is no null value.) It is possible to provide a default value which is used when the key cannot be found.
XML is also a suitable format for storing configuration information, especially if the data becomes more complex. For instance, lists of values can be stored in a natural way by just repeating tags. The example file for this section defines some directory paths that are to be processed by an application. It is named paths.xml and looks as follows:
<?xml version="1.0" encoding="ISO-8859-1" ?> <configuration> <processing stage="qa"> <paths> <path>/data/path1</path> <path>/data/otherpath</path> <path>/var/log</path> </paths> </processing> </configuration>
Reading this file works analogously to reading a properties file. Again a Configurations instance is needed (by the way, this class is thread-safe, and an instance can be shared and reused to read multiple configuration sources), but this time we use the xml() method rather than properties():
Configurations configs = new Configurations(); try { XMLConfiguration config = configs.xml("paths.xml"); // access configuration properties ... } catch (ConfigurationException cex) { // Something went wrong }
The xml() method returns an object of type XMLConfiguration. This class implements the Configuration interface, but offers some more functionality to access properties in a more structured manner. The reader may also have noticed that we passed a string to xml() while we used a java.io.File object in the properties example. All these methods come in several overloaded variants allowing the caller to specify the configuration source in different ways: as a file, as a URL, or as a string. In the latter case, the file is searched for in various places, including at an absolute file path, at a relative file path, as a resource in the classpath, or in the current user‘s home directory.
Accessing properties in a XML configuration (or any other hierarchical configuration) supports the same query methods as for regular configurations. There are some additional facilities that take the hierarchical nature of these sources into account. The properties in the example configuration can be read in the following way:
String stage = config.getString("processing[@stage]"); List<String> paths = config.getList(String.class, "processing.paths.path");
The keys for properties are generated by concatening the possibly nested tag names in the XML document (ignoring the root element). For attributes, there is a special syntax as shown for thestage property. Because the path element appears multiple times it actually defines a list. With the getList() method all values can be queried at once.
Hierarchical configurations support an advanced syntax for keys that allows a navigation to a specific element in the source document. This is achieved by adding numeric indices in parentheses after the single key parts. For instance, in order to reference the second path element in the list, the following key can be used (indices are 0-based):
String secondPath = config.getString("processing.paths.path(1)");
For elements which are not repeated such indices can be dropped. It is also possible to set an alternative expression engine - the component that evaluates and interprets configuration keys. There is an implementation available which can deal with XPath expressions. Refer to Expression engines for further details.
The Configuration interface defines some methods for manipulating configuration properties. Typical CRUD operations are available for all properties. The following code fragment shows how the example properties configuration can be changed. The port of the database is changed to a new value, and a new property is added:
config.setProperty("database.port", 8200);
config.addProperty("database.type", "production");
addProperty() always adds a new value to the configuration. If the affected key already exists, the value is added to this key, so that it becomes a list. setProperty() in contrast overrides an existing value (or creates a new one if the key does not exist). Both methods can be passed an arbitrary value object. This can also be an array or a collection, which makes it possible to add multiple values in a single step.
After a configuration has been manipulated, it should probably be saved again to make the changes persistent. Otherwise, the changes are only in memory. If configurations are to be changed, it is preferrable to obtain them via a different mechanism: a configuration builder. Builders are the most powerful and flexible way to construct configurations. They support many settings that impact the way the configuration data is loaded and the resulting configuration object behaves. Builders for file-based configurations also offer a save() method that writes all configuration data back to disk. Configuration builders are typically created using a fluent API which allows a convenient and flexible configuration of the builder. This API is described in the section Configuration builders. For simple use cases, the Configurations class we have already used has again some convenience methods. The following code fragment shows how a configuration is read via a builder, manipulated, and finally saved again:
Configurations configs = new Configurations(); try { // obtain the configuration FileBasedConfigurationBuilder<XMLConfiguration> builder = configs.xmlBuilder("paths.xml"); XMLConfiguration config = builder.getConfiguration(); // update property config.addProperty("newProperty", "newValue"); // save configuration builder.save(); } catch (ConfigurationException cex) { // Something went wrong }
Commons Configuration2 - Quick start guide
标签:
原文地址:http://www.cnblogs.com/huey/p/5633543.html