CSVObjects

The main class for the framework is {@link net.sf.anupam.csv.CSVParser CSVParser}, which provides a {@link java.lang.Iterable Iterable} interface for parsing and retrieving the POJO beans from the CSV records transparently. Configuration of the parser is done via XML mapping files.

The {@link net.sf.anupam.csv.CSVParserFactory CSVParserFactory} is the main factory for creation and instantiation on the CSV parsers.

Implementation of the framework is based on Stephen Ostermiller's excellent CSV parsing engine.

Configuration

The parser framework provides convenience Xdoclet tagsets and templates for auto-generation of the mapping XML file directly from the POJOs being mapped. This is similar to the Xdoclet support for Hibernate persistance framework, and behaves in a similar fashion in terms of the mapping and configuration generation.

Format of the file is:
<?xml version="1.0"?>
<csv-mapping>
<bean-mapping name="declarative name of the POJO bean to be mapped"
class="fully qualified name of the POJO's class">
</bean-mapping>
</csv-mapping>

Logging and Debugging

The framework uses the Jakarta Commons logging as its logging framework. Any standard setup (including Log4j) can be used as the actual logger.

Notes

• Note that the current implementation is read-only and does not offer any mechanism for writing out Java objects to CSV files.
• The current implementation is Java 1.5 only

In most usage situations, no formatting is required on the CSV fields during parsing, and data formatting can be performed on the source CSV data itself.

However, minor formatting is necessary in certain circumstances, e.g., when only part of a CSV field value needs to be mapped to a POJO attribute. Common examples include extracting the first name from a full name, or changing the case of the input string value. This functionality is enabled by this package.

Default Formatters

The following formatters are present in the default system. The declarative formatter name is in brackets. Details on each formatter is present in the respective class' documentation.

• {@link net.sf.anupam.csv.formatters.DoNothingFormatter DoNothingFormatter} (none)
• {@link net.sf.anupam.csv.formatters.DoNothingFormatter firstWordFormatter} (firstWord)
• {@link net.sf.anupam.csv.formatters.DoNothingFormatter lastWordFormatter} (lastWord)
• {@link net.sf.anupam.csv.formatters.DoNothingFormatter AllUpperCaseFormatter} (AllUpperCase)

Configuration of Formatters

The formatters are registered into the CSV parser framework via the csv-formatter-config.xml XML file. The file defines the logical formatter name, and the formatter implementation class name.

Custom Formatters

Customer formatters can be created by implementing the {@link net.sf.anupam.csv.formatters.CSVFieldFormatter CSVFieldFormatter} interface, and its sole method: String format(String). This method acts as the transformation/formatting implementation. It is recommended that the custom formatters be stateless and provide re-entrant implementations.

Once the custom formatters are defined, they can be registered in the Framework by updating the csv-formatter-config.xml configuration file. A custom Xdoclet template/tagset is provided for ease of generating the configuration XML file.