Home

Awesome

csv2rdf

Ruth Helfinstein 6/24/2012

CSV to RDF translator Does a simple CSV to RDF translation.

Outputs the rdf file as <filename>.rdf NOTE: It does not currently check if the output file already exists and will overwrite it if it does.

Looks for a configuration file called <filename>-config.csv, where <filename>.csv is the input file. Each line in the configuration file describes one item in the header line of the input csv file <csv_name>, <type>(<optional rdf_name>), <class>

where <csv_name> is the name seen in the header of the column in the csv file. <type> is either "class" or "property" or "ignore" (without quotes) <rdf_name> in parenthesis is optional, indicates name to use for this property or class in the rdf file <class> is the class (for a property) or superclass (for a class)

the lines in the config file do not have to be in the same order as the columns in the csv file with the exception of any columns with a blank name. These will be renamed "unlabeled1" "unlabeled2" and will only match if they are in the same order.

NOTE: any columns NOT listed in the configuration file will be ignored in the rdf file.

If no configuration file is found, the program will create a basic one

Files: csv2rdf.jar example.csv - sample input file example-config.csv - sample config file src - source files

To run CSV to RDF from the console: java -jar csv2rdf.jar <input csv file name> *** NOTE You can specify the name of the .csv file to use on the command line, otherwise it will default to input.csv

for example java -jar csv2rdf.jar input.csv creates input.rdf

Changes 2012.08.02 • Add configuration file

Changes 2012.07.04 • Ignore leading or trailing spaces in the csv • Add a unlabeled1,unlabled2, unlabeld3... to any attribute in the csv that is not named. • If there is a blank in the data, do not add the property name to instance
• Add additional input question: Do you want an 'all' class? If the user answers yes an "all" class is created as the only subclass to Thing and everything is subclassed to "all".