org.aksw.qa.commons.qald

Class IQuestionCsvParser

java.lang.Object

org.aksw.qa.commons.qald.IQuestionCsvParser

public class IQuestionCsvParser
extends Object

Convert IQuestions to CSVs, in both directions.

Author:

jhuth

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
(package private) static class	IQuestionCsvParser.Column Wrapper for enum to make different enum states possible;

Field Summary

Fields

Modifier and Type	Field and Description
static String	ARRAY_SPLIT Some Columns can hold multiple values.
static String	NULL_DEFAULT The value that will be written if a column is declared for a csv but corresponding field in IQuestion is null

Constructor Summary

Constructors

Constructor and Description
IQuestionCsvParser()

Method Summary

Modifier and Type	Method and Description
private static IQuestion	csvRowToQuestion(String[] csvRow, IQuestionCsvParser.Column... columns) Works as described in csvToQuestionList(CSVReader, Column...)
static List<IQuestion>	csvToQuestionList(au.com.bytecode.opencsv.CSVReader reader, IQuestionCsvParser.Column... columns) Reads a CSV which source is defined by given CSVReader, which also defines the separation and quotation chars.
private static Boolean	parseBoolean(String boolStr) Parses a boolean.
private static List<String>	parseStringArray(String arrayStr) Splits a string on ARRAY_SPLIT.
static void	questionListToCsv(au.com.bytecode.opencsv.CSVWriter writer, boolean columnDescriptorgRow, List<IQuestion> questions, IQuestionCsvParser.Column... columns) Writes all data specified by the columns from all given IQuestions to the csvwriter.
static au.com.bytecode.opencsv.CSVReader	readerForGoogleDocsCsvExports(Reader reader, int skipLines) Returns reader which uses comma(,) as separator and (") as quote.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

ARRAY_SPLIT

public static String ARRAY_SPLIT

Some Columns can hold multiple values. So, a String representing this column content will be split on this char.

NULL_DEFAULT

public static String NULL_DEFAULT

The value that will be written if a column is declared for a csv but corresponding field in IQuestion is null

Constructor Detail

IQuestionCsvParser

public IQuestionCsvParser()

Method Detail

csvRowToQuestion

private static IQuestion csvRowToQuestion(String[] csvRow,
                                          IQuestionCsvParser.Column... columns)
                                   throws IOException

Works as described in csvToQuestionList(CSVReader, Column...)

Parameters:

csvRow - A row of a csv, with decoded terminators and split up

columns -

Returns:

Throws:

IOException

csvToQuestionList

public static List<IQuestion> csvToQuestionList(au.com.bytecode.opencsv.CSVReader reader,
                                                IQuestionCsvParser.Column... columns)
                                         throws IOException

Reads a CSV which source is defined by given CSVReader, which also defines the separation and quotation chars. CSVReader can also be set up to skip the first n lines.

Define the Column structure if the CSV by passig IQuestionCsvParser.Column Objects. The order has to fit the column order of the csv.

For example, following call:

csvToQuestionList(reader,Column.ID(), Column.ignore(), Column.question("en") )

Will parse the first columnn of the csv to IQuestion#setId() , ignore the second column, and will parse column 3 as question in english.

Trailing csv columns (no IQuestionCsvParser.Column present) will be ignored, also, additionally defined IQuestionCsvParser.Columns will be ignored.

IQuestionCsvParser.Columns can appear multiple times. For a column with one value(e.g. a boolean flag, sparql query) the rightmost column is used. For Columns holding possible array contents (e.g. golden answers, keywords), all defined columns will be parsed.

Some columns can hold arrays, e.g. IQuestionCsvParser.Column.goldenAnswers(), IQuestionCsvParser.Column.keywords(String). The array separator string can be set via ARRAY_SPLIT Arrays can be preceeded and concluded by square and curly brackets. So, a toString on most Collections would be a valid input format.

Parameters:

reader - A CSVReader fitting to your needs. For GoogleDocs CSV export, you can use readerForGoogleDocsCsvExports(Reader, int)

columns - as described above

Returns:

A list of IQuestions with the parseable information set.

Throws:

IOException - if something bad happens

questionListToCsv

public static void questionListToCsv(au.com.bytecode.opencsv.CSVWriter writer,
                                     boolean columnDescriptorgRow,
                                     List<IQuestion> questions,
                                     IQuestionCsvParser.Column... columns)
                              throws IOException

Writes all data specified by the columns from all given IQuestions to the csvwriter. Get more information from the doc of csvToQuestionList(CSVReader, Column...)

If or a declared IQuestionCsvParser.Column no information in IQuestion is present(e.g. field is null), NULL_DEFAULT will be written.

Parameters:

writer -

columnDescriptorgRow - - the first row will contain the names of the columns

questions -

columns -

Throws:

IOException

parseBoolean

private static Boolean parseBoolean(String boolStr)
                             throws IOException

Parses a boolean. Difference to Boolean.parseBoolean(String) is that no other values than {"true","false"} are accepted, ignoring case.

Parameters:

boolStr -

Returns:

a parsed boolean

Throws:

IOException - if boolStr not in {"true","false"}, ignoring case

parseStringArray

private static List<String> parseStringArray(String arrayStr)

Splits a string on ARRAY_SPLIT. First and last char can be square or curly brackets, they will be removed

Parameters:

arrayStr -

Returns:

readerForGoogleDocsCsvExports

public static au.com.bytecode.opencsv.CSVReader readerForGoogleDocsCsvExports(Reader reader,
                                                                              int skipLines)

Returns reader which uses comma(,) as separator and (") as quote. Sufficient to read data which was generated with googleDocs csv-export.

Parameters:

reader -

skipLines - skip the first n lines

Returns: