com.google.refine.expr.functions.strings.DetectLanguage Maven / Gradle / Ivy

Go to download

Show more of this group Show more artifacts with this name
Show all versions of main Show documentation

OpenRefine is a free, open source power tool for working with messy data and improving it

There is a newer version: 3.8.7


package com.google.refine.expr.functions.strings;

import java.io.IOException;
import java.util.Properties;

import com.google.common.base.Optional;
import com.optimaize.langdetect.i18n.LdLocale;

import com.google.refine.expr.EvalError;
import com.google.refine.grel.ControlFunctionRegistry;
import com.google.refine.grel.EvalErrorMessage;
import com.google.refine.grel.Function;
import com.google.refine.grel.FunctionDescription;
import com.google.refine.util.DetectLanguageUtils;

public class DetectLanguage implements Function {

    /**
     * Detects the language of the given string and provides the language code.
     * 
     * @param bindings
     *            bindings
     * @param args
     *            arguments
     * @return the language code of the string
     */
    @Override
    public Object call(Properties bindings, Object[] args) {
        if (args.length == 1) {
            Object obj = args[0]; // get the first argument
            if (obj instanceof String) { // if it is a string
                String text = (String) obj; // get the string
                if (text.length() > 0) { // if the string is not empty
                    try { // try to detect the language
                        Optional lang = DetectLanguageUtils.detect(text); // detect the language
                        if (lang.isPresent()) { // if the language is detected
                            return lang.get().getLanguage(); // return the language code
                        } else { // if the language is not detected
                            return new EvalError(EvalErrorMessage.language_detect_failed(ControlFunctionRegistry.getFunctionName(this)));
                        }
                    } catch (IOException e) { // if the language detection failed
                        e.printStackTrace(); // print the stack trace
                    }
                }
            }
        }
        return new EvalError(EvalErrorMessage.expects_one_string(ControlFunctionRegistry.getFunctionName(this)));
    }

    @Override
    public String getDescription() {
        return FunctionDescription.str_detect_language();
    }

    @Override
    public String getParams() {
        return "string s";
    }

    @Override
    public String getReturns() {
        return "string";
    }
}