BigQuery RegExp functions in Spark

BigQuery_re2_udfs

https://cloud.google.com/bigquery/docs/reference/standard-sql/string_functions#regexp_contains
https://cloud.google.com/bigquery/docs/reference/standard-sql/string_functions#regexp_extract
https://cloud.google.com/bigquery/docs/reference/standard-sql/string_functions#regexp_replace

RegexpContains.java

package re2_udfs;

import org.apache.spark.sql.api.java.UDF2;

import com.google.re2j.Pattern;

public class RegexpContains implements UDF2<String, String, Boolean> {
    private static final long serialVersionUID = -4480474501284793639L;

    public Boolean call(String input, String pattern) throws Exception {
        if (input == null) {
            return null;
        }

        return Pattern.compile(pattern).matcher(input).find();
    }
}

RegexpExtract.java

package re2_udfs;

import org.apache.spark.sql.api.java.UDF2;

import com.google.re2j.Matcher;
import com.google.re2j.Pattern;

public class RegexpExtract implements UDF2<String, String, String> {
    private static final long serialVersionUID = 1L;

    public String call(String input, String pattern) throws Exception {
        if (input == null) {
            return null;
        }

        Matcher m = Pattern.compile(pattern).matcher(input);

        if (m.find()) {
            return m.group(1);
        } else {
            return null;
        }
    }
}

RegexpReplace.java

package re2_udfs;

import org.apache.spark.sql.api.java.UDF3;

import com.google.re2j.Pattern;

public class RegexpReplace implements UDF3<String, String, String, String> {
    private static final long serialVersionUID = -4978515450337236613L;

    public String call(String input, String pattern, String replacement) throws Exception {
        if (input == null) {
            return null;
        }

        return Pattern.compile(pattern).matcher(input).replaceAll(replacement);
    }
}