Added line end delimiter and null string support
parent
f315c1ae4b
commit
8daf918fc8
|
@ -1,15 +1,7 @@
|
|||
package com.bizo.hive.serde.csv;
|
||||
|
||||
import java.io.CharArrayReader;
|
||||
import java.io.IOException;
|
||||
import java.io.Reader;
|
||||
import java.io.StringWriter;
|
||||
import java.io.Writer;
|
||||
import java.util.ArrayList;
|
||||
import java.util.Arrays;
|
||||
import java.util.List;
|
||||
import java.util.Properties;
|
||||
|
||||
import au.com.bytecode.opencsv.CSVReader;
|
||||
import au.com.bytecode.opencsv.CSVWriter;
|
||||
import org.apache.hadoop.conf.Configuration;
|
||||
import org.apache.hadoop.hive.serde.Constants;
|
||||
import org.apache.hadoop.hive.serde2.SerDe;
|
||||
|
@ -26,8 +18,11 @@ import org.apache.hadoop.hive.serde2.typeinfo.TypeInfoUtils;
|
|||
import org.apache.hadoop.io.Text;
|
||||
import org.apache.hadoop.io.Writable;
|
||||
|
||||
import au.com.bytecode.opencsv.CSVReader;
|
||||
import au.com.bytecode.opencsv.CSVWriter;
|
||||
import java.io.*;
|
||||
import java.util.ArrayList;
|
||||
import java.util.Arrays;
|
||||
import java.util.List;
|
||||
import java.util.Properties;
|
||||
|
||||
|
||||
/**
|
||||
|
@ -45,7 +40,8 @@ public final class CSVSerde implements SerDe {
|
|||
private char separatorChar;
|
||||
private char quoteChar;
|
||||
private char escapeChar;
|
||||
|
||||
private String lineEnd;
|
||||
private String nullDefinedAs;
|
||||
|
||||
@Override
|
||||
public void initialize(final Configuration conf, final Properties tbl) throws SerDeException {
|
||||
|
@ -71,6 +67,8 @@ public final class CSVSerde implements SerDe {
|
|||
separatorChar = getProperty(tbl, "separatorChar", CSVWriter.DEFAULT_SEPARATOR);
|
||||
quoteChar = getProperty(tbl, "quoteChar", CSVWriter.DEFAULT_QUOTE_CHARACTER);
|
||||
escapeChar = getProperty(tbl, "escapeChar", CSVWriter.DEFAULT_ESCAPE_CHARACTER);
|
||||
lineEnd = tbl.getProperty("lineEnd", CSVWriter.DEFAULT_LINE_END);
|
||||
nullDefinedAs = tbl.getProperty("nullDefinedAs", "\\N");
|
||||
}
|
||||
|
||||
private final char getProperty(final Properties tbl, final String property, final char def) {
|
||||
|
@ -104,10 +102,13 @@ public final class CSVSerde implements SerDe {
|
|||
// Convert the field to Java class String, because objects of String type
|
||||
// can be stored in String, Text, or some other classes.
|
||||
outputFields[c] = fieldStringOI.getPrimitiveJavaObject(field);
|
||||
if (outputFields[c] == null) {
|
||||
outputFields[c] = nullDefinedAs;
|
||||
}
|
||||
}
|
||||
|
||||
final StringWriter writer = new StringWriter();
|
||||
final CSVWriter csv = newWriter(writer, separatorChar, quoteChar, escapeChar);
|
||||
final CSVWriter csv = newWriter(writer, separatorChar, quoteChar, escapeChar, lineEnd);
|
||||
|
||||
try {
|
||||
csv.writeNext(outputFields);
|
||||
|
@ -130,7 +131,7 @@ public final class CSVSerde implements SerDe {
|
|||
|
||||
for (int i=0; i< numCols; i++) {
|
||||
if (read != null && i < read.length) {
|
||||
row.set(i, read[i]);
|
||||
row.set(i, read[i] == nullDefinedAs ? null : read[i]);
|
||||
} else {
|
||||
row.set(i, null);
|
||||
}
|
||||
|
@ -150,7 +151,7 @@ public final class CSVSerde implements SerDe {
|
|||
}
|
||||
}
|
||||
|
||||
private CSVReader newReader(final Reader reader, char separator, char quote, char escape) {
|
||||
private static CSVReader newReader(final Reader reader, char separator, char quote, char escape) {
|
||||
// CSVReader will throw an exception if any of separator, quote, or escape is the same, but
|
||||
// the CSV format specifies that the escape character and quote char are the same... very weird
|
||||
if (CSVWriter.DEFAULT_ESCAPE_CHARACTER == escape) {
|
||||
|
@ -160,11 +161,11 @@ public final class CSVSerde implements SerDe {
|
|||
}
|
||||
}
|
||||
|
||||
private CSVWriter newWriter(final Writer writer, char separator, char quote, char escape) {
|
||||
private static CSVWriter newWriter(final Writer writer, char separator, char quote, char escape, String lineEnd) {
|
||||
if (CSVWriter.DEFAULT_ESCAPE_CHARACTER == escape) {
|
||||
return new CSVWriter(writer, separator, quote, "");
|
||||
return new CSVWriter(writer, separator, quote, lineEnd);
|
||||
} else {
|
||||
return new CSVWriter(writer, separator, quote, escape, "");
|
||||
return new CSVWriter(writer, separator, quote, escape, lineEnd);
|
||||
}
|
||||
}
|
||||
|
||||
|
|
Loading…
Reference in New Issue