I think the changes for adding byte arrays to mutation are great and
appreciate them.
I do not want to start another dicussion about encoding. The javadoc
for mutation states the following:
Convenience methods which takes columns
and value as CharSequence (String implements
CharSequence) are provided. CharSequence is
converted to UTF-8 by constructing a new Text
object.
This is the documented behavior of Mutation in released code. Whether
its good or bad is a seperate issue I do not want to discuss :)
Changing the behavior could break existing code that relies on the
documented behavior. I will commit a change in a bit for the
CharSequence related methods. Just want to point why I am doing this.
Keith
On Tue, Nov 13, 2012 at 2:20 PM, <[email protected]> wrote:
> Author: medined
> Date: Tue Nov 13 19:19:59 2012
> New Revision: 1408900
>
> URL: http://svn.apache.org/viewvc?rev=1408900&view=rev
> Log:
> ACCUMULO-851: Add Mutation Constructor Accepting Byte Array. Followup to have
> put() use byte arrays too.
>
> Modified:
>
> accumulo/trunk/core/src/main/java/org/apache/accumulo/core/data/Mutation.java
>
> Modified:
> accumulo/trunk/core/src/main/java/org/apache/accumulo/core/data/Mutation.java
> URL:
> http://svn.apache.org/viewvc/accumulo/trunk/core/src/main/java/org/apache/accumulo/core/data/Mutation.java?rev=1408900&r1=1408899&r2=1408900&view=diff
> ==============================================================================
> ---
> accumulo/trunk/core/src/main/java/org/apache/accumulo/core/data/Mutation.java
> (original)
> +++
> accumulo/trunk/core/src/main/java/org/apache/accumulo/core/data/Mutation.java
> Tue Nov 13 19:19:59 2012
> @@ -27,7 +27,6 @@ import java.util.List;
> import org.apache.accumulo.core.data.thrift.TMutation;
> import org.apache.accumulo.core.security.ColumnVisibility;
> import org.apache.accumulo.core.util.ByteBufferUtil;
> -import org.apache.accumulo.core.util.TextUtil;
> import org.apache.hadoop.io.Text;
> import org.apache.hadoop.io.Writable;
> import org.apache.hadoop.io.WritableUtils;
> @@ -87,12 +86,6 @@ public class Mutation implements Writabl
>
> }
>
> - void add(byte[] b) {
> - reserve(b.length);
> - System.arraycopy(b, 0, data, offset, b.length);
> - offset += b.length;
> - }
> -
> public void add(byte[] bytes, int off, int length) {
> reserve(length);
> System.arraycopy(bytes, off, data, offset, length);
> @@ -238,14 +231,20 @@ public class Mutation implements Writabl
> return row;
> }
>
> + public static String toHexString(byte[] ba) {
> + StringBuilder str = new StringBuilder();
> + for (int i = 0; i < ba.length; i++)
> + str.append(String.format("%x", ba[i]));
> + return str.toString();
> + }
> +
> private void put(byte b[]) {
> - buffer.writeVLong(b.length);
> - buffer.add(b);
> + put(b, b.length);
> }
>
> - private void put(Text t) {
> - buffer.writeVLong(t.getLength());
> - buffer.add(t.getBytes(), 0, t.getLength());
> + private void put(byte b[], int length) {
> + buffer.writeVLong(length);
> + buffer.add(b, 0, length);
> }
>
> private void put(boolean b) {
> @@ -260,24 +259,36 @@ public class Mutation implements Writabl
> buffer.writeVLong(l);
> }
>
> + private void put(byte[] cf, byte[] cq, byte[] cv, boolean hasts, long ts,
> boolean deleted, byte[] val) {
> + put(cf, cf.length, cq, cq.length, cv, hasts, ts, deleted, val);
> + }
> +
> + /*
> + * When dealing with Text object the length must be gotten from the
> object, not from the byte array.
> + */
> private void put(Text cf, Text cq, byte[] cv, boolean hasts, long ts,
> boolean deleted, byte[] val) {
> -
> - if (buffer == null)
> + put(cf.getBytes(), cf.getLength(), cq.getBytes(), cq.getLength(), cv,
> hasts, ts, deleted, val);
> + }
> +
> + private void put(byte[] cf, int cfLength, byte[] cq, int cqLength, byte[]
> cv, boolean hasts, long ts, boolean deleted, byte[] val) {
> + if (buffer == null) {
> throw new IllegalStateException("Can not add to mutation after
> serializing it");
> -
> - put(cf);
> - put(cq);
> + }
> + put(cf, cfLength);
> + put(cq, cqLength);
> put(cv);
> put(hasts);
> - if (hasts)
> + if (hasts) {
> put(ts);
> + }
> put(deleted);
>
> if (val.length < VALUE_SIZE_COPY_CUTOFF) {
> put(val);
> } else {
> - if (values == null)
> + if (values == null) {
> values = new ArrayList<byte[]>();
> + }
> byte copy[] = new byte[val.length];
> System.arraycopy(val, 0, copy, 0, val.length);
> values.add(copy);
> @@ -287,14 +298,6 @@ public class Mutation implements Writabl
> entries++;
> }
>
> - private void put(CharSequence cf, CharSequence cq, byte[] cv, boolean
> hasts, long ts, boolean deleted, byte[] val) {
> - put(new Text(cf.toString()), new Text(cq.toString()), cv, hasts, ts,
> deleted, val);
> - }
> -
> - private void put(CharSequence cf, CharSequence cq, byte[] cv, boolean
> hasts, long ts, boolean deleted, CharSequence val) {
> - put(cf, cq, cv, hasts, ts, deleted, TextUtil.getBytes(new
> Text(val.toString())));
> - }
> -
> public void put(Text columnFamily, Text columnQualifier, Value value) {
> put(columnFamily, columnQualifier, EMPTY_BYTES, false, 0l, false,
> value.get());
> }
> @@ -328,51 +331,51 @@ public class Mutation implements Writabl
> }
>
> public void put(CharSequence columnFamily, CharSequence columnQualifier,
> Value value) {
> - put(columnFamily, columnQualifier, EMPTY_BYTES, false, 0l, false,
> value.get());
> + put(columnFamily.toString().getBytes(),
> columnQualifier.toString().getBytes(), EMPTY_BYTES, false, 0l, false,
> value.get());
> }
>
> public void put(CharSequence columnFamily, CharSequence columnQualifier,
> ColumnVisibility columnVisibility, Value value) {
> - put(columnFamily, columnQualifier, columnVisibility.getExpression(),
> false, 0l, false, value.get());
> + put(columnFamily.toString().getBytes(),
> columnQualifier.toString().getBytes(), columnVisibility.getExpression(),
> false, 0l, false, value.get());
> }
>
> public void put(CharSequence columnFamily, CharSequence columnQualifier,
> long timestamp, Value value) {
> - put(columnFamily, columnQualifier, EMPTY_BYTES, true, timestamp, false,
> value.get());
> + put(columnFamily.toString().getBytes(),
> columnQualifier.toString().getBytes(), EMPTY_BYTES, true, timestamp, false,
> value.get());
> }
>
> public void put(CharSequence columnFamily, CharSequence columnQualifier,
> ColumnVisibility columnVisibility, long timestamp, Value value) {
> - put(columnFamily, columnQualifier, columnVisibility.getExpression(),
> true, timestamp, false, value.get());
> + put(columnFamily.toString().getBytes(),
> columnQualifier.toString().getBytes(), columnVisibility.getExpression(),
> true, timestamp, false, value.get());
> }
>
> public void putDelete(CharSequence columnFamily, CharSequence
> columnQualifier) {
> - put(columnFamily, columnQualifier, EMPTY_BYTES, false, 0l, true,
> EMPTY_BYTES);
> + put(columnFamily.toString().getBytes(),
> columnQualifier.toString().getBytes(), EMPTY_BYTES, false, 0l, true,
> EMPTY_BYTES);
> }
>
> public void putDelete(CharSequence columnFamily, CharSequence
> columnQualifier, ColumnVisibility columnVisibility) {
> - put(columnFamily, columnQualifier, columnVisibility.getExpression(),
> false, 0l, true, EMPTY_BYTES);
> + put(columnFamily.toString().getBytes(),
> columnQualifier.toString().getBytes(), columnVisibility.getExpression(),
> false, 0l, true, EMPTY_BYTES);
> }
>
> public void putDelete(CharSequence columnFamily, CharSequence
> columnQualifier, long timestamp) {
> - put(columnFamily, columnQualifier, EMPTY_BYTES, true, timestamp, true,
> EMPTY_BYTES);
> + put(columnFamily.toString().getBytes(),
> columnQualifier.toString().getBytes(), EMPTY_BYTES, true, timestamp, true,
> EMPTY_BYTES);
> }
>
> public void putDelete(CharSequence columnFamily, CharSequence
> columnQualifier, ColumnVisibility columnVisibility, long timestamp) {
> - put(columnFamily, columnQualifier, columnVisibility.getExpression(),
> true, timestamp, true, EMPTY_BYTES);
> + put(columnFamily.toString().getBytes(),
> columnQualifier.toString().getBytes(), columnVisibility.getExpression(),
> true, timestamp, true, EMPTY_BYTES);
> }
>
> public void put(CharSequence columnFamily, CharSequence columnQualifier,
> CharSequence value) {
> - put(columnFamily, columnQualifier, EMPTY_BYTES, false, 0l, false, value);
> + put(columnFamily.toString().getBytes(),
> columnQualifier.toString().getBytes(), EMPTY_BYTES, false, 0l, false,
> value.toString().getBytes());
> }
>
> public void put(CharSequence columnFamily, CharSequence columnQualifier,
> ColumnVisibility columnVisibility, CharSequence value) {
> - put(columnFamily, columnQualifier, columnVisibility.getExpression(),
> false, 0l, false, value);
> + put(columnFamily.toString().getBytes(),
> columnQualifier.toString().getBytes(), columnVisibility.getExpression(),
> false, 0l, false, value.toString().getBytes());
> }
>
> public void put(CharSequence columnFamily, CharSequence columnQualifier,
> long timestamp, CharSequence value) {
> - put(columnFamily, columnQualifier, EMPTY_BYTES, true, timestamp, false,
> value);
> + put(columnFamily.toString().getBytes(),
> columnQualifier.toString().getBytes(), EMPTY_BYTES, true, timestamp, false,
> value.toString().getBytes());
> }
>
> public void put(CharSequence columnFamily, CharSequence columnQualifier,
> ColumnVisibility columnVisibility, long timestamp, CharSequence value) {
> - put(columnFamily, columnQualifier, columnVisibility.getExpression(),
> true, timestamp, false, value);
> + put(columnFamily.toString().getBytes(),
> columnQualifier.toString().getBytes(), columnVisibility.getExpression(),
> true, timestamp, false, value.toString().getBytes());
> }
>
> private byte[] oldReadBytes(SimpleReader in) {
>
>
