List of usage examples for org.apache.hadoop.io Text set
public void set(byte[] utf8, int start, int len)
From source file:com.huayu.metis.flume.sink.hdfs.HDFSTextSerializer.java
License:Apache License
private Text makeText(Event e) { Text textObject = new Text(); textObject.set(e.getBody(), 0, e.getBody().length); return textObject; }
From source file:com.ibm.jaql.lang.expr.del.JsonToDel.java
License:Apache License
/** * Converts the given JSON value for a del line into a Text. * //from www . ja va 2 s. c o m * @param src JSON value * @return text text * @throws IOException * @throws IllegalArgumentException If no field names are provided for JSON * record. */ public void convert(JsonValue src, Text text) throws IOException { /* * 1. If quoted is false and src can be extract as a JSON string, its * internal UTF-8 byte array is set to text. * 2. Extract JSON values into an array list. Looping the JSON values in the * array list: * 2.1 quotes is false: If the value is a JSON string, its internal * UTF-8 byte array is written to out. Otherwise, it is serialized * to out. * 2.2 quoted is true: If the value does not contain characters which * need double-quote escaping or backslash escaping, it is * serialized to out. For JSON string, its internal UTF-8 byte array is * written to out byte-by-byte after escaping. For the value of other * types, it is serialized to a field buffer. Then the array backing the * field buffer is written to out byte-by-bye after escaping. * 2.3 Set the content of the buffer backing out to text. */ try { if (!quoted) { if (src instanceof JsonRecord && fieldNames.length == 1) { JsonRecord rec = (JsonRecord) src; JsonValue only = rec.get(fieldNames[0]); if (only instanceof JsonString) { JsonString js = (JsonString) only; text.set(js.getInternalBytes(), js.bytesOffset(), js.bytesLength()); return; } } else if (src instanceof JsonArray) { JsonArray arr = (JsonArray) src; if (arr.count() == 1) { JsonValue only = arr.get(0); if (only instanceof JsonString) { JsonString js = (JsonString) only; text.set(js.getInternalBytes(), js.bytesOffset(), js.bytesLength()); return; } } } else if (src instanceof JsonString) { JsonString js = (JsonString) src; text.set(js.getInternalBytes(), js.bytesOffset(), js.bytesLength()); return; } } } catch (Exception e) { throw new RuntimeException(e); } values.clear(); if (src instanceof JsonRecord) { JsonRecord rec = (JsonRecord) src; if (fieldNames.length < 1) throw new IllegalArgumentException("fields are required to convert A JSON record into a del line."); for (JsonString n : fieldNames) values.add(rec.get(n)); } else if (src instanceof JsonArray) { JsonArray arr = (JsonArray) src; for (JsonValue value : arr) values.add(value); } else { // If the value is not JsonRecord or JsonArray, then write it directly. values.add(src); } out.reset(); for (int i = 0; i < values.size(); i++) { JsonValue value = values.get(i); if (i != 0) out.write(delimiter); if (value == null) continue; if (quoted) printFieldQuoted(out, value, escape); else printFieldUnquoted(out, value); } out.flush(); text.set(out.toString()); }
From source file:com.kylinolap.job.hadoop.cube.CubeReducerTest.java
License:Apache License
private Text newValueText(MeasureCodec codec, String sum, String min, String max, int count) { Object[] values = new Object[] { new BigDecimal(sum), new BigDecimal(min), new BigDecimal(max), new LongWritable(count) }; buf.clear();/*from ww w . j av a2 s . c om*/ codec.encode(values, buf); Text t = new Text(); t.set(buf.array(), 0, buf.position()); return t; }
From source file:com.ricemap.spateDB.core.RTree.java
License:Apache License
/** * Builds the RTree given a serialized list of elements. It uses the given * stockObject to deserialize these elements and build the tree. Also writes * the created tree to the disk directly. * //from w w w .j av a2s .co m * @param elements * - serialization of elements to be written * @param offset * - index of the first element to use in the elements array * @param len * - number of bytes to user from the elements array * @param bytesAvailable * - size available (in bytes) to store the tree structures * @param dataOut * - an output to use for writing the tree to * @param fast_sort * - setting this to <code>true</code> allows the method to run * faster by materializing the offset of each element in the list * which speeds up the comparison. However, this requires an * additional 16 bytes per element. So, for each 1M elements, the * method will require an additional 16 M bytes (approximately). */ public void bulkLoadWrite(final byte[] element_bytes, final int offset, final int len, final int degree, DataOutput dataOut, final boolean fast_sort, final boolean columnarStorage) { try { columnar = columnarStorage; //TODO: the order of fields should be stable under Oracle JVM, but not guaranteed Field[] fields = stockObject.getClass().getDeclaredFields(); // Count number of elements in the given text int i_start = offset; final Text line = new Text(); while (i_start < offset + len) { int i_end = skipToEOL(element_bytes, i_start); // Extract the line without end of line character line.set(element_bytes, i_start, i_end - i_start - 1); stockObject.fromText(line); elementCount++; i_start = i_end; } LOG.info("Bulk loading an RTree with " + elementCount + " elements"); // It turns out the findBestDegree returns the best degree when the // whole // tree is loaded to memory when processed. However, as current // algorithms // process the tree while it's on disk, a higher degree should be // selected // such that a node fits one file block (assumed to be 4K). // final int degree = findBestDegree(bytesAvailable, elementCount); LOG.info("Writing an RTree with degree " + degree); int height = Math.max(1, (int) Math.ceil(Math.log(elementCount) / Math.log(degree))); int leafNodeCount = (int) Math.pow(degree, height - 1); if (elementCount < 2 * leafNodeCount && height > 1) { height--; leafNodeCount = (int) Math.pow(degree, height - 1); } int nodeCount = (int) ((Math.pow(degree, height) - 1) / (degree - 1)); int nonLeafNodeCount = nodeCount - leafNodeCount; // Keep track of the offset of each element in the text final int[] offsets = new int[elementCount]; final int[] ids = new int[elementCount]; final double[] ts = fast_sort ? new double[elementCount] : null; final double[] xs = fast_sort ? new double[elementCount] : null; final double[] ys = fast_sort ? new double[elementCount] : null; //initialize columnar data output ByteArrayOutputStream index_bos = new ByteArrayOutputStream(); DataOutputStream index_dos = new DataOutputStream(index_bos); ByteArrayOutputStream[] bos = new ByteArrayOutputStream[fields.length]; DataOutputStream[] dos = new DataOutputStream[fields.length]; for (int i = 0; i < bos.length; i++) { bos[i] = new ByteArrayOutputStream(); dos[i] = new DataOutputStream(bos[i]); } i_start = offset; line.clear(); for (int i = 0; i < elementCount; i++) { offsets[i] = i_start; ids[i] = i; int i_end = skipToEOL(element_bytes, i_start); if (xs != null) { // Extract the line with end of line character line.set(element_bytes, i_start, i_end - i_start - 1); stockObject.fromText(line); // Sample center of the shape ts[i] = (stockObject.getMBR().t1 + stockObject.getMBR().t2) / 2; xs[i] = (stockObject.getMBR().x1 + stockObject.getMBR().x2) / 2; ys[i] = (stockObject.getMBR().y1 + stockObject.getMBR().y2) / 2; //build columnar storage if (stockObject instanceof Point3d) { index_dos.writeDouble(ts[i]); index_dos.writeDouble(xs[i]); index_dos.writeDouble(ys[i]); } else { throw new RuntimeException("Indexing non-point shape with RTREE is not supported yet"); } for (int j = 0; j < fields.length; j++) { if (fields[j].getType().equals(Integer.TYPE)) { dos[j].writeInt(fields[j].getInt(stockObject)); } else if (fields[j].getType().equals(Double.TYPE)) { dos[j].writeDouble(fields[j].getDouble(stockObject)); } else if (fields[j].getType().equals(Long.TYPE)) { dos[j].writeLong(fields[j].getLong(stockObject)); } else { continue; //throw new RuntimeException("Field type is not supported yet"); } } } i_start = i_end; } index_dos.close(); for (int i = 0; i < dos.length; i++) { dos[i].close(); } /** A struct to store information about a split */ class SplitStruct extends Prism { /** Start and end index for this split */ int index1, index2; /** Direction of this split */ byte direction; /** Index of first element on disk */ int offsetOfFirstElement; static final byte DIRECTION_T = 0; static final byte DIRECTION_X = 1; static final byte DIRECTION_Y = 2; SplitStruct(int index1, int index2, byte direction) { this.index1 = index1; this.index2 = index2; this.direction = direction; } @Override public void write(DataOutput out) throws IOException { // if (columnarStorage) out.writeInt(index1); else out.writeInt(offsetOfFirstElement); super.write(out); } void partition(Queue<SplitStruct> toBePartitioned) { IndexedSortable sortableT; IndexedSortable sortableX; IndexedSortable sortableY; if (fast_sort) { // Use materialized xs[] and ys[] to do the comparisons sortableT = new IndexedSortable() { @Override public void swap(int i, int j) { // Swap ts double tempt = ts[i]; ts[i] = ts[j]; ts[j] = tempt; // Swap xs double tempx = xs[i]; xs[i] = xs[j]; xs[j] = tempx; // Swap ys double tempY = ys[i]; ys[i] = ys[j]; ys[j] = tempY; // Swap id int tempid = offsets[i]; offsets[i] = offsets[j]; offsets[j] = tempid; tempid = ids[i]; ids[i] = ids[j]; ids[j] = tempid; } @Override public int compare(int i, int j) { if (ts[i] < ts[j]) return -1; if (ts[i] > ts[j]) return 1; return 0; } }; sortableX = new IndexedSortable() { @Override public void swap(int i, int j) { // Swap ts double tempt = ts[i]; ts[i] = ts[j]; ts[j] = tempt; // Swap xs double tempx = xs[i]; xs[i] = xs[j]; xs[j] = tempx; // Swap ys double tempY = ys[i]; ys[i] = ys[j]; ys[j] = tempY; // Swap id int tempid = offsets[i]; offsets[i] = offsets[j]; offsets[j] = tempid; tempid = ids[i]; ids[i] = ids[j]; ids[j] = tempid; } @Override public int compare(int i, int j) { if (ts[i] < ts[j]) return -1; if (xs[i] < xs[j]) return -1; if (xs[i] > xs[j]) return 1; return 0; } }; sortableY = new IndexedSortable() { @Override public void swap(int i, int j) { // Swap ts double tempt = ts[i]; ts[i] = ts[j]; ts[j] = tempt; // Swap xs double tempx = xs[i]; xs[i] = xs[j]; xs[j] = tempx; // Swap ys double tempY = ys[i]; ys[i] = ys[j]; ys[j] = tempY; // Swap id int tempid = offsets[i]; offsets[i] = offsets[j]; offsets[j] = tempid; tempid = ids[i]; ids[i] = ids[j]; ids[j] = tempid; } @Override public int compare(int i, int j) { if (ys[i] < ys[j]) return -1; if (ys[i] > ys[j]) return 1; return 0; } }; } else { // No materialized xs and ys. Always deserialize objects // to compare sortableT = new IndexedSortable() { @Override public void swap(int i, int j) { // Swap id int tempid = offsets[i]; offsets[i] = offsets[j]; offsets[j] = tempid; tempid = ids[i]; ids[i] = ids[j]; ids[j] = tempid; } @Override public int compare(int i, int j) { // Get end of line int eol = skipToEOL(element_bytes, offsets[i]); line.set(element_bytes, offsets[i], eol - offsets[i] - 1); stockObject.fromText(line); double ti = (stockObject.getMBR().t1 + stockObject.getMBR().t2) / 2; eol = skipToEOL(element_bytes, offsets[j]); line.set(element_bytes, offsets[j], eol - offsets[j] - 1); stockObject.fromText(line); double tj = (stockObject.getMBR().t1 + stockObject.getMBR().t2) / 2; if (ti < tj) return -1; if (ti > tj) return 1; return 0; } }; sortableX = new IndexedSortable() { @Override public void swap(int i, int j) { // Swap id int tempid = offsets[i]; offsets[i] = offsets[j]; offsets[j] = tempid; tempid = ids[i]; ids[i] = ids[j]; ids[j] = tempid; } @Override public int compare(int i, int j) { // Get end of line int eol = skipToEOL(element_bytes, offsets[i]); line.set(element_bytes, offsets[i], eol - offsets[i] - 1); stockObject.fromText(line); double xi = (stockObject.getMBR().x1 + stockObject.getMBR().x2) / 2; eol = skipToEOL(element_bytes, offsets[j]); line.set(element_bytes, offsets[j], eol - offsets[j] - 1); stockObject.fromText(line); double xj = (stockObject.getMBR().x1 + stockObject.getMBR().x2) / 2; if (xi < xj) return -1; if (xi > xj) return 1; return 0; } }; sortableY = new IndexedSortable() { @Override public void swap(int i, int j) { // Swap id int tempid = offsets[i]; offsets[i] = offsets[j]; offsets[j] = tempid; tempid = ids[i]; ids[i] = ids[j]; ids[j] = tempid; } @Override public int compare(int i, int j) { int eol = skipToEOL(element_bytes, offsets[i]); line.set(element_bytes, offsets[i], eol - offsets[i] - 1); stockObject.fromText(line); double yi = (stockObject.getMBR().y1 + stockObject.getMBR().y2) / 2; eol = skipToEOL(element_bytes, offsets[j]); line.set(element_bytes, offsets[j], eol - offsets[j] - 1); stockObject.fromText(line); double yj = (stockObject.getMBR().y1 + stockObject.getMBR().y2) / 2; if (yi < yj) return -1; if (yi > yj) return 1; return 0; } }; } final IndexedSorter sorter = new QuickSort(); final IndexedSortable[] sortables = new IndexedSortable[3]; sortables[SplitStruct.DIRECTION_T] = sortableT; sortables[SplitStruct.DIRECTION_X] = sortableX; sortables[SplitStruct.DIRECTION_Y] = sortableY; sorter.sort(sortables[direction], index1, index2); // Partition into maxEntries partitions (equally) and // create a SplitStruct for each partition int i1 = index1; for (int iSplit = 0; iSplit < degree; iSplit++) { int i2 = index1 + (index2 - index1) * (iSplit + 1) / degree; SplitStruct newSplit; if (direction == 0) { newSplit = new SplitStruct(i1, i2, (byte) 1); } else if (direction == 1) { newSplit = new SplitStruct(i1, i2, (byte) 2); } else { newSplit = new SplitStruct(i1, i2, (byte) 0); } toBePartitioned.add(newSplit); i1 = i2; } } } // All nodes stored in level-order traversal Vector<SplitStruct> nodes = new Vector<SplitStruct>(); final Queue<SplitStruct> toBePartitioned = new LinkedList<SplitStruct>(); toBePartitioned.add(new SplitStruct(0, elementCount, SplitStruct.DIRECTION_X)); while (!toBePartitioned.isEmpty()) { SplitStruct split = toBePartitioned.poll(); if (nodes.size() < nonLeafNodeCount) { // This is a non-leaf split.partition(toBePartitioned); } nodes.add(split); } if (nodes.size() != nodeCount) { throw new RuntimeException( "Expected node count: " + nodeCount + ". Real node count: " + nodes.size()); } // Now we have our data sorted in the required order. Start building // the tree. // Store the offset of each leaf node in the tree FSDataOutputStream fakeOut = new FSDataOutputStream(new java.io.OutputStream() { // Null output stream @Override public void write(int b) throws IOException { // Do nothing } @Override public void write(byte[] b, int off, int len) throws IOException { // Do nothing } @Override public void write(byte[] b) throws IOException { // Do nothing } }, null, TreeHeaderSize + nodes.size() * NodeSize); for (int i_leaf = nonLeafNodeCount, i = 0; i_leaf < nodes.size(); i_leaf++) { nodes.elementAt(i_leaf).offsetOfFirstElement = (int) fakeOut.getPos(); if (i != nodes.elementAt(i_leaf).index1) throw new RuntimeException(); double t1, x1, y1, t2, x2, y2; // Initialize MBR to first object int eol = skipToEOL(element_bytes, offsets[i]); fakeOut.write(element_bytes, offsets[i], eol - offsets[i]); line.set(element_bytes, offsets[i], eol - offsets[i] - 1); stockObject.fromText(line); Prism mbr = stockObject.getMBR(); t1 = mbr.t1; x1 = mbr.x1; y1 = mbr.y1; t2 = mbr.t2; x2 = mbr.x2; y2 = mbr.y2; i++; while (i < nodes.elementAt(i_leaf).index2) { eol = skipToEOL(element_bytes, offsets[i]); fakeOut.write(element_bytes, offsets[i], eol - offsets[i]); line.set(element_bytes, offsets[i], eol - offsets[i] - 1); stockObject.fromText(line); mbr = stockObject.getMBR(); if (mbr.t1 < t1) t1 = mbr.t1; if (mbr.x1 < x1) x1 = mbr.x1; if (mbr.y1 < y1) y1 = mbr.y1; if (mbr.t2 > t2) t2 = mbr.t2; if (mbr.x2 > x2) x2 = mbr.x2; if (mbr.y2 > y2) y2 = mbr.y2; i++; } nodes.elementAt(i_leaf).set(t1, x1, y1, t2, x2, y2); } fakeOut.close(); fakeOut = null; // Calculate MBR and offsetOfFirstElement for non-leaves for (int i_node = nonLeafNodeCount - 1; i_node >= 0; i_node--) { int i_first_child = i_node * degree + 1; nodes.elementAt(i_node).offsetOfFirstElement = nodes.elementAt(i_first_child).offsetOfFirstElement; int i_child = 0; Prism mbr; mbr = nodes.elementAt(i_first_child + i_child); double t1 = mbr.t1; double x1 = mbr.x1; double y1 = mbr.y1; double t2 = mbr.t2; double x2 = mbr.x2; double y2 = mbr.y2; i_child++; while (i_child < degree) { mbr = nodes.elementAt(i_first_child + i_child); if (mbr.t1 < t1) t1 = mbr.t1; if (mbr.x1 < x1) x1 = mbr.x1; if (mbr.y1 < y1) y1 = mbr.y1; if (mbr.t2 > t2) t2 = mbr.t2; if (mbr.x2 > x2) x2 = mbr.x2; if (mbr.y2 > y2) y2 = mbr.y2; i_child++; } nodes.elementAt(i_node).set(t1, x1, y1, t2, x2, y2); } // Start writing the tree // write tree header (including size) // Total tree size. (== Total bytes written - 8 bytes for the size // itself) dataOut.writeInt(TreeHeaderSize + NodeSize * nodeCount + len); // Tree height dataOut.writeInt(height); // Degree dataOut.writeInt(degree); dataOut.writeInt(elementCount); //isColumnar dataOut.writeInt(columnarStorage ? 1 : 0); // write nodes for (SplitStruct node : nodes) { node.write(dataOut); } // write elements if (columnarStorage) { byte[] index_bs = index_bos.toByteArray(); byte[][] bss = new byte[bos.length][]; for (int i = 0; i < bss.length; i++) { bss[i] = bos[i].toByteArray(); } for (int element_i = 0; element_i < elementCount; element_i++) { //int eol = skipToEOL(element_bytes, offsets[element_i]); //dataOut.write(element_bytes, offsets[element_i], eol - offsets[element_i]); dataOut.write(index_bs, ids[element_i] * IndexUnitSize, IndexUnitSize); } for (int i = 0; i < fields.length; i++) { int fieldSize = 0; if (fields[i].getType().equals(Integer.TYPE)) { fieldSize = 4; } else if (fields[i].getType().equals(Long.TYPE)) { fieldSize = 8; } else if (fields[i].getType().equals(Double.TYPE)) { fieldSize = 8; } else { //throw new RuntimeException("Unsupported field type: " + fields[i].getType().getName()); continue; } for (int element_i = 0; element_i < elementCount; element_i++) { //int eol = skipToEOL(element_bytes, offsets[element_i]); //dataOut.write(element_bytes, offsets[element_i], eol - offsets[element_i]); dataOut.write(bss[i], ids[element_i] * fieldSize, fieldSize); } } } else { for (int element_i = 0; element_i < elementCount; element_i++) { int eol = skipToEOL(element_bytes, offsets[element_i]); dataOut.write(element_bytes, offsets[element_i], eol - offsets[element_i]); } } } catch (IOException e) { e.printStackTrace(); } catch (IllegalArgumentException e) { // TODO Auto-generated catch block e.printStackTrace(); } catch (IllegalAccessException e) { // TODO Auto-generated catch block e.printStackTrace(); } }
From source file:com.ricemap.spateDB.io.TextSerializerHelper.java
License:Apache License
/** * Deserializes and consumes a long from the given text. Consuming means all * characters read for deserialization are removed from the given text. * If separator is non-zero, a long is read and consumed up to the first * occurrence of this separator. The separator is also consumed. * @param text/*from w ww. j a v a 2s . c o m*/ * @param separator * @return */ public static long consumeHexLong(Text text, char separator) { int i = 0; byte[] bytes = text.getBytes(); // Skip until the separator or end of text while (i < text.getLength() && HexadecimalChars[bytes[i]]) i++; long l = deserializeHexLong(bytes, 0, i); // If the first char after the long is the separator, skip it if (i < text.getLength() && bytes[i] == separator) i++; // Shift bytes after the long System.arraycopy(bytes, i, bytes, 0, text.getLength() - i); text.set(bytes, 0, text.getLength() - i); return l; }
From source file:com.ricemap.spateDB.io.TextSerializerHelper.java
License:Apache License
/** * Deserializes and consumes a double from the given text. Consuming means all * characters read for deserialization are removed from the given text. * If separator is non-zero, a double is read and consumed up to the first * occurrence of this separator. The separator is also consumed. * @param text/*from w ww .j a va2 s . c om*/ * @param separator * @return */ public static double consumeDouble(Text text, char separator) { int i = 0; byte[] bytes = text.getBytes(); // Skip until the separator or end of text while (i < text.getLength() && ((bytes[i] >= '0' && bytes[i] <= '9') || bytes[i] == 'e' || bytes[i] == 'E' || bytes[i] == '-' || bytes[i] == '+' || bytes[i] == '.')) i++; double d = deserializeDouble(bytes, 0, i); if (i < text.getLength() && bytes[i] == separator) i++; System.arraycopy(bytes, i, bytes, 0, text.getLength() - i); text.set(bytes, 0, text.getLength() - i); return d; }
From source file:com.ricemap.spateDB.io.TextSerializerHelper.java
License:Apache License
public static long consumeLong(Text text, char separator) { int i = 0;//from ww w .j a v a2 s .c om byte[] bytes = text.getBytes(); // Skip until the separator or end of text while (i < text.getLength() && DecimalChars[bytes[i]]) i++; long l = deserializeLong(bytes, 0, i); // If the first char after the long is the separator, skip it if (i < text.getLength() && bytes[i] == separator) i++; // Shift bytes after the long System.arraycopy(bytes, i, bytes, 0, text.getLength() - i); text.set(bytes, 0, text.getLength() - i); return l; }
From source file:com.ricemap.spateDB.io.TextSerializerHelper.java
License:Apache License
public static int consumeInt(Text text, char separator) { int i = 0;//from w w w.j a v a 2s .c om byte[] bytes = text.getBytes(); // Skip until the separator or end of text while (i < text.getLength() && DecimalChars[bytes[i]]) i++; int l = deserializeInt(bytes, 0, i); // If the first char after the long is the separator, skip it if (i < text.getLength() && bytes[i] == separator) i++; // Shift bytes after the long System.arraycopy(bytes, i, bytes, 0, text.getLength() - i); text.set(bytes, 0, text.getLength() - i); return l; }
From source file:com.ricemap.spateDB.io.TextSerializerHelper.java
License:Apache License
public static void consumeMap(Text text, Map<String, String> tags) { tags.clear();// w ww .jav a2s . c o m if (text.getLength() > 0) { byte[] tagsBytes = text.getBytes(); if (tagsBytes[0] != Separators[MapStart]) return; int i1 = 1; while (i1 < text.getLength() && tagsBytes[i1] != Separators[MapEnd]) { int i2 = i1 + 1; while (i2 < text.getLength() && tagsBytes[i2] != Separators[KeyValueSeparator]) i2++; String key = new String(tagsBytes, i1, i2 - i1); i1 = i2 + 1; i2 = i1 + 1; while (i2 < text.getLength() && tagsBytes[i2] != Separators[FieldSeparator] && tagsBytes[i2] != Separators[MapEnd]) i2++; String value = new String(tagsBytes, i1, i2 - i1); tags.put(key, value); i1 = i2; if (i1 < text.getLength() && tagsBytes[i1] == Separators[FieldSeparator]) i1++; } text.set(tagsBytes, i1, text.getLength() - i1); } }
From source file:de.l3s.streamcorpus.terrier.ThriftFileCollectionRecordReader.java
License:Apache License
/** * Reading a bunch of lines of file paths in a list. * The code in this method is redistributed from Hadoop LineRecordReader * // www .ja va 2 s. c o m * @throws IOException */ private void loadPathsFromInputSplit(InputSplit split, Configuration conf) throws IOException { FileSplit fileSplit = (FileSplit) split; Path path = fileSplit.getPath(); long begin = fileSplit.getStart(); long end = begin + fileSplit.getLength(); LOG.info("Reading paths in file " + path.getName()); // First check the compression codec CompressionCodecFactory compressionCodec = new CompressionCodecFactory(conf); CompressionCodec codec = compressionCodec.getCodec(path); FSDataInputStream fis = fs.open(path); SplitLineReader in; Seekable filePosition; boolean compressed = false; Decompressor decompressor = null; if (null != codec) { compressed = true; decompressor = CodecPool.getDecompressor(codec); if (codec instanceof SplittableCompressionCodec) { final SplitCompressionInputStream cIn = ((SplittableCompressionCodec) codec).createInputStream(fis, decompressor, begin, end, SplittableCompressionCodec.READ_MODE.BYBLOCK); in = new CompressedSplitLineReader(cIn, conf, (byte[]) null); begin = cIn.getAdjustedStart(); end = cIn.getAdjustedEnd(); filePosition = cIn; } else { in = new SplitLineReader(codec.createInputStream(fis, decompressor), conf, null); filePosition = fis; } } else { fis.seek(begin); in = new SplitLineReader(fis, conf, (byte[]) null); filePosition = fis; } // If this is not the first split, we always throw away first record // because we always (except the last split) read one extra line in // next() method. if (begin != 0) { begin += in.readLine(new Text(), 0, maxBytesToConsume(compressed, begin, end)); } long pos = begin; int newSize = 0; final Text nextLine = new Text(); paths = new ArrayList<>(); while (getFilePosition(compressed, filePosition, pos) <= end || in.needAdditionalRecordAfterSplit()) { if (pos == 0) { // Strip BOM(Byte Order Mark) // Text only support UTF-8, we only need to check UTF-8 BOM // (0xEF,0xBB,0xBF) at the start of the text stream. newSize = in.readLine(nextLine, Integer.MAX_VALUE, Integer.MAX_VALUE); pos += newSize; int textLength = nextLine.getLength(); byte[] textBytes = nextLine.getBytes(); if ((textLength >= 3) && (textBytes[0] == (byte) 0xEF) && (textBytes[1] == (byte) 0xBB) && (textBytes[2] == (byte) 0xBF)) { // find UTF-8 BOM, strip it. LOG.info("Found UTF-8 BOM and skipped it"); textLength -= 3; newSize -= 3; if (textLength > 0) { // It may work to use the same buffer and // not do the copyBytes textBytes = nextLine.copyBytes(); nextLine.set(textBytes, 3, textLength); } else { nextLine.clear(); } } } else { newSize = in.readLine(nextLine, Integer.MAX_VALUE, maxBytesToConsume(compressed, pos, end)); pos += newSize; } paths.add(nextLine.toString()); // line too long. try again LOG.info("Skipped line of size " + newSize + " at pos " + (pos - newSize)); } try { if (in != null) { in.close(); } if (fis != null) { fis.close(); } } finally { if (decompressor != null) { CodecPool.returnDecompressor(decompressor); } } }