Merge pull request #57 from jamesmudd/string-dataset-fix

String dataset fix

README.md

@@ -1,7 +1,7 @@
 # jHDF - Pure Java HDF5 library
 [![Build Status](https://dev.azure.com/jamesmudd/jhdf/_apis/build/status/jhdf-CI)](https://dev.azure.com/jamesmudd/jhdf/_build/latest?definitionId=3) [![Quality Gate Status](https://sonarcloud.io/api/project_badges/measure?project=jamesmudd_jhdf&metric=alert_status)](https://sonarcloud.io/dashboard?id=jamesmudd_jhdf) [![Coverage](https://sonarcloud.io/api/project_badges/measure?project=jamesmudd_jhdf&metric=coverage)](https://sonarcloud.io/dashboard?id=jamesmudd_jhdf) [ ![Download](https://api.bintray.com/packages/jamesmudd/jhdf/jhdf/images/download.svg) ](https://bintray.com/jamesmudd/jhdf/jhdf/_latestVersion) [![Javadocs](http://javadoc.io/badge/io.jhdf/jhdf.svg)](http://javadoc.io/doc/io.jhdf/jhdf)
 
-This is a hobby project for me to see if I can write a loader for HDF5 files. The file format specification is available from the HDF Group [here](https://support.hdfgroup.org/HDF5/doc/H5.format.html).
+This is a hobby project for me to see if I can write a loader for HDF5 files. The file format specification is available from the HDF Group [here](https://support.hdfgroup.org/HDF5/doc/H5.format.html). More information on the format is available on [Wikipedia](https://en.wikipedia.org/wiki/Hierarchical_Data_Format)
 
 The intension is to make a clean Java API to access HDF5 data. Currently the project is targeting HDF5 1.8 read-only compatibility. For progress see the [change log](CHANGES.md)
 

jhdf/src/main/java/io/jhdf/ObjectHeader.java

@@ -273,7 +273,7 @@ private ObjectHeaderV2(FileChannel fc, Superblock sb, long address) {
 		}
 
 		private ByteBuffer readMessages(FileChannel fc, Superblock sb, ByteBuffer bb) throws IOException {
-			while (bb.remaining() > 4) {
+			while (bb.remaining() >= 8) {
 				Message m = Message.readObjectHeaderV2Message(bb, sb);
 				messages.add(m);
 

jhdf/src/main/java/io/jhdf/dataset/DatasetReader.java

@@ -1,5 +1,7 @@
 package io.jhdf.dataset;
 
+import static java.nio.charset.StandardCharsets.US_ASCII;
+
 import java.lang.reflect.Array;
 import java.math.BigInteger;
 import java.nio.ByteBuffer;
@@ -9,7 +11,6 @@
 import java.nio.IntBuffer;
 import java.nio.LongBuffer;
 import java.nio.ShortBuffer;
-import java.nio.charset.Charset;
 import java.util.Arrays;
 
 import io.jhdf.Utils;
@@ -18,7 +19,6 @@
 import io.jhdf.object.datatype.FixedPoint;
 import io.jhdf.object.datatype.FloatingPoint;
 import io.jhdf.object.datatype.StringData;
-import io.jhdf.object.datatype.VariableLength;
 
 /**
  * <p>
@@ -108,11 +108,6 @@ private DatasetReader() {
 			StringData stringData = (StringData) type;
 			int stringLength = stringData.getSize();
 			fillFixedLentghStringData(data, dimensions, buffer, stringLength);
-		} else if (type instanceof VariableLength) {
-			VariableLength variableLentgh = (VariableLength) type;
-			Charset encoding = variableLentgh.getEncoding();
-			int stringLength = variableLentgh.getSize();
-			fillVariableLentghStringData(data, dimensions, buffer, stringLength, encoding);
 		}
 
 		return data;
@@ -270,24 +265,8 @@ private static void fillFixedLentghStringData(Object data, int[] dims, ByteBuffe
 		} else {
 			for (int i = 0; i < dims[0]; i++) {
 				buffer.position(i * stringLength);
-				Array.set(data, i, Utils.readUntilNull(buffer));
-			}
-		}
-	}
-
-	private static void fillVariableLentghStringData(Object data, int[] dims, ByteBuffer buffer, int stringLength,
-			Charset charset) {
-		if (dims.length > 1) {
-			for (int i = 0; i < dims[0]; i++) {
-				Object newArray = Array.get(data, i);
-				fillData(newArray, stripLeadingIndex(dims), buffer);
-			}
-		} else {
-			byte[] elementBuffer = new byte[stringLength];
-			for (int i = 0; i < dims[0]; i++) {
-				buffer.position(i * stringLength);
-				buffer.get(elementBuffer);
-				Array.set(data, i, charset.decode(ByteBuffer.wrap(elementBuffer)).toString());
+				ByteBuffer elementBuffer = Utils.createSubBuffer(buffer, stringLength);
+				Array.set(data, i, US_ASCII.decode(elementBuffer).toString().trim());
 			}
 		}
 	}

jhdf/src/test/java/io/jhdf/dataset/StringDatasetTest.java

@@ -44,12 +44,16 @@ Collection<DynamicNode> stringDataset1DTests() {
 		return Arrays.asList(
 				dynamicTest("earliest - fixed ASCII",
 						createTest(earliestHdfFile, "/fixed_lentgh_ascii")),
+				dynamicTest("earliest - fixed ASCII 1 char",
+						createTest(earliestHdfFile, "/fixed_lentgh_ascii_1_char")),
 				dynamicTest("earliest - variable ASCII",
 						createTest(earliestHdfFile, "/variable_lentgh_ascii")),
 				dynamicTest("earliest - variable UTF8",
 						createTest(earliestHdfFile, "/variable_lentgh_utf8")),
 				dynamicTest("latest - fixed ASCII",
 						createTest(latestHdfFile, "/fixed_lentgh_ascii")),
+				dynamicTest("latest - fixed ASCII 1 char",
+						createTest(latestHdfFile, "/fixed_lentgh_ascii_1_char")),
 				dynamicTest("latest - variable ASCII",
 						createTest(latestHdfFile, "/variable_lentgh_ascii")),
 				dynamicTest("latest - variable UTF8",

jhdf/src/test/resources/io/jhdf/string_datasets.py

@@ -12,6 +12,12 @@ def write_string_datasets(f):
     for i in range(10):
         fixed_ds[i] = ('string number ' + str(i)).encode('ascii')
 
+    # Fixed length (15) ASCII dataset the exact lentgh of 'string number 0'
+    fixed_length = 'S15'
+    fixed_ds = f.create_dataset('fixed_lentgh_ascii_1_char', (10,), dtype=fixed_length)
+    for i in range(10):
+        fixed_ds[i] = ('string number ' + str(i)).encode('ascii')
+
     # Variable length ASCII dataset
     ascii = h5py.special_dtype(vlen=bytes)
     varaible_ascii_ds = f.create_dataset('variable_lentgh_ascii', (10,), dtype=ascii)