LUCENE-3707: Add a Lucene3x private SegmentInfosFormat implementation

git-svn-id: https://svn.apache.org/repos/asf/lucene/dev/trunk@1233380 13f79535-47bb-0310-9956-ffa450edef68

lucene/src/java/org/apache/lucene/codecs/lucene3x/Lucene3xCodec.java

@@ -30,9 +30,7 @@ import org.apache.lucene.codecs.PostingsFormat;
 import org.apache.lucene.codecs.SegmentInfosFormat;
 import org.apache.lucene.codecs.StoredFieldsFormat;
 import org.apache.lucene.codecs.TermVectorsFormat;
-import org.apache.lucene.codecs.lucene40.Lucene40SegmentInfosFormat;
 import org.apache.lucene.codecs.lucene40.Lucene40StoredFieldsFormat;
-import org.apache.lucene.codecs.lucene40.Lucene40TermVectorsFormat;
 import org.apache.lucene.index.PerDocWriteState;
 import org.apache.lucene.index.SegmentInfo;
 import org.apache.lucene.index.SegmentReadState;
@@ -55,12 +53,8 @@ public class Lucene3xCodec extends Codec {
   
   private final FieldInfosFormat fieldInfosFormat = new Lucene3xFieldInfosFormat();
 
-  // TODO: this should really be a different impl
-  // also if we want preflex to *really* be read-only it should throw exception for the writer?
-  // this way IR.commit fails on delete/undelete/setNorm/etc ?
-  private final SegmentInfosFormat infosFormat = new Lucene40SegmentInfosFormat();
+  private final SegmentInfosFormat infosFormat = new Lucene3xSegmentInfosFormat();
   
-  // TODO: this should really be a different impl
   private final NormsFormat normsFormat = new Lucene3xNormsFormat();
   
   // 3.x doesn't support docvalues

lucene/src/java/org/apache/lucene/codecs/lucene3x/Lucene3xFieldInfosFormat.java

@@ -27,8 +27,12 @@ import org.apache.lucene.index.SegmentInfo;
 import org.apache.lucene.store.Directory;
 
 /**
+ * Lucene3x ReadOnly FieldInfosFromat implementation
+ * @deprecated (4.0) This is only used to read indexes created
+ * before 4.0.
  * @lucene.experimental
  */
+@Deprecated
 public class Lucene3xFieldInfosFormat extends FieldInfosFormat {
   private final FieldInfosReader reader = new Lucene3xFieldInfosReader();
   
@@ -39,7 +43,7 @@ public class Lucene3xFieldInfosFormat extends FieldInfosFormat {
 
   @Override
   public FieldInfosWriter getFieldInfosWriter() throws IOException {
-    throw new IllegalArgumentException("this codec can only be used for reading");
+    throw new UnsupportedOperationException("this codec can only be used for reading");
   }
 
   @Override

lucene/src/java/org/apache/lucene/codecs/lucene3x/Lucene3xNormsFormat.java

@@ -29,10 +29,12 @@ import org.apache.lucene.index.SegmentReadState;
 import org.apache.lucene.store.Directory;
 
 /**
- * Read-Only Lucene 3.x Norms Format
- * 
+ * Lucene3x ReadOnly NormsFormat implementation
+ * @deprecated (4.0) This is only used to read indexes created
+ * before 4.0.
  * @lucene.experimental
  */
+@Deprecated
 public class Lucene3xNormsFormat extends NormsFormat {
 
 
@@ -49,7 +51,7 @@ public class Lucene3xNormsFormat extends NormsFormat {
 
   @Override
   public PerDocConsumer docsConsumer(PerDocWriteState state) throws IOException {
-    throw new IllegalArgumentException("this codec can only be used for reading");
+    throw new UnsupportedOperationException("this codec can only be used for reading");
   }
 
   @Override

lucene/src/java/org/apache/lucene/codecs/lucene3x/Lucene3xPostingsFormat.java

@@ -57,7 +57,7 @@ public class Lucene3xPostingsFormat extends PostingsFormat {
   
   @Override
   public FieldsConsumer fieldsConsumer(SegmentWriteState state) throws IOException {
-    throw new IllegalArgumentException("this codec can only be used for reading");
+    throw new UnsupportedOperationException("this codec can only be used for reading");
   }
 
   @Override

lucene/src/java/org/apache/lucene/codecs/lucene3x/Lucene3xSegmentInfosFormat.java

@@ -0,0 +1,43 @@
+package org.apache.lucene.codecs.lucene3x;
+
+import org.apache.lucene.codecs.SegmentInfosFormat;
+import org.apache.lucene.codecs.SegmentInfosReader;
+import org.apache.lucene.codecs.SegmentInfosWriter;
+
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+/**
+ * Lucene3x ReadOnly SegmentInfosFormat implementation
+ * @deprecated (4.0) This is only used to read indexes created
+ * before 4.0.
+ * @lucene.experimental
+ */
+@Deprecated
+public class Lucene3xSegmentInfosFormat extends SegmentInfosFormat {
+  private final SegmentInfosReader reader = new Lucene3xSegmentInfosReader();
+  
+  @Override
+  public SegmentInfosReader getSegmentInfosReader() {
+    return reader;
+  }
+
+  @Override
+  public SegmentInfosWriter getSegmentInfosWriter() {
+    throw new UnsupportedOperationException("this codec can only be used for reading");
+  }
+}

lucene/src/java/org/apache/lucene/codecs/lucene3x/Lucene3xSegmentInfosReader.java

@@ -0,0 +1,170 @@
+package org.apache.lucene.codecs.lucene3x;
+
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+import java.io.IOException;
+import java.util.HashMap;
+import java.util.Map;
+
+import org.apache.lucene.codecs.Codec;
+import org.apache.lucene.codecs.SegmentInfosReader;
+// TODO we need a 3x version of stored fields reader
+import org.apache.lucene.codecs.lucene40.Lucene40StoredFieldsReader;
+import org.apache.lucene.index.IndexFileNames;
+import org.apache.lucene.index.IndexFormatTooOldException;
+import org.apache.lucene.index.SegmentInfo;
+import org.apache.lucene.index.SegmentInfos;
+import org.apache.lucene.store.ChecksumIndexInput;
+import org.apache.lucene.store.CompoundFileDirectory;
+import org.apache.lucene.store.Directory;
+import org.apache.lucene.store.IOContext;
+
+/**
+ * Lucene 3x implementation of {@link SegmentInfosReader}.
+ * @lucene.experimental
+ */
+public class Lucene3xSegmentInfosReader extends SegmentInfosReader {
+
+  @Override
+  public void read(Directory directory, String segmentsFileName, ChecksumIndexInput input, SegmentInfos infos, IOContext context) throws IOException { 
+    infos.version = input.readLong(); // read version
+    infos.counter = input.readInt(); // read counter
+    final int format = infos.getFormat();
+    for (int i = input.readInt(); i > 0; i--) { // read segmentInfos
+      SegmentInfo si = readSegmentInfo(directory, format, input);
+      if (si.getVersion() == null) {
+        // Could be a 3.0 - try to open the doc stores - if it fails, it's a
+        // 2.x segment, and an IndexFormatTooOldException will be thrown,
+        // which is what we want.
+        Directory dir = directory;
+        if (si.getDocStoreOffset() != -1) {
+          if (si.getDocStoreIsCompoundFile()) {
+            dir = new CompoundFileDirectory(dir, IndexFileNames.segmentFileName(
+                si.getDocStoreSegment(), "",
+                IndexFileNames.COMPOUND_FILE_STORE_EXTENSION), context, false);
+          }
+        } else if (si.getUseCompoundFile()) {
+          dir = new CompoundFileDirectory(dir, IndexFileNames.segmentFileName(
+              si.name, "", IndexFileNames.COMPOUND_FILE_EXTENSION), context, false);
+        }
+
+        try {
+          Lucene40StoredFieldsReader.checkCodeVersion(dir, si.getDocStoreSegment());
+        } finally {
+          // If we opened the directory, close it
+          if (dir != directory) dir.close();
+        }
+          
+        // Above call succeeded, so it's a 3.0 segment. Upgrade it so the next
+        // time the segment is read, its version won't be null and we won't
+        // need to open FieldsReader every time for each such segment.
+        si.setVersion("3.0");
+      } else if (si.getVersion().equals("2.x")) {
+        // If it's a 3x index touched by 3.1+ code, then segments record their
+        // version, whether they are 2.x ones or not. We detect that and throw
+        // appropriate exception.
+        throw new IndexFormatTooOldException("segment " + si.name + " in resource " + input, si.getVersion());
+      }
+      infos.add(si);
+    }
+      
+    infos.userData = input.readStringStringMap();
+  }
+  
+  // if we make a preflex impl we can remove a lot of this hair...
+  public SegmentInfo readSegmentInfo(Directory dir, int format, ChecksumIndexInput input) throws IOException {
+    final String version;
+    if (format <= SegmentInfos.FORMAT_3_1) {
+      version = input.readString();
+    } else {
+      version = null;
+    }
+    final String name = input.readString();
+    final int docCount = input.readInt();
+    final long delGen = input.readLong();
+    final int docStoreOffset = input.readInt();
+    final String docStoreSegment;
+    final boolean docStoreIsCompoundFile;
+    if (docStoreOffset != -1) {
+      docStoreSegment = input.readString();
+      docStoreIsCompoundFile = input.readByte() == SegmentInfo.YES;
+    } else {
+      docStoreSegment = name;
+      docStoreIsCompoundFile = false;
+    }
+
+    // pre-4.0 indexes write a byte if there is a single norms file
+    byte b = input.readByte();
+    assert 1 == b : "expected 1 but was: "+ b + " format: " + format;
+
+    final int numNormGen = input.readInt();
+    final Map<Integer,Long> normGen;
+    if (numNormGen == SegmentInfo.NO) {
+      normGen = null;
+    } else {
+      normGen = new HashMap<Integer, Long>();
+      for(int j=0;j<numNormGen;j++) {
+        normGen.put(j, input.readLong());
+      }
+    }
+    final boolean isCompoundFile = input.readByte() == SegmentInfo.YES;
+
+    final int delCount = input.readInt();
+    assert delCount <= docCount;
+
+    final int hasProx = input.readByte();
+
+    final Codec codec = Codec.forName("Lucene3x");
+    final Map<String,String> diagnostics = input.readStringStringMap();
+
+    final int hasVectors;
+    if (format <= SegmentInfos.FORMAT_HAS_VECTORS) {
+      hasVectors = input.readByte();
+    } else {
+      final String storesSegment;
+      final String ext;
+      final boolean storeIsCompoundFile;
+      if (docStoreOffset != -1) {
+        storesSegment = docStoreSegment;
+        storeIsCompoundFile = docStoreIsCompoundFile;
+        ext = IndexFileNames.COMPOUND_FILE_STORE_EXTENSION;
+      } else {
+        storesSegment = name;
+        storeIsCompoundFile = isCompoundFile;
+        ext = IndexFileNames.COMPOUND_FILE_EXTENSION;
+      }
+      final Directory dirToTest;
+      if (storeIsCompoundFile) {
+        dirToTest = new CompoundFileDirectory(dir, IndexFileNames.segmentFileName(storesSegment, "", ext), IOContext.READONCE, false);
+      } else {
+        dirToTest = dir;
+      }
+      try {
+        hasVectors = dirToTest.fileExists(IndexFileNames.segmentFileName(storesSegment, "", Lucene3xTermVectorsReader.VECTORS_INDEX_EXTENSION)) ? SegmentInfo.YES : SegmentInfo.NO;
+      } finally {
+        if (isCompoundFile) {
+          dirToTest.close();
+        }
+      }
+    }
+    
+    return new SegmentInfo(dir, version, name, docCount, delGen, docStoreOffset,
+      docStoreSegment, docStoreIsCompoundFile, normGen, isCompoundFile,
+      delCount, hasProx, codec, diagnostics, hasVectors);
+  }
+}

lucene/src/java/org/apache/lucene/codecs/lucene3x/Lucene3xTermVectorsFormat.java

@@ -28,6 +28,13 @@ import org.apache.lucene.index.SegmentInfo;
 import org.apache.lucene.store.Directory;
 import org.apache.lucene.store.IOContext;
 
+/**
+ * Lucene3x ReadOnly TermVectorsFormat implementation
+ * @deprecated (4.0) This is only used to read indexes created
+ * before 4.0.
+ * @lucene.experimental
+ */
+@Deprecated
 public class Lucene3xTermVectorsFormat extends TermVectorsFormat {
 
   @Override
@@ -37,8 +44,7 @@ public class Lucene3xTermVectorsFormat extends TermVectorsFormat {
 
   @Override
   public TermVectorsWriter vectorsWriter(Directory directory, String segment, IOContext context) throws IOException {
-    // TODO all these IAEs in preflex should be UOEs?
-    throw new IllegalArgumentException("this codec can only be used for reading");
+    throw new UnsupportedOperationException("this codec can only be used for reading");
   }
 
   @Override

lucene/src/java/org/apache/lucene/codecs/lucene40/Lucene40SegmentInfosReader.java

@@ -23,13 +23,10 @@ import java.util.Map;
 
 import org.apache.lucene.codecs.Codec;
 import org.apache.lucene.codecs.SegmentInfosReader;
-import org.apache.lucene.codecs.lucene40.Lucene40TermVectorsReader;
-import org.apache.lucene.index.IndexFileNames;
 import org.apache.lucene.index.IndexFormatTooOldException;
 import org.apache.lucene.index.SegmentInfo;
 import org.apache.lucene.index.SegmentInfos;
 import org.apache.lucene.store.ChecksumIndexInput;
-import org.apache.lucene.store.CompoundFileDirectory;
 import org.apache.lucene.store.Directory;
 import org.apache.lucene.store.IOContext;
 
@@ -39,88 +36,37 @@ import org.apache.lucene.store.IOContext;
  */
 public class Lucene40SegmentInfosReader extends SegmentInfosReader {
 
-  // TODO: shove all backwards code to preflex!
-  // this is a little tricky, because of IR.commit(), two options:
-  // 1. PreFlex writes 4.x SIS format, but reads both 3.x and 4.x
-  //    (and maybe RW always only writes the 3.x one? for that to work well,
-  //     we have to move .fnx file to codec too, not too bad but more work).
-  //     or we just have crappier RW testing like today.
-  // 2. PreFlex writes 3.x SIS format, and only reads 3.x
-  //    (in this case we have to move .fnx file to codec as well)
   @Override
   public void read(Directory directory, String segmentsFileName, ChecksumIndexInput input, SegmentInfos infos, IOContext context) throws IOException { 
     infos.version = input.readLong(); // read version
     infos.counter = input.readInt(); // read counter
     final int format = infos.getFormat();
+    assert format <= SegmentInfos.FORMAT_4_0;
     for (int i = input.readInt(); i > 0; i--) { // read segmentInfos
       SegmentInfo si = readSegmentInfo(directory, format, input);
-      if (si.getVersion() == null) {
-        // Could be a 3.0 - try to open the doc stores - if it fails, it's a
-        // 2.x segment, and an IndexFormatTooOldException will be thrown,
-        // which is what we want.
-        Directory dir = directory;
-        if (si.getDocStoreOffset() != -1) {
-          if (si.getDocStoreIsCompoundFile()) {
-            dir = new CompoundFileDirectory(dir, IndexFileNames.segmentFileName(
-                si.getDocStoreSegment(), "",
-                IndexFileNames.COMPOUND_FILE_STORE_EXTENSION), context, false);
-          }
-        } else if (si.getUseCompoundFile()) {
-          dir = new CompoundFileDirectory(dir, IndexFileNames.segmentFileName(
-              si.name, "", IndexFileNames.COMPOUND_FILE_EXTENSION), context, false);
-        }
-
-        try {
-          Lucene40StoredFieldsReader.checkCodeVersion(dir, si.getDocStoreSegment());
-        } finally {
-          // If we opened the directory, close it
-          if (dir != directory) dir.close();
-        }
-          
-        // Above call succeeded, so it's a 3.0 segment. Upgrade it so the next
-        // time the segment is read, its version won't be null and we won't
-        // need to open FieldsReader every time for each such segment.
-        si.setVersion("3.0");
-      } else if (si.getVersion().equals("2.x")) {
-        // If it's a 3x index touched by 3.1+ code, then segments record their
-        // version, whether they are 2.x ones or not. We detect that and throw
-        // appropriate exception.
-        throw new IndexFormatTooOldException("segment " + si.name + " in resource " + input, si.getVersion());
-      }
+      assert si.getVersion() != null;
       infos.add(si);
     }
       
     infos.userData = input.readStringStringMap();
   }
   
-  // if we make a preflex impl we can remove a lot of this hair...
   public SegmentInfo readSegmentInfo(Directory dir, int format, ChecksumIndexInput input) throws IOException {
-    final String version;
-    if (format <= SegmentInfos.FORMAT_3_1) {
-      version = input.readString();
-    } else {
-      version = null;
-    }
+    final String version = input.readString();
     final String name = input.readString();
     final int docCount = input.readInt();
     final long delGen = input.readLong();
+    // this is still written in 4.0 if we open a 3.x and upgrade the SI
     final int docStoreOffset = input.readInt();
     final String docStoreSegment;
     final boolean docStoreIsCompoundFile;
-    if (docStoreOffset != -1) {
+    if (docStoreOffset != -1) { 
       docStoreSegment = input.readString();
       docStoreIsCompoundFile = input.readByte() == SegmentInfo.YES;
     } else {
       docStoreSegment = name;
       docStoreIsCompoundFile = false;
     }
-
-    if (format > SegmentInfos.FORMAT_4_0) {
-      // pre-4.0 indexes write a byte if there is a single norms file
-      byte b = input.readByte();
-      assert 1 == b;
-    }
-
     final int numNormGen = input.readInt();
     final Map<Integer,Long> normGen;
     if (numNormGen == SegmentInfo.NO) {
@@ -128,61 +74,17 @@ public class Lucene40SegmentInfosReader extends SegmentInfosReader {
     } else {
       normGen = new HashMap<Integer, Long>();
       for(int j=0;j<numNormGen;j++) {
-        int fieldNumber = j;
-        if (format <= SegmentInfos.FORMAT_4_0) {
-          fieldNumber = input.readInt();
-        }
-
-        normGen.put(fieldNumber, input.readLong());
+        normGen.put(input.readInt(), input.readLong());
       }
     }
     final boolean isCompoundFile = input.readByte() == SegmentInfo.YES;
 
     final int delCount = input.readInt();
     assert delCount <= docCount;
-
     final int hasProx = input.readByte();
-
-    final Codec codec;
-    // note: if the codec is not available: Codec.forName will throw an exception.
-    if (format <= SegmentInfos.FORMAT_4_0) {
-      codec = Codec.forName(input.readString());
-    } else {
-      codec = Codec.forName("Lucene3x");
-    }
+    final Codec codec = Codec.forName(input.readString());
     final Map<String,String> diagnostics = input.readStringStringMap();
-
-    final int hasVectors;
-    if (format <= SegmentInfos.FORMAT_HAS_VECTORS) {
-      hasVectors = input.readByte();
-    } else {
-      final String storesSegment;
-      final String ext;
-      final boolean storeIsCompoundFile;
-      if (docStoreOffset != -1) {
-        storesSegment = docStoreSegment;
-        storeIsCompoundFile = docStoreIsCompoundFile;
-        ext = IndexFileNames.COMPOUND_FILE_STORE_EXTENSION;
-      } else {
-        storesSegment = name;
-        storeIsCompoundFile = isCompoundFile;
-        ext = IndexFileNames.COMPOUND_FILE_EXTENSION;
-      }
-      final Directory dirToTest;
-      if (storeIsCompoundFile) {
-        dirToTest = new CompoundFileDirectory(dir, IndexFileNames.segmentFileName(storesSegment, "", ext), IOContext.READONCE, false);
-      } else {
-        dirToTest = dir;
-      }
-      try {
-        // TODO: remove this manual file check or push to preflex codec
-        hasVectors = dirToTest.fileExists(IndexFileNames.segmentFileName(storesSegment, "", Lucene40TermVectorsReader.VECTORS_INDEX_EXTENSION)) ? SegmentInfo.YES : SegmentInfo.NO;
-      } finally {
-        if (isCompoundFile) {
-          dirToTest.close();
-        }
-      }
-    }
+    final int hasVectors = input.readByte();
     
     return new SegmentInfo(dir, version, name, docCount, delGen, docStoreOffset,
       docStoreSegment, docStoreIsCompoundFile, normGen, isCompoundFile,

lucene/src/java/org/apache/lucene/codecs/lucene40/Lucene40SegmentInfosWriter.java

@@ -43,6 +43,10 @@ public class Lucene40SegmentInfosWriter extends SegmentInfosWriter {
     IndexOutput out = createOutput(dir, segmentFileName, new IOContext(new FlushInfo(infos.size(), infos.totalDocCount())));
     boolean success = false;
     try {
+      /*
+       * TODO its not ideal that we write the format and the codecID inside the
+       * codec private classes but we read it in SegmentInfos.
+       */
       out.writeInt(SegmentInfos.FORMAT_CURRENT); // write FORMAT
       out.writeString(codecID); // write codecID
       out.writeLong(infos.version);
@@ -69,7 +73,7 @@ public class Lucene40SegmentInfosWriter extends SegmentInfosWriter {
     output.writeString(si.name);
     output.writeInt(si.docCount);
     output.writeLong(si.getDelGen());
-
+    // we still need to write this in 4.0 since we can open a 3.x with shared docStores
     output.writeInt(si.getDocStoreOffset());
     if (si.getDocStoreOffset() != -1) {
       output.writeString(si.getDocStoreSegment());

lucene/src/test-framework/java/org/apache/lucene/codecs/preflexrw/PreFlexRWCodec.java

@@ -20,6 +20,7 @@ package org.apache.lucene.codecs.preflexrw;
 import org.apache.lucene.codecs.FieldInfosFormat;
 import org.apache.lucene.codecs.NormsFormat;
 import org.apache.lucene.codecs.PostingsFormat;
+import org.apache.lucene.codecs.SegmentInfosFormat;
 import org.apache.lucene.codecs.TermVectorsFormat;
 import org.apache.lucene.codecs.lucene3x.Lucene3xCodec;
 import org.apache.lucene.util.LuceneTestCase;
@@ -33,6 +34,7 @@ public class PreFlexRWCodec extends Lucene3xCodec {
   private final NormsFormat norms = new PreFlexRWNormsFormat();
   private final FieldInfosFormat fieldInfos = new PreFlexRWFieldInfosFormat();
   private final TermVectorsFormat termVectors = new PreFlexRWTermVectorsFormat();
+  private final SegmentInfosFormat segmentInfos = new PreFlexRWSegmentInfosFormat();
   
   @Override
   public PostingsFormat postingsFormat() {
@@ -52,6 +54,15 @@ public class PreFlexRWCodec extends Lucene3xCodec {
     }
   }
 
+  @Override
+  public SegmentInfosFormat segmentInfosFormat() {
+    if (LuceneTestCase.PREFLEX_IMPERSONATION_IS_ACTIVE) {
+      return segmentInfos ;
+    } else {
+      return super.segmentInfosFormat();
+    }
+  }
+
   @Override
   public FieldInfosFormat fieldInfosFormat() {
     if (LuceneTestCase.PREFLEX_IMPERSONATION_IS_ACTIVE) {

lucene/src/test-framework/java/org/apache/lucene/codecs/preflexrw/PreFlexRWSegmentInfosFormat.java

@@ -0,0 +1,33 @@
+package org.apache.lucene.codecs.preflexrw;
+
+import org.apache.lucene.codecs.SegmentInfosWriter;
+import org.apache.lucene.codecs.lucene3x.Lucene3xSegmentInfosFormat;
+
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+/**
+ * @lucene.experimental
+ */
+public class PreFlexRWSegmentInfosFormat extends Lucene3xSegmentInfosFormat {
+  private final SegmentInfosWriter writer = new PreFlexRWSegmentInfosWriter();
+  
+  @Override
+  public SegmentInfosWriter getSegmentInfosWriter() {
+    return writer;
+  }
+}

lucene/src/test-framework/java/org/apache/lucene/codecs/preflexrw/PreFlexRWSegmentInfosWriter.java

@@ -0,0 +1,115 @@
+package org.apache.lucene.codecs.preflexrw;
+
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+import java.io.IOException;
+import java.util.Map;
+import java.util.Map.Entry;
+
+import org.apache.lucene.codecs.SegmentInfosWriter;
+import org.apache.lucene.index.SegmentInfo;
+import org.apache.lucene.index.SegmentInfos;
+import org.apache.lucene.store.ChecksumIndexOutput;
+import org.apache.lucene.store.Directory;
+import org.apache.lucene.store.FlushInfo;
+import org.apache.lucene.store.IOContext;
+import org.apache.lucene.store.IndexOutput;
+import org.apache.lucene.util.IOUtils;
+
+/**
+ * PreFlex implementation of {@link SegmentInfosWriter}.
+ * @lucene.experimental
+ */
+public class PreFlexRWSegmentInfosWriter extends SegmentInfosWriter {
+
+  @Override
+  public IndexOutput writeInfos(Directory dir, String segmentFileName, String codecID, SegmentInfos infos, IOContext context)
+          throws IOException {
+    IndexOutput out = createOutput(dir, segmentFileName, new IOContext(new FlushInfo(infos.size(), infos.totalDocCount())));
+    boolean success = false;
+    try {
+      out.writeInt(SegmentInfos.FORMAT_3_1); // write FORMAT
+      // we don't write a codec - this is 3.x
+      out.writeLong(infos.version);
+      out.writeInt(infos.counter); // write counter
+      out.writeInt(infos.size()); // write infos
+      for (SegmentInfo si : infos) {
+        writeInfo(out, si);
+      }
+      out.writeStringStringMap(infos.getUserData());
+      success = true;
+      return out;
+    } finally {
+      if (!success) {
+        IOUtils.closeWhileHandlingException(out);
+      }
+    }
+  }
+  
+  /** Save a single segment's info. */
+  private void writeInfo(IndexOutput output, SegmentInfo si) throws IOException {
+    assert si.getDelCount() <= si.docCount: "delCount=" + si.getDelCount() + " docCount=" + si.docCount + " segment=" + si.name;
+    // Write the Lucene version that created this segment, since 3.1
+    output.writeString(si.getVersion());
+    output.writeString(si.name);
+    output.writeInt(si.docCount);
+    output.writeLong(si.getDelGen());
+
+    output.writeInt(si.getDocStoreOffset());
+    if (si.getDocStoreOffset() != -1) {
+      output.writeString(si.getDocStoreSegment());
+      output.writeByte((byte) (si.getDocStoreIsCompoundFile() ? 1:0));
+    }
+    // pre-4.0 indexes write a byte if there is a single norms file
+    output.writeByte((byte) 1);
+
+    Map<Integer,Long> normGen = si.getNormGen();
+    if (normGen == null) {
+      output.writeInt(SegmentInfo.NO);
+    } else {
+      output.writeInt(normGen.size());
+      for (Entry<Integer,Long> entry : normGen.entrySet()) {
+        output.writeLong(entry.getValue());
+      }
+    }
+
+    output.writeByte((byte) (si.getUseCompoundFile() ? SegmentInfo.YES : SegmentInfo.NO));
+    output.writeInt(si.getDelCount());
+    output.writeByte((byte) (si.getHasProxInternal()));
+    output.writeStringStringMap(si.getDiagnostics());
+    output.writeByte((byte) (si.getHasVectorsInternal()));
+  }
+  
+  protected IndexOutput createOutput(Directory dir, String segmentFileName, IOContext context)
+      throws IOException {
+    IndexOutput plainOut = dir.createOutput(segmentFileName, context);
+    ChecksumIndexOutput out = new ChecksumIndexOutput(plainOut);
+    return out;
+  }
+
+  @Override
+  public void prepareCommit(IndexOutput segmentOutput) throws IOException {
+    ((ChecksumIndexOutput)segmentOutput).prepareCommit();
+  }
+
+  @Override
+  public void finishCommit(IndexOutput out) throws IOException {
+    ((ChecksumIndexOutput)out).finishCommit();
+    out.close();
+  }
+}