Remove FST constructor

dungba88 · dungba88 · commit 730e934c5795 · 2023-11-14T08:36:40.000+09:00
diff --git a/lucene/analysis/kuromoji/src/java/org/apache/lucene/analysis/ja/dict/TokenInfoDictionary.java b/lucene/analysis/kuromoji/src/java/org/apache/lucene/analysis/ja/dict/TokenInfoDictionary.java
@@ -103,7 +103,8 @@ private TokenInfoDictionary(
     FST<Long> fst;
     try (InputStream is = new BufferedInputStream(fstResource.get())) {
       DataInput in = new InputStreamDataInput(is);
-      fst = new FST<>(in, in, PositiveIntOutputs.getSingleton());
+      FST.FSTMetadata<Long> metadata = FST.readMetadata(in, PositiveIntOutputs.getSingleton());
+      fst = new FST<>(metadata, in, PositiveIntOutputs.getSingleton());
     }
     // TODO: some way to configure?
     this.fst = new TokenInfoFST(fst, true);
diff --git a/lucene/analysis/nori/src/java/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary.java b/lucene/analysis/nori/src/java/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary.java
@@ -102,7 +102,8 @@ private TokenInfoDictionary(
     FST<Long> fst;
     try (InputStream is = new BufferedInputStream(fstResource.get())) {
       DataInput in = new InputStreamDataInput(is);
-      fst = new FST<>(in, in, PositiveIntOutputs.getSingleton());
+      FST.FSTMetadata<Long> metadata = FST.readMetadata(in, PositiveIntOutputs.getSingleton());
+      fst = new FST<>(metadata, in, PositiveIntOutputs.getSingleton());
     }
     this.fst = new TokenInfoFST(fst);
   }
diff --git a/lucene/backward-codecs/src/java/org/apache/lucene/backward_codecs/lucene40/blocktree/FieldReader.java b/lucene/backward-codecs/src/java/org/apache/lucene/backward_codecs/lucene40/blocktree/FieldReader.java
@@ -89,9 +89,13 @@ public final class FieldReader extends Terms {
     final IndexInput clone = indexIn.clone();
     clone.seek(indexStartFP);
     if (metaIn == indexIn) { // Only true before Lucene 8.6
-      index = new FST<>(clone, clone, ByteSequenceOutputs.getSingleton(), new OffHeapFSTStore());
+      FST.FSTMetadata<BytesRef> metadata =
+          FST.readMetadata(clone, ByteSequenceOutputs.getSingleton());
+      index = new FST<>(metadata, clone, ByteSequenceOutputs.getSingleton(), new OffHeapFSTStore());
     } else {
-      index = new FST<>(metaIn, clone, ByteSequenceOutputs.getSingleton(), new OffHeapFSTStore());
+      FST.FSTMetadata<BytesRef> metadata =
+          FST.readMetadata(metaIn, ByteSequenceOutputs.getSingleton());
+      index = new FST<>(metadata, clone, ByteSequenceOutputs.getSingleton(), new OffHeapFSTStore());
     }
     /*
      if (false) {
diff --git a/lucene/codecs/src/java/org/apache/lucene/codecs/blockterms/VariableGapTermsIndexReader.java b/lucene/codecs/src/java/org/apache/lucene/codecs/blockterms/VariableGapTermsIndexReader.java
@@ -154,7 +154,7 @@ private final class FieldIndexData implements Accountable {
     public FieldIndexData(IndexInput in, FieldInfo fieldInfo, long indexStart) throws IOException {
       IndexInput clone = in.clone();
       clone.seek(indexStart);
-      fst = new FST<>(clone, clone, fstOutputs);
+      fst = new FST<>(FST.readMetadata(clone, fstOutputs), clone, fstOutputs);
       clone.close();
 
       /*
diff --git a/lucene/codecs/src/java/org/apache/lucene/codecs/blocktreeords/OrdsFieldReader.java b/lucene/codecs/src/java/org/apache/lucene/codecs/blocktreeords/OrdsFieldReader.java
@@ -85,7 +85,9 @@ final class OrdsFieldReader extends Terms {
       final IndexInput clone = indexIn.clone();
       // System.out.println("start=" + indexStartFP + " field=" + fieldInfo.name);
       clone.seek(indexStartFP);
-      index = new FST<>(clone, clone, OrdsBlockTreeTermsWriter.FST_OUTPUTS);
+      FST.FSTMetadata<Output> metadata =
+          FST.readMetadata(clone, OrdsBlockTreeTermsWriter.FST_OUTPUTS);
+      index = new FST<>(metadata, clone, OrdsBlockTreeTermsWriter.FST_OUTPUTS);
 
       /*
       if (true) {
diff --git a/lucene/codecs/src/java/org/apache/lucene/codecs/memory/FSTTermsReader.java b/lucene/codecs/src/java/org/apache/lucene/codecs/memory/FSTTermsReader.java
@@ -194,7 +194,8 @@ final class TermsReader extends Terms {
       this.sumDocFreq = sumDocFreq;
       this.docCount = docCount;
       OffHeapFSTStore offHeapFSTStore = new OffHeapFSTStore();
-      this.dict = new FST<>(in, in, new FSTTermOutputs(fieldInfo), offHeapFSTStore);
+      FSTTermOutputs outputs = new FSTTermOutputs(fieldInfo);
+      this.dict = new FST<>(FST.readMetadata(in, outputs), in, outputs, offHeapFSTStore);
       in.skipBytes(offHeapFSTStore.size());
     }
 
diff --git a/lucene/codecs/src/java/org/apache/lucene/codecs/uniformsplit/FSTDictionary.java b/lucene/codecs/src/java/org/apache/lucene/codecs/uniformsplit/FSTDictionary.java
@@ -89,10 +89,11 @@ protected static FSTDictionary read(
       isFSTOnHeap = true;
     }
     PositiveIntOutputs fstOutputs = PositiveIntOutputs.getSingleton();
+    FST.FSTMetadata<Long> metadata = FST.readMetadata(fstDataInput, fstOutputs);
     FST<Long> fst =
         isFSTOnHeap
-            ? new FST<>(fstDataInput, fstDataInput, fstOutputs)
-            : new FST<>(fstDataInput, fstDataInput, fstOutputs, new OffHeapFSTStore());
+            ? new FST<>(metadata, fstDataInput, fstOutputs)
+            : new FST<>(metadata, fstDataInput, fstOutputs, new OffHeapFSTStore());
     return new FSTDictionary(fst);
   }
 
diff --git a/lucene/core/src/java/org/apache/lucene/codecs/lucene90/blocktree/FieldReader.java b/lucene/core/src/java/org/apache/lucene/codecs/lucene90/blocktree/FieldReader.java
@@ -91,7 +91,9 @@ public final class FieldReader extends Terms {
     // Initialize FST always off-heap.
     final IndexInput clone = indexIn.clone();
     clone.seek(indexStartFP);
-    index = new FST<>(metaIn, clone, ByteSequenceOutputs.getSingleton(), new OffHeapFSTStore());
+    FST.FSTMetadata<BytesRef> metadata =
+        FST.readMetadata(metaIn, ByteSequenceOutputs.getSingleton());
+    index = new FST<>(metadata, clone, ByteSequenceOutputs.getSingleton(), new OffHeapFSTStore());
     /*
      if (false) {
      final String dotFileName = segment + "_" + fieldInfo.name + ".dot";
diff --git a/lucene/core/src/java/org/apache/lucene/util/fst/FST.java b/lucene/core/src/java/org/apache/lucene/util/fst/FST.java
@@ -404,18 +404,8 @@ private static boolean flag(int flags, int bit) {
    * Load a previously saved FST with a DataInput for metdata using an {@link OnHeapFSTStore} with
    * maxBlockBits set to {@link #DEFAULT_MAX_BLOCK_BITS}
    */
-  public FST(DataInput metaIn, DataInput in, Outputs<T> outputs) throws IOException {
-    this(metaIn, in, outputs, new OnHeapFSTStore(DEFAULT_MAX_BLOCK_BITS));
-  }
-
-  /**
-   * Load a previously saved FST with a DataInput for metdata and a FSTStore. If using {@link
-   * OnHeapFSTStore}, setting maxBlockBits allows you to control the size of the byte[] pages used
-   * to hold the FST bytes.
-   */
-  public FST(DataInput metaIn, DataInput in, Outputs<T> outputs, FSTStore fstStore)
-      throws IOException {
-    this(readMetadata(metaIn, outputs), in, outputs, fstStore);
+  public FST(FSTMetadata<T> metadata, DataInput in, Outputs<T> outputs) throws IOException {
+    this(metadata, in, outputs, new OnHeapFSTStore(DEFAULT_MAX_BLOCK_BITS));
   }
 
   /**
@@ -574,7 +564,7 @@ public void save(final Path path) throws IOException {
   public static <T> FST<T> read(Path path, Outputs<T> outputs) throws IOException {
     try (InputStream is = Files.newInputStream(path)) {
       DataInput in = new InputStreamDataInput(new BufferedInputStream(is));
-      return new FST<>(in, in, outputs);
+      return new FST<>(readMetadata(in, outputs), in, outputs);
     }
   }
 
diff --git a/lucene/core/src/test/org/apache/lucene/util/fst/Test2BFST.java b/lucene/core/src/test/org/apache/lucene/util/fst/Test2BFST.java
@@ -140,7 +140,7 @@ public void test() throws Exception {
             fst.save(out, out);
             out.close();
             IndexInput in = dir.openInput("fst", IOContext.DEFAULT);
-            fst = new FST<>(in, in, outputs);
+            fst = new FST<>(FST.readMetadata(in, outputs), in, outputs);
             in.close();
           } else {
             dir.deleteFile("fst");
@@ -226,7 +226,7 @@ public void test() throws Exception {
             fst.save(out, out);
             out.close();
             IndexInput in = dir.openInput("fst", IOContext.DEFAULT);
-            fst = new FST<>(in, in, outputs);
+            fst = new FST<>(FST.readMetadata(in, outputs), in, outputs);
             in.close();
           } else {
             dir.deleteFile("fst");
@@ -317,7 +317,7 @@ public void test() throws Exception {
             fst.save(out, out);
             out.close();
             IndexInput in = dir.openInput("fst", IOContext.DEFAULT);
-            fst = new FST<>(in, in, outputs);
+            fst = new FST<>(FST.readMetadata(in, outputs), in, outputs);
             in.close();
           } else {
             dir.deleteFile("fst");
diff --git a/lucene/core/src/test/org/apache/lucene/util/fst/TestFSTDirectAddressing.java b/lucene/core/src/test/org/apache/lucene/util/fst/TestFSTDirectAddressing.java
@@ -218,7 +218,8 @@ public static void main(String... args) throws Exception {
   private static void countFSTArcs(String fstFilePath) throws IOException {
     byte[] buf = Files.readAllBytes(Paths.get(fstFilePath));
     DataInput in = new ByteArrayDataInput(buf);
-    FST<BytesRef> fst = new FST<>(in, in, ByteSequenceOutputs.getSingleton());
+    FST.FSTMetadata<BytesRef> metadata = FST.readMetadata(in, ByteSequenceOutputs.getSingleton());
+    FST<BytesRef> fst = new FST<>(metadata, in, ByteSequenceOutputs.getSingleton());
     BytesRefFSTEnum<BytesRef> fstEnum = new BytesRefFSTEnum<>(fst);
     int binarySearchArcCount = 0,
         directAddressingArcCount = 0,
@@ -285,7 +286,8 @@ private static void recompileAndWalk(String fstFilePath) throws IOException {
 
       System.out.println("Reading FST");
       long startTimeMs = System.nanoTime();
-      FST<CharsRef> originalFst = new FST<>(in, in, CharSequenceOutputs.getSingleton());
+      FST.FSTMetadata<CharsRef> metadata = FST.readMetadata(in, CharSequenceOutputs.getSingleton());
+      FST<CharsRef> originalFst = new FST<>(metadata, in, CharSequenceOutputs.getSingleton());
       long endTimeMs = System.nanoTime();
       System.out.println(
           "time = " + TimeUnit.NANOSECONDS.toMillis(endTimeMs - startTimeMs) + " ms");
diff --git a/lucene/core/src/test/org/apache/lucene/util/fst/TestFSTs.java b/lucene/core/src/test/org/apache/lucene/util/fst/TestFSTs.java
@@ -1219,7 +1219,7 @@ public void testSaveDifferentMetaOut() throws Exception {
 
     // load the FST, which will force it to use FSTStore instead of BytesStore
     ByteArrayDataInput in = new ByteArrayDataInput(outOS.toByteArray());
-    FST<Long> loadedFST = new FST<>(in, in, outputs);
+    FST<Long> loadedFST = new FST<>(FST.readMetadata(in, outputs), in, outputs);
 
     // now save the FST again, this time to different DataOutput for meta
     ByteArrayOutputStream metdataOS = new ByteArrayOutputStream();
@@ -1231,7 +1231,7 @@ public void testSaveDifferentMetaOut() throws Exception {
     // finally load it again
     ByteArrayDataInput metaIn = new ByteArrayDataInput(metdataOS.toByteArray());
     ByteArrayDataInput dataIn = new ByteArrayDataInput(dataOS.toByteArray());
-    loadedFST = new FST<>(metaIn, dataIn, outputs);
+    loadedFST = new FST<>(FST.readMetadata(metaIn, outputs), dataIn, outputs);
 
     assertEquals(22L, Util.get(loadedFST, Util.toIntsRef(newBytesRef("aab"), scratch)).longValue());
     assertEquals(7L, Util.get(loadedFST, Util.toIntsRef(newBytesRef("aac"), scratch)).longValue());
@@ -1291,7 +1291,7 @@ public void testNonFinalStopNode() throws Exception {
     out.close();
 
     IndexInput in = dir.openInput("fst", IOContext.DEFAULT);
-    final FST<Long> fst2 = new FST<>(in, in, outputs);
+    final FST<Long> fst2 = new FST<>(FST.readMetadata(in, outputs), in, outputs);
     checkStopNodes(fst2, outputs);
     in.close();
     dir.close();
diff --git a/lucene/demo/src/java/org/apache/lucene/demo/knn/KnnVectorDict.java b/lucene/demo/src/java/org/apache/lucene/demo/knn/KnnVectorDict.java
@@ -58,7 +58,8 @@ public class KnnVectorDict implements Closeable {
    */
   public KnnVectorDict(Directory directory, String dictName) throws IOException {
     try (IndexInput fstIn = directory.openInput(dictName + ".fst", IOContext.READ)) {
-      fst = new FST<>(fstIn, fstIn, PositiveIntOutputs.getSingleton());
+      FST.FSTMetadata<Long> metadata = FST.readMetadata(fstIn, PositiveIntOutputs.getSingleton());
+      fst = new FST<>(metadata, fstIn, PositiveIntOutputs.getSingleton());
     }
 
     vectors = directory.openInput(dictName + ".bin", IOContext.READ);
diff --git a/lucene/sandbox/src/java/org/apache/lucene/sandbox/codecs/idversion/VersionFieldReader.java b/lucene/sandbox/src/java/org/apache/lucene/sandbox/codecs/idversion/VersionFieldReader.java
@@ -86,7 +86,11 @@ final class VersionFieldReader extends Terms {
       final IndexInput clone = indexIn.clone();
       // System.out.println("start=" + indexStartFP + " field=" + fieldInfo.name);
       clone.seek(indexStartFP);
-      index = new FST<>(clone, clone, VersionBlockTreeTermsWriter.FST_OUTPUTS);
+      index =
+          new FST<>(
+              FST.readMetadata(clone, VersionBlockTreeTermsWriter.FST_OUTPUTS),
+              clone,
+              VersionBlockTreeTermsWriter.FST_OUTPUTS);
 
       /*
         if (false) {
diff --git a/lucene/suggest/src/java/org/apache/lucene/search/suggest/analyzing/AnalyzingSuggester.java b/lucene/suggest/src/java/org/apache/lucene/search/suggest/analyzing/AnalyzingSuggester.java
@@ -612,12 +612,9 @@ public boolean store(DataOutput output) throws IOException {
   @Override
   public boolean load(DataInput input) throws IOException {
     count = input.readVLong();
-    this.fst =
-        new FST<>(
-            input,
-            input,
-            new PairOutputs<>(
-                PositiveIntOutputs.getSingleton(), ByteSequenceOutputs.getSingleton()));
+    PairOutputs<Long, BytesRef> outputs =
+        new PairOutputs<>(PositiveIntOutputs.getSingleton(), ByteSequenceOutputs.getSingleton());
+    this.fst = new FST<>(FST.readMetadata(input, outputs), input, outputs);
     maxAnalyzedPathsForOneInput = input.readVInt();
     hasPayloads = input.readByte() == 1;
     return true;
diff --git a/lucene/suggest/src/java/org/apache/lucene/search/suggest/analyzing/FreeTextSuggester.java b/lucene/suggest/src/java/org/apache/lucene/search/suggest/analyzing/FreeTextSuggester.java
@@ -383,7 +383,8 @@ public boolean load(DataInput input) throws IOException {
     }
     totTokens = input.readVLong();
 
-    fst = new FST<>(input, input, PositiveIntOutputs.getSingleton());
+    FST.FSTMetadata<Long> metadata = FST.readMetadata(input, PositiveIntOutputs.getSingleton());
+    fst = new FST<>(metadata, input, PositiveIntOutputs.getSingleton());
 
     return true;
   }
diff --git a/lucene/suggest/src/java/org/apache/lucene/search/suggest/document/NRTSuggester.java b/lucene/suggest/src/java/org/apache/lucene/search/suggest/document/NRTSuggester.java
@@ -337,25 +337,16 @@ private static boolean shouldLoadFSTOffHeap(IndexInput input, FSTLoadMode fstLoa
    */
   public static NRTSuggester load(IndexInput input, FSTLoadMode fstLoadMode) throws IOException {
     final FST<Pair<Long, BytesRef>> fst;
+    PairOutputs<Long, BytesRef> outputs =
+        new PairOutputs<>(PositiveIntOutputs.getSingleton(), ByteSequenceOutputs.getSingleton());
     if (shouldLoadFSTOffHeap(input, fstLoadMode)) {
       OffHeapFSTStore store = new OffHeapFSTStore();
       IndexInput clone = input.clone();
       clone.seek(input.getFilePointer());
-      fst =
-          new FST<>(
-              clone,
-              clone,
-              new PairOutputs<>(
-                  PositiveIntOutputs.getSingleton(), ByteSequenceOutputs.getSingleton()),
-              store);
+      fst = new FST<>(FST.readMetadata(clone, outputs), clone, outputs, store);
       input.seek(clone.getFilePointer() + store.size());
     } else {
-      fst =
-          new FST<>(
-              input,
-              input,
-              new PairOutputs<>(
-                  PositiveIntOutputs.getSingleton(), ByteSequenceOutputs.getSingleton()));
+      fst = new FST<>(FST.readMetadata(input, outputs), input, outputs);
     }
 
     /* read some meta info */
diff --git a/lucene/suggest/src/java/org/apache/lucene/search/suggest/fst/FSTCompletionLookup.java b/lucene/suggest/src/java/org/apache/lucene/search/suggest/fst/FSTCompletionLookup.java
@@ -300,8 +300,9 @@ public synchronized boolean store(DataOutput output) throws IOException {
   @Override
   public synchronized boolean load(DataInput input) throws IOException {
     count = input.readVLong();
+    FST.FSTMetadata<Object> metadata = FST.readMetadata(input, NoOutputs.getSingleton());
     this.higherWeightsCompletion =
-        new FSTCompletion(new FST<>(input, input, NoOutputs.getSingleton()));
+        new FSTCompletion(new FST<>(metadata, input, NoOutputs.getSingleton()));
     this.normalCompletion =
         new FSTCompletion(higherWeightsCompletion.getFST(), false, exactMatchFirst);
     return true;
diff --git a/lucene/suggest/src/java/org/apache/lucene/search/suggest/fst/WFSTCompletionLookup.java b/lucene/suggest/src/java/org/apache/lucene/search/suggest/fst/WFSTCompletionLookup.java
@@ -140,7 +140,8 @@ public boolean store(DataOutput output) throws IOException {
   @Override
   public boolean load(DataInput input) throws IOException {
     count = input.readVLong();
-    this.fst = new FST<>(input, input, PositiveIntOutputs.getSingleton());
+    FST.FSTMetadata<Long> metadata = FST.readMetadata(input, PositiveIntOutputs.getSingleton());
+    this.fst = new FST<>(metadata, input, PositiveIntOutputs.getSingleton());
     return true;
   }
 
diff --git a/lucene/test-framework/src/java/org/apache/lucene/tests/util/fst/FSTTester.java b/lucene/test-framework/src/java/org/apache/lucene/tests/util/fst/FSTTester.java
@@ -281,7 +281,7 @@ public FST<T> doTest() throws IOException {
         fst.save(out, out);
       }
       try (IndexInput in = dir.openInput("fst.bin", context)) {
-        fst = new FST<>(in, in, outputs);
+        fst = new FST<>(FST.readMetadata(in, outputs), in, outputs);
       } finally {
         dir.deleteFile("fst.bin");
       }

Original file line number	Diff line number	Diff line change
`@@ -102,7 +102,8 @@ private TokenInfoDictionary(`
`102`	`102`	`FST<Long> fst;`
`103`	`103`	`try (InputStream is = new BufferedInputStream(fstResource.get())) {`
`104`	`104`	`DataInput in = new InputStreamDataInput(is);`
`105`		`- fst = new FST<>(in, in, PositiveIntOutputs.getSingleton());`
	`105`	`+ FST.FSTMetadata<Long> metadata = FST.readMetadata(in, PositiveIntOutputs.getSingleton());`
	`106`	`+ fst = new FST<>(metadata, in, PositiveIntOutputs.getSingleton());`
`106`	`107`	`}`
`107`	`108`	`this.fst = new TokenInfoFST(fst);`
`108`	`109`	`}`
Original file line number	Diff line number	Diff line change
`@@ -194,7 +194,8 @@ final class TermsReader extends Terms {`
`194`	`194`	`this.sumDocFreq = sumDocFreq;`
`195`	`195`	`this.docCount = docCount;`
`196`	`196`	`OffHeapFSTStore offHeapFSTStore = new OffHeapFSTStore();`
`197`		`- this.dict = new FST<>(in, in, new FSTTermOutputs(fieldInfo), offHeapFSTStore);`
	`197`	`+ FSTTermOutputs outputs = new FSTTermOutputs(fieldInfo);`
	`198`	`+ this.dict = new FST<>(FST.readMetadata(in, outputs), in, outputs, offHeapFSTStore);`
`198`	`199`	`in.skipBytes(offHeapFSTStore.size());`
`199`	`200`	`}`
`200`	`201`
Original file line number	Diff line number	Diff line change
`@@ -89,10 +89,11 @@ protected static FSTDictionary read(`
`89`	`89`	`isFSTOnHeap = true;`
`90`	`90`	`}`
`91`	`91`	`PositiveIntOutputs fstOutputs = PositiveIntOutputs.getSingleton();`
	`92`	`+ FST.FSTMetadata<Long> metadata = FST.readMetadata(fstDataInput, fstOutputs);`
`92`	`93`	`FST<Long> fst =`
`93`	`94`	`isFSTOnHeap`
`94`		`- ? new FST<>(fstDataInput, fstDataInput, fstOutputs)`
`95`		`- : new FST<>(fstDataInput, fstDataInput, fstOutputs, new OffHeapFSTStore());`
	`95`	`+ ? new FST<>(metadata, fstDataInput, fstOutputs)`
	`96`	`+ : new FST<>(metadata, fstDataInput, fstOutputs, new OffHeapFSTStore());`
`96`	`97`	`return new FSTDictionary(fst);`
`97`	`98`	`}`
`98`	`99`