AVRO-4249 provide a cache of schema to avoid building

Author: mkeskells

lang/java/avro/src/main/java/org/apache/avro/file/DataFileStream.java

@@ -118,6 +118,12 @@ void validateMagic(byte[] magic) throws InvalidAvroMagicException {
 
   /** Initialize the stream by reading from its head. */
   void initialize(InputStream in, byte[] magic) throws IOException {
+    initialize(in, magic, SchemaCache.NO_CACHE);
+  }
+
+
+  /** Initialize the stream by reading from its head. */
+  protected void initialize(InputStream in, byte[] magic, SchemaCache schemaCache) throws IOException {
     this.header = new Header();
     this.vin = DecoderFactory.get().binaryDecoder(in, vin);
     magic = (magic == null) ? readMagic() : magic;
@@ -140,6 +146,8 @@ void initialize(InputStream in, byte[] magic) throws IOException {
 
     // finalize the header
     header.metaKeyList = Collections.unmodifiableList(header.metaKeyList);
+    header.schema = schemaCache.getOrParseSchema(getMetaString(DataFileConstants.SCHEMA));
+
     header.schema = new Schema.Parser(NameValidator.NO_VALIDATION).setValidateDefaults(false)
         .parse(getMetaString(DataFileConstants.SCHEMA));
     this.codec = resolveCodec();

lang/java/avro/src/main/java/org/apache/avro/file/SchemaCache.java

@@ -0,0 +1,33 @@
+package org.apache.avro.file;
+
+import org.apache.avro.NameValidator;
+import org.apache.avro.Schema;
+import java.util.concurrent.ConcurrentHashMap;
+import java.util.concurrent.ConcurrentMap;
+
+public abstract class SchemaCache {
+  public abstract Schema getOrParseSchema(String metaString);
+
+  protected Schema parse(String metaString) {
+    return new Schema.Parser(NameValidator.NO_VALIDATION).setValidateDefaults(false)
+        .parse(metaString);
+  }
+
+  public static final SchemaCache NO_CACHE = new SchemaCache() {
+    @Override
+    public Schema getOrParseSchema(String metaString) {
+      return parse(metaString);
+    }
+  };
+  public static SchemaCache createConcurrentCache() {
+    return new SchemaCache() {
+      private final ConcurrentMap<String, Schema> cache = new ConcurrentHashMap<>();
+
+      @Override
+      public Schema getOrParseSchema(String metaString) {
+        return cache.computeIfAbsent(metaString, this::parse);
+      }
+    };
+  }
+
+}