trinodb · electrum · Mar 15, 2023 · Mar 8, 2023
diff --git a/plugin/trino-hive/src/main/java/io/trino/plugin/hive/HiveErrorCode.java b/plugin/trino-hive/src/main/java/io/trino/plugin/hive/HiveErrorCode.java
@@ -33,7 +33,7 @@ public enum HiveErrorCode
     HIVE_PARTITION_OFFLINE(6, USER_ERROR),
     HIVE_BAD_DATA(7, EXTERNAL),
     HIVE_PARTITION_SCHEMA_MISMATCH(8, EXTERNAL),
-    HIVE_MISSING_DATA(9, EXTERNAL),
+    // HIVE_MISSING_DATA(9, EXTERNAL) is deprecated
     HIVE_INVALID_PARTITION_VALUE(10, EXTERNAL),
     HIVE_TIMEZONE_MISMATCH(11, EXTERNAL),
     HIVE_INVALID_METADATA(12, EXTERNAL),

diff --git a/plugin/trino-hive/src/main/java/io/trino/plugin/hive/line/LinePageSourceFactory.java b/plugin/trino-hive/src/main/java/io/trino/plugin/hive/line/LinePageSourceFactory.java
@@ -40,9 +40,7 @@
 import io.trino.spi.predicate.TupleDomain;
 import org.apache.hadoop.conf.Configuration;
 import org.apache.hadoop.fs.Path;
-import org.apache.hadoop.hdfs.BlockMissingException;
 
-import java.io.FileNotFoundException;
 import java.io.InputStream;
 import java.util.List;
 import java.util.Optional;
@@ -51,12 +49,10 @@
 import java.util.function.Predicate;
 
 import static com.google.common.base.Preconditions.checkArgument;
-import static com.google.common.base.Strings.nullToEmpty;
 import static com.google.common.collect.ImmutableList.toImmutableList;
 import static io.trino.hive.formats.line.LineDeserializer.EMPTY_LINE_DESERIALIZER;
 import static io.trino.hive.thrift.metastore.hive_metastoreConstants.FILE_INPUT_FORMAT;
 import static io.trino.plugin.hive.HiveErrorCode.HIVE_CANNOT_OPEN_SPLIT;
-import static io.trino.plugin.hive.HiveErrorCode.HIVE_MISSING_DATA;
 import static io.trino.plugin.hive.HivePageSourceProvider.projectBaseColumns;
 import static io.trino.plugin.hive.ReaderPageSource.noProjectionAdaptation;
 import static io.trino.plugin.hive.util.HiveUtil.getDeserializerClassName;
@@ -163,10 +159,6 @@ public Optional<ReaderPageSource> createPageSource(
             throw e;
         }
         catch (Exception e) {
-            if (nullToEmpty(e.getMessage()).trim().equals("Filesystem closed") ||
-                    e instanceof FileNotFoundException) {
-                throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, e);
-            }
             throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, splitError(e, path, start, length), e);
         }
 
@@ -180,15 +172,11 @@ public Optional<ReaderPageSource> createPageSource(
             LinePageSource pageSource = new LinePageSource(lineReader, lineDeserializer, lineReaderFactory.createLineBuffer(), path.toString());
             return Optional.of(new ReaderPageSource(pageSource, readerProjections));
         }
-        catch (Throwable e) {
-            if (e instanceof TrinoException) {
-                throw (TrinoException) e;
-            }
-            String message = splitError(e, path, start, length);
-            if (e instanceof BlockMissingException) {
-                throw new TrinoException(HIVE_MISSING_DATA, message, e);
-            }
-            throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, message, e);
+        catch (TrinoException e) {
+            throw e;
+        }
+        catch (Exception e) {
+            throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, splitError(e, path, start, length), e);
         }
     }
 

diff --git a/plugin/trino-hive/src/main/java/io/trino/plugin/hive/orc/HdfsOrcDataSource.java b/plugin/trino-hive/src/main/java/io/trino/plugin/hive/orc/HdfsOrcDataSource.java
@@ -21,12 +21,10 @@
 import io.trino.orc.OrcReaderOptions;
 import io.trino.plugin.hive.FileFormatDataSourceStats;
 import io.trino.spi.TrinoException;
-import org.apache.hadoop.hdfs.BlockMissingException;
 
 import java.io.IOException;
 
 import static io.trino.plugin.hive.HiveErrorCode.HIVE_FILESYSTEM_ERROR;
-import static io.trino.plugin.hive.HiveErrorCode.HIVE_MISSING_DATA;
 import static io.trino.plugin.hive.HiveErrorCode.HIVE_UNKNOWN_ERROR;
 import static java.lang.String.format;
 import static java.util.Objects.requireNonNull;
@@ -82,9 +80,6 @@ protected void readInternal(long position, byte[] buffer, int bufferOffset, int
         }
         catch (Exception e) {
             String message = format("Error reading from %s at position %s", this, position);
-            if (e instanceof BlockMissingException) {
-                throw new TrinoException(HIVE_MISSING_DATA, message, e);
-            }
             if (e instanceof IOException) {
                 throw new TrinoException(HIVE_FILESYSTEM_ERROR, message, e);
             }

diff --git a/plugin/trino-hive/src/main/java/io/trino/plugin/hive/orc/OrcDeleteDeltaPageSource.java b/plugin/trino-hive/src/main/java/io/trino/plugin/hive/orc/OrcDeleteDeltaPageSource.java
@@ -30,25 +30,22 @@
 import io.trino.spi.TrinoException;
 import io.trino.spi.connector.ConnectorPageSource;
 import org.apache.hadoop.fs.Path;
-import org.apache.hadoop.hdfs.BlockMissingException;
 
-import java.io.FileNotFoundException;
 import java.io.IOException;
 import java.io.UncheckedIOException;
 import java.util.List;
 import java.util.Map;
 import java.util.Optional;
 
 import static com.google.common.base.MoreObjects.toStringHelper;
-import static com.google.common.base.Strings.nullToEmpty;
 import static com.google.common.collect.Maps.uniqueIndex;
 import static io.trino.memory.context.AggregatedMemoryContext.newSimpleAggregatedMemoryContext;
 import static io.trino.orc.OrcReader.MAX_BATCH_SIZE;
 import static io.trino.orc.OrcReader.createOrcReader;
 import static io.trino.orc.OrcReader.fullyProjectedLayout;
 import static io.trino.plugin.base.util.Closables.closeAllSuppress;
+import static io.trino.plugin.hive.HiveErrorCode.HIVE_BAD_DATA;
 import static io.trino.plugin.hive.HiveErrorCode.HIVE_CANNOT_OPEN_SPLIT;
-import static io.trino.plugin.hive.HiveErrorCode.HIVE_MISSING_DATA;
 import static io.trino.plugin.hive.acid.AcidSchema.ACID_COLUMN_BUCKET;
 import static io.trino.plugin.hive.acid.AcidSchema.ACID_COLUMN_ORIGINAL_TRANSACTION;
 import static io.trino.plugin.hive.acid.AcidSchema.ACID_COLUMN_ROW_ID;
@@ -87,10 +84,6 @@ public static Optional<ConnectorPageSource> createOrcDeleteDeltaPageSource(
                     stats);
         }
         catch (Exception e) {
-            if (nullToEmpty(e.getMessage()).trim().equals("Filesystem closed") ||
-                    e instanceof FileNotFoundException) {
-                throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, e);
-            }
             throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, openError(e, path), e);
         }
 
@@ -111,11 +104,10 @@ public static Optional<ConnectorPageSource> createOrcDeleteDeltaPageSource(
             if (e instanceof TrinoException) {
                 throw (TrinoException) e;
             }
-            String message = openError(e, path);
-            if (e instanceof BlockMissingException) {
-                throw new TrinoException(HIVE_MISSING_DATA, message, e);
+            if (e instanceof OrcCorruptionException) {
+                throw new TrinoException(HIVE_BAD_DATA, e);
             }
-            throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, message, e);
+            throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, openError(e, path), e);
         }
     }
 

diff --git a/plugin/trino-hive/src/main/java/io/trino/plugin/hive/orc/OrcPageSourceFactory.java b/plugin/trino-hive/src/main/java/io/trino/plugin/hive/orc/OrcPageSourceFactory.java
@@ -23,6 +23,7 @@
 import io.trino.memory.context.AggregatedMemoryContext;
 import io.trino.orc.NameBasedFieldMapper;
 import io.trino.orc.OrcColumn;
+import io.trino.orc.OrcCorruptionException;
 import io.trino.orc.OrcDataSource;
 import io.trino.orc.OrcDataSourceId;
 import io.trino.orc.OrcReader;
@@ -52,12 +53,10 @@
 import io.trino.spi.type.Type;
 import org.apache.hadoop.conf.Configuration;
 import org.apache.hadoop.fs.Path;
-import org.apache.hadoop.hdfs.BlockMissingException;
 import org.joda.time.DateTimeZone;
 
 import javax.inject.Inject;
 
-import java.io.FileNotFoundException;
 import java.io.IOException;
 import java.util.ArrayList;
 import java.util.List;
@@ -69,7 +68,6 @@
 import java.util.stream.Collectors;
 
 import static com.google.common.base.Preconditions.checkArgument;
-import static com.google.common.base.Strings.nullToEmpty;
 import static com.google.common.collect.ImmutableList.toImmutableList;
 import static com.google.common.collect.ImmutableMap.toImmutableMap;
 import static com.google.common.collect.Maps.uniqueIndex;
@@ -86,7 +84,6 @@
 import static io.trino.plugin.hive.HiveErrorCode.HIVE_BAD_DATA;
 import static io.trino.plugin.hive.HiveErrorCode.HIVE_CANNOT_OPEN_SPLIT;
 import static io.trino.plugin.hive.HiveErrorCode.HIVE_FILE_MISSING_COLUMN_NAMES;
-import static io.trino.plugin.hive.HiveErrorCode.HIVE_MISSING_DATA;
 import static io.trino.plugin.hive.HivePageSourceProvider.projectBaseColumns;
 import static io.trino.plugin.hive.HiveSessionProperties.getOrcLazyReadSmallRanges;
 import static io.trino.plugin.hive.HiveSessionProperties.getOrcMaxBufferSize;
@@ -270,10 +267,6 @@ private ConnectorPageSource createOrcPageSource(
                     stats);
         }
         catch (Exception e) {
-            if (nullToEmpty(e.getMessage()).trim().equals("Filesystem closed") ||
-                    e instanceof FileNotFoundException) {
-                throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, e);
-            }
             throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, splitError(e, path, start, length), e);
         }
 
@@ -453,11 +446,10 @@ else if (column.getBaseHiveColumnIndex() < fileColumns.size()) {
             if (e instanceof TrinoException) {
                 throw (TrinoException) e;
             }
-            String message = splitError(e, path, start, length);
-            if (e instanceof BlockMissingException) {
-                throw new TrinoException(HIVE_MISSING_DATA, message, e);
+            if (e instanceof OrcCorruptionException) {
+                throw new TrinoException(HIVE_BAD_DATA, e);
             }
-            throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, message, e);
+            throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, splitError(e, path, start, length), e);
         }
     }
 

diff --git a/plugin/trino-hive/src/main/java/io/trino/plugin/hive/parquet/ParquetPageSourceFactory.java b/plugin/trino-hive/src/main/java/io/trino/plugin/hive/parquet/ParquetPageSourceFactory.java
@@ -46,7 +46,6 @@
 import io.trino.spi.predicate.TupleDomain;
 import org.apache.hadoop.conf.Configuration;
 import org.apache.hadoop.fs.Path;
-import org.apache.hadoop.hdfs.BlockMissingException;
 import org.apache.parquet.column.ColumnDescriptor;
 import org.apache.parquet.hadoop.metadata.BlockMetaData;
 import org.apache.parquet.hadoop.metadata.ColumnChunkMetaData;
@@ -61,7 +60,6 @@
 
 import javax.inject.Inject;
 
-import java.io.FileNotFoundException;
 import java.io.IOException;
 import java.util.HashSet;
 import java.util.List;
@@ -73,7 +71,6 @@
 import java.util.Set;
 
 import static com.google.common.base.Preconditions.checkArgument;
-import static com.google.common.base.Strings.nullToEmpty;
 import static com.google.common.collect.ImmutableList.toImmutableList;
 import static com.google.common.collect.ImmutableSet.toImmutableSet;
 import static io.trino.memory.context.AggregatedMemoryContext.newSimpleAggregatedMemoryContext;
@@ -88,7 +85,6 @@
 import static io.trino.plugin.hive.HiveColumnHandle.ColumnType.REGULAR;
 import static io.trino.plugin.hive.HiveErrorCode.HIVE_BAD_DATA;
 import static io.trino.plugin.hive.HiveErrorCode.HIVE_CANNOT_OPEN_SPLIT;
-import static io.trino.plugin.hive.HiveErrorCode.HIVE_MISSING_DATA;
 import static io.trino.plugin.hive.HivePageSourceProvider.projectBaseColumns;
 import static io.trino.plugin.hive.HivePageSourceProvider.projectSufficientColumns;
 import static io.trino.plugin.hive.HiveSessionProperties.getParquetMaxReadBlockRowCount;
@@ -315,14 +311,7 @@ && predicateMatches(
             if (e instanceof ParquetCorruptionException) {
                 throw new TrinoException(HIVE_BAD_DATA, e);
             }
-            if (nullToEmpty(e.getMessage()).trim().equals("Filesystem closed") ||
-                    e instanceof FileNotFoundException) {
-                throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, e);
-            }
             String message = format("Error opening Hive split %s (offset=%s, length=%s): %s", inputFile.location(), start, length, e.getMessage());
-            if (e instanceof BlockMissingException) {
-                throw new TrinoException(HIVE_MISSING_DATA, message, e);
-            }
             throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, message, e);
         }
     }

diff --git a/plugin/trino-hive/src/main/java/io/trino/plugin/hive/rcfile/RcFilePageSourceFactory.java b/plugin/trino-hive/src/main/java/io/trino/plugin/hive/rcfile/RcFilePageSourceFactory.java
@@ -48,24 +48,20 @@
 import io.trino.spi.type.TypeManager;
 import org.apache.hadoop.conf.Configuration;
 import org.apache.hadoop.fs.Path;
-import org.apache.hadoop.hdfs.BlockMissingException;
 import org.joda.time.DateTimeZone;
 
 import javax.inject.Inject;
 
-import java.io.FileNotFoundException;
 import java.io.InputStream;
 import java.util.List;
 import java.util.Optional;
 import java.util.OptionalInt;
 import java.util.Properties;
 
 import static com.google.common.base.Preconditions.checkArgument;
-import static com.google.common.base.Strings.nullToEmpty;
 import static com.google.common.collect.ImmutableList.toImmutableList;
 import static io.trino.plugin.hive.HiveErrorCode.HIVE_BAD_DATA;
 import static io.trino.plugin.hive.HiveErrorCode.HIVE_CANNOT_OPEN_SPLIT;
-import static io.trino.plugin.hive.HiveErrorCode.HIVE_MISSING_DATA;
 import static io.trino.plugin.hive.HivePageSourceProvider.projectBaseColumns;
 import static io.trino.plugin.hive.HiveSessionProperties.getTimestampPrecision;
 import static io.trino.plugin.hive.ReaderPageSource.noProjectionAdaptation;
@@ -163,10 +159,6 @@ else if (deserializerClassName.equals(COLUMNAR_SERDE_CLASS)) {
             throw e;
         }
         catch (Exception e) {
-            if (nullToEmpty(e.getMessage()).trim().equals("Filesystem closed") ||
-                    e instanceof FileNotFoundException) {
-                throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, e);
-            }
             throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, splitError(e, path, start, length), e);
         }
 
@@ -192,17 +184,14 @@ else if (deserializerClassName.equals(COLUMNAR_SERDE_CLASS)) {
             ConnectorPageSource pageSource = new RcFilePageSource(rcFileReader, projectedReaderColumns);
             return Optional.of(new ReaderPageSource(pageSource, readerProjections));
         }
+        catch (TrinoException e) {
+            throw e;
+        }
         catch (Throwable e) {
-            if (e instanceof TrinoException) {
-                throw (TrinoException) e;
-            }
             String message = splitError(e, path, start, length);
             if (e instanceof FileCorruptionException) {
                 throw new TrinoException(HIVE_BAD_DATA, message, e);
             }
-            if (e instanceof BlockMissingException) {
-                throw new TrinoException(HIVE_MISSING_DATA, message, e);
-            }
             throw new TrinoException(HIVE_CANNOT_OPEN_SPLIT, message, e);
         }
     }

diff --git a/plugin/trino-hudi/src/main/java/io/trino/plugin/hudi/HudiErrorCode.java b/plugin/trino-hudi/src/main/java/io/trino/plugin/hudi/HudiErrorCode.java
@@ -25,7 +25,7 @@ public enum HudiErrorCode
     HUDI_UNKNOWN_TABLE_TYPE(0, EXTERNAL),
     HUDI_INVALID_PARTITION_VALUE(1, EXTERNAL),
     HUDI_BAD_DATA(2, EXTERNAL),
-    HUDI_MISSING_DATA(3, EXTERNAL),
+    // HUDI_MISSING_DATA(3, EXTERNAL) is deprecated
     HUDI_CANNOT_OPEN_SPLIT(4, EXTERNAL),
     HUDI_UNSUPPORTED_FILE_FORMAT(5, EXTERNAL),
     HUDI_CURSOR_ERROR(6, EXTERNAL);

diff --git a/plugin/trino-hudi/src/main/java/io/trino/plugin/hudi/HudiPageSourceProvider.java b/plugin/trino-hudi/src/main/java/io/trino/plugin/hudi/HudiPageSourceProvider.java
@@ -46,7 +46,6 @@
 import io.trino.spi.type.Decimals;
 import io.trino.spi.type.TypeSignature;
 import org.apache.hadoop.fs.Path;
-import org.apache.hadoop.hdfs.BlockMissingException;
 import org.apache.hudi.common.model.HoodieFileFormat;
 import org.apache.parquet.column.ColumnDescriptor;
 import org.apache.parquet.hadoop.metadata.BlockMetaData;
@@ -89,7 +88,6 @@
 import static io.trino.plugin.hudi.HudiErrorCode.HUDI_CANNOT_OPEN_SPLIT;
 import static io.trino.plugin.hudi.HudiErrorCode.HUDI_CURSOR_ERROR;
 import static io.trino.plugin.hudi.HudiErrorCode.HUDI_INVALID_PARTITION_VALUE;
-import static io.trino.plugin.hudi.HudiErrorCode.HUDI_MISSING_DATA;
 import static io.trino.plugin.hudi.HudiErrorCode.HUDI_UNSUPPORTED_FILE_FORMAT;
 import static io.trino.plugin.hudi.HudiSessionProperties.isParquetOptimizedNestedReaderEnabled;
 import static io.trino.plugin.hudi.HudiSessionProperties.isParquetOptimizedReaderEnabled;
@@ -264,16 +262,10 @@ && predicateMatches(parquetPredicate, block, dataSource, descriptorsByPath, parq
             if (e instanceof TrinoException) {
                 throw (TrinoException) e;
             }
-            String message = format("Error opening Hudi split %s (offset=%s, length=%s): %s",
-                    path, start, length, e.getMessage());
-
             if (e instanceof ParquetCorruptionException) {
-                throw new TrinoException(HUDI_BAD_DATA, message, e);
-            }
-
-            if (e instanceof BlockMissingException) {
-                throw new TrinoException(HUDI_MISSING_DATA, message, e);
+                throw new TrinoException(HUDI_BAD_DATA, e);
             }
+            String message = "Error opening Hudi split %s (offset=%s, length=%s): %s".formatted(path, start, length, e.getMessage());
             throw new TrinoException(HUDI_CANNOT_OPEN_SPLIT, message, e);
         }
     }

diff --git a/plugin/trino-iceberg/src/main/java/io/trino/plugin/iceberg/IcebergErrorCode.java b/plugin/trino-iceberg/src/main/java/io/trino/plugin/iceberg/IcebergErrorCode.java
@@ -29,7 +29,7 @@ public enum IcebergErrorCode
     ICEBERG_TOO_MANY_OPEN_PARTITIONS(2, USER_ERROR),
     ICEBERG_INVALID_PARTITION_VALUE(3, EXTERNAL),
     ICEBERG_BAD_DATA(4, EXTERNAL),
-    ICEBERG_MISSING_DATA(5, EXTERNAL),
+    // ICEBERG_MISSING_DATA(5, EXTERNAL) is deprecated
     ICEBERG_CANNOT_OPEN_SPLIT(6, EXTERNAL),
     ICEBERG_WRITER_OPEN_ERROR(7, EXTERNAL),
     ICEBERG_FILESYSTEM_ERROR(8, EXTERNAL),