Fix deletes which trigger file rewrites in partitions with special ch…

…aracters
trinodb · Oct 3, 2023 · 0cdb524 · 0cdb524
1 parent b02a15b
commit 0cdb524
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 5 deletions.
diff --git a/plugin/trino-delta-lake/src/main/java/io/trino/plugin/deltalake/DeltaLakeMetadata.java b/plugin/trino-delta-lake/src/main/java/io/trino/plugin/deltalake/DeltaLakeMetadata.java
@@ -1978,7 +1978,7 @@ public void finishMerge(ConnectorSession session, ConnectorMergeTableHandle merg
             }
 
             for (String file : oldFiles) {
-                transactionLogWriter.appendRemoveFileEntry(new RemoveFileEntry(file, writeTimestamp, true));
+                transactionLogWriter.appendRemoveFileEntry(new RemoveFileEntry(toUriFormat(file), writeTimestamp, true));
             }
 
             appendAddFileEntries(transactionLogWriter, newFiles, partitionColumns, getExactColumnNames(handle.getMetadataEntry()), true);
@@ -2169,7 +2169,7 @@ private void finishOptimize(ConnectorSession session, DeltaLakeTableExecuteHandl
 
             for (String scannedPath : scannedPaths) {
                 String relativePath = relativePath(tableLocation, scannedPath);
-                transactionLogWriter.appendRemoveFileEntry(new RemoveFileEntry(relativePath, writeTimestamp, false));
+                transactionLogWriter.appendRemoveFileEntry(new RemoveFileEntry(toUriFormat(relativePath), writeTimestamp, false));
             }
 
             // Note: during writes we want to preserve original case of partition columns
@@ -2632,7 +2632,7 @@ static String relativePath(String basePath, String path)
         String basePathDirectory = basePath.endsWith("/") ? basePath : basePath + "/";
         checkArgument(path.startsWith(basePathDirectory) && (path.length() > basePathDirectory.length()),
                 "path [%s] must be a subdirectory of basePath [%s]", path, basePath);
-        return toUriFormat(path.substring(basePathDirectory.length()));
+        return path.substring(basePathDirectory.length());
     }
 
     public void rollback()

diff --git a/plugin/trino-delta-lake/src/test/java/io/trino/plugin/deltalake/TestDeltaLakeDelete.java b/plugin/trino-delta-lake/src/test/java/io/trino/plugin/deltalake/TestDeltaLakeDelete.java
@@ -67,9 +67,13 @@ public void testTargetedDeleteWhenTableIsPartitionedWithColumnContainingSpecialC
                 "AS VALUES " +
                 "(1, 'with-hyphen'), " +
                 "(2, 'with:colon'), " +
-                "(3, 'with?question')", 3);
+                "(3, 'with:colon'), " + // create two rows in a single file to trigger parquet file rewrite on delete
+                "(4, 'with?question')", 4);
+        assertQuery("SELECT count(*), count(DISTINCT \"$path\"), col_name FROM " + tableName + " GROUP BY 3", "VALUES (1, 1, 'with-hyphen'), (2, 1, 'with:colon'), (1, 1, 'with?question')");
         assertUpdate("DELETE FROM " + tableName + " WHERE id = 2", 1);
-        assertQuery("SELECT * FROM " + tableName, "VALUES(1, 'with-hyphen'), (3, 'with?question')");
+        assertQuery("SELECT * FROM " + tableName, "VALUES (1, 'with-hyphen'), (3, 'with:colon'), (4, 'with?question')");
+        assertUpdate("DELETE FROM " + tableName, 3);
+        assertQueryReturnsEmptyResult("SELECT * FROM " + tableName);
     }
 
     @Test