Throw error before AIOUB

cloudsufi · Dec 5, 2023 · 3984b8e · 3984b8e
1 parent 55af0b7
commit 3984b8e
Show file tree

Hide file tree

Showing 3 changed files with 17 additions and 5 deletions.
diff --git a/format-xls/src/main/java/io/cdap/plugin/format/xls/input/XlsInputFormat.java b/format-xls/src/main/java/io/cdap/plugin/format/xls/input/XlsInputFormat.java
@@ -118,7 +118,7 @@ public void initialize(InputSplit genericSplit, TaskAttemptContext context) thro
       rowIndex = 0;
       isRowNull = false;
 
-      boolean skipFirstRow = job.getBoolean("skipHeader", true);
+      boolean skipFirstRow = job.getBoolean(NAME_SKIP_HEADER, true);
       if (skipFirstRow) {
         Preconditions.checkArgument(lastRowNum != -1, "No rows found on sheet %s", sheetValue);
         rowIndex = 1;
@@ -140,16 +140,23 @@ public boolean nextKeyValue() {
 
       isRowNull = true;
       for (int cellIndex = 0; cellIndex < row.getLastCellNum(); cellIndex++) {
+        if (cellIndex >= fields.size()) {
+         throw new IllegalArgumentException(
+           String.format("Schema contains less fields than the number of columns in the excel file. " +
+                           "Schema fields: %s, Excel columns: %s", fields.size(), row.getLastCellNum()));
+        }
         Cell cell = row.getCell(cellIndex, Row.MissingCellPolicy.RETURN_BLANK_AS_NULL);
         if (cell == null) {
           continue;
         }
-        isRowNull = false;
         Schema.Field field = fields.get(cellIndex);
-        String result = formatter.formatCellValue(cell);
+        Schema.Type type = field.getSchema().isNullable() ?
+                field.getSchema().getNonNullable().getType() : field.getSchema().getType();
+        String result = formatter.formatCellValue(cell, type);
         if (result == null) {
           continue;
         }
+        isRowNull = false;
         builder.convertAndSet(field.getName(), result);
       }
       value = builder.build();

diff --git a/format-xls/src/main/java/io/cdap/plugin/format/xls/input/XlsInputFormatConfig.java b/format-xls/src/main/java/io/cdap/plugin/format/xls/input/XlsInputFormatConfig.java
@@ -90,7 +90,7 @@ public class XlsInputFormatConfig extends PathTrackingConfig {
   @Nullable
   @Name(NAME_TERMINATE_IF_EMPTY_ROW)
   @Description(DESC_TERMINATE_ROW)
-  private static Boolean terminateIfEmptyRow;
+  private Boolean terminateIfEmptyRow;
 
   public XlsInputFormatConfig() {
     super();

diff --git a/format-xls/src/main/java/io/cdap/plugin/format/xls/input/XlsInputFormatDataFormatter.java b/format-xls/src/main/java/io/cdap/plugin/format/xls/input/XlsInputFormatDataFormatter.java
@@ -45,7 +45,7 @@ public XlsInputFormatDataFormatter(FormulaEvaluator evaluator) {
    * @param cell the cell to format
    * @return the formatted cell value
    */
-  public String formatCellValue(Cell cell) {
+  public String formatCellValue(Cell cell, Schema.Type type) {
     if (cell == null) {
       return null;
     }
@@ -66,6 +66,11 @@ public String formatCellValue(Cell cell) {
         }
         return Double.toString(cell.getNumericCellValue());
       case STRING:
+        if (type == Schema.Type.DOUBLE) {
+          // Edge case when schema was inferred as double but the cell is actually a string
+          // this can be caused by an error formula cell, as the error value is stored as a string
+          return null;
+        }
         return cell.getRichStringCellValue().getString();
       case BOOLEAN:
         return cell.getBooleanCellValue() ? "TRUE" : "FALSE";