exceptionfactory commented on code in PR #9379:
URL: https://github.com/apache/nifi/pull/9379#discussion_r1801634966


##########
nifi-extension-bundles/nifi-poi-bundle/nifi-poi-services/src/main/java/org/apache/nifi/excel/StandardCellFieldTypeReader.java:
##########
@@ -0,0 +1,129 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.nifi.excel;
+
+import org.apache.nifi.schema.inference.FieldTypeInference;
+import org.apache.nifi.schema.inference.TimeValueInference;
+import org.apache.nifi.serialization.record.DataType;
+import org.apache.nifi.serialization.record.RecordFieldType;
+import org.apache.poi.ss.usermodel.Cell;
+import org.apache.poi.ss.usermodel.CellType;
+import org.apache.poi.ss.usermodel.DateUtil;
+
+import java.util.Map;
+import java.util.Objects;
+import java.util.Optional;
+
+/**
+ * Standard implementation of Cell Field Type Reader uses Cell Type and Cell 
Value information with inference based on Timestamp formats
+ */
+class StandardCellFieldTypeReader implements CellFieldTypeReader {
+    private final TimeValueInference timeValueInference;
+
+    StandardCellFieldTypeReader(final TimeValueInference timeValueInference) {
+        this.timeValueInference = Objects.requireNonNull(timeValueInference, 
"Time Value Inference required");
+    }
+
+    /**
+     * Infer Cell Field Type and append possible Data Types to mapped Field 
Type Inference information
+     *
+     * @param cell Spreadsheet Cell required
+     * @param fieldName Cell field name for tracking in Field Types
+     * @param fieldTypes Map of Field Name to Field Type Inference information
+     */
+    @Override
+    public void inferCellFieldType(final Cell cell, final String fieldName, 
final Map<String, FieldTypeInference> fieldTypes) {
+        Objects.requireNonNull(cell, "Cell required");
+        Objects.requireNonNull(fieldName, "Field Name required");
+        Objects.requireNonNull(fieldTypes, "Field Types required");
+
+        final FieldTypeInference fieldTypeInference = 
fieldTypes.computeIfAbsent(fieldName, key -> new FieldTypeInference());
+        final DataType dataType = getCellDataType(cell);
+        fieldTypeInference.addPossibleDataType(dataType);
+    }
+
+    /**
+     * Get Record Data Type from Spreadsheet Cell Type and additional 
resolution of NUMERIC and STRING types
+     *
+     * @param cell Spreadsheet Cell required
+     * @return Record Data Type
+     */
+    @Override
+    public DataType getCellDataType(final Cell cell) {
+        Objects.requireNonNull(cell, "Cell required");
+
+        final CellType cellType = cell.getCellType();
+
+        final DataType dataType;
+
+        if (CellType.NUMERIC == cellType) {
+            // Date Formatting check limited to NUMERIC Cell Types
+            if (DateUtil.isCellDateFormatted(cell)) {
+                // Default to TIMESTAMP for date formatted values using 
standard Date Time Pattern
+                dataType = RecordFieldType.TIMESTAMP.getDataType();
+            } else {
+                final double numericCellValue = cell.getNumericCellValue();
+                final long roundedCellValue = (long) numericCellValue;
+                if (roundedCellValue == numericCellValue) {
+                    dataType = RecordFieldType.LONG.getDataType();
+                } else {
+                    // Default to DOUBLE for NUMERIC values following 
cell.getNumericCellValue()
+                    dataType = RecordFieldType.DOUBLE.getDataType();
+                }
+            }
+        } else if (CellType.BOOLEAN == cellType) {
+            dataType = RecordFieldType.BOOLEAN.getDataType();
+        } else if (CellType.STRING == cellType) {
+            final String cellValue = cell.getStringCellValue();
+            // Attempt Time Value inference for STRING cell values
+            final Optional<DataType> timeDataType = 
timeValueInference.getDataType(cellValue);

Review Comment:
   Reviewing the current `additionalDetails.md` documentation, the description 
actually seems to reflect the behavior accurately:
   
   > If the "Time Format," "Timestamp Format," or "Date Format" properties are 
configured, any value that would otherwise be considered a STRING type is first 
checked against the configured formats to see if it matches any of them. If the 
value matches the Timestamp Format, the value is considered a Timestamp field. 
If it matches the Date Format, it is considered a Date field. If it matches the 
Time Format, it is considered a Time field. In the unlikely event that the 
value matches more than one of the configured formats, they will be matched in 
the order: Timestamp, Date, Time. I.e., if a value matched both the Timestamp 
Format and the Date Format, the type that is inferred will be Timestamp. 
Because parsing dates and times can be expensive, it is advisable not to 
configure these formats if dates, times, and timestamps are not expected, or if 
processing the data as a STRING is acceptable. For use cases when this is 
important, though, the inference engine is intelligent enough to opt
 imize the parsing by first checking several very cheap conditions. For 
example, the string's length is examined to see if it is too long or too short 
to match the pattern. This results in far more efficient processing than would 
result if attempting to parse each string value as a timestamp.
   
   I'm not sure how valuable it is to go into some additional implementation 
details at the moment, if future revisions are necessary. However, if you have 
a suggestion for an additional bullet point, glad to include.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to