Merge pull request #2880 from alibaba/developing

* 提高xlsx读取兼容性：在存在第一行很多空列的情况下，忽略空列

README.md

@@ -36,7 +36,7 @@ Java解析、生成Excel比较有名的框架有Apache poi、jxl。但他们都
         <dependency>
             <groupId>com.alibaba</groupId>
             <artifactId>easyexcel</artifactId>
-            <version>3.1.3</version>
+            <version>3.1.4</version>
         </dependency>
 ```
 

easyexcel-core/src/main/java/com/alibaba/excel/read/listener/ModelBuildEventListener.java

@@ -25,14 +25,6 @@
  */
 public class ModelBuildEventListener implements IgnoreExceptionReadListener<Map<Integer, ReadCellData<?>>> {
 
-    @Override
-    public void invokeHead(Map<Integer, ReadCellData<?>> cellDataMap, AnalysisContext context) {
-        if (context.readSheetHolder().getMaxDataHeadSize() == null
-            || context.readSheetHolder().getMaxDataHeadSize() < CollectionUtils.size(cellDataMap)) {
-            context.readSheetHolder().setMaxDataHeadSize(CollectionUtils.size(cellDataMap));
-        }
-    }
-
     @Override
     public void invoke(Map<Integer, ReadCellData<?>> cellDataMap, AnalysisContext context) {
         ReadSheetHolder readSheetHolder = context.readSheetHolder();
@@ -73,8 +65,8 @@ private int calculateHeadSize(ReadSheetHolder readSheetHolder) {
         if (readSheetHolder.excelReadHeadProperty().getHeadMap().size() > 0) {
             return readSheetHolder.excelReadHeadProperty().getHeadMap().size();
         }
-        if (readSheetHolder.getMaxDataHeadSize() != null) {
-            return readSheetHolder.getMaxDataHeadSize();
+        if (readSheetHolder.getMaxNotEmptyDataHeadSize() != null) {
+            return readSheetHolder.getMaxNotEmptyDataHeadSize();
         }
         return 0;
     }

easyexcel-core/src/main/java/com/alibaba/excel/read/metadata/holder/ReadSheetHolder.java

@@ -65,7 +65,7 @@ public class ReadSheetHolder extends AbstractReadHolder {
      * Read the size of the largest head in sheet head data.
      * see https://github.com/alibaba/easyexcel/issues/2014
      */
-    private Integer maxDataHeadSize;
+    private Integer maxNotEmptyDataHeadSize;
 
     public ReadSheetHolder(ReadSheet readSheet, ReadWorkbookHolder readWorkbookHolder) {
         super(readSheet, readWorkbookHolder);

easyexcel-core/src/main/java/com/alibaba/excel/read/processor/DefaultAnalysisEventProcessor.java

@@ -3,8 +3,11 @@
 import java.util.HashMap;
 import java.util.List;
 import java.util.Map;
+import java.util.Map.Entry;
+import java.util.stream.Collectors;
 
 import com.alibaba.excel.context.AnalysisContext;
+import com.alibaba.excel.enums.CellDataTypeEnum;
 import com.alibaba.excel.enums.HeadKindEnum;
 import com.alibaba.excel.enums.RowTypeEnum;
 import com.alibaba.excel.exception.ExcelAnalysisException;
@@ -17,6 +20,8 @@
 import com.alibaba.excel.util.ConverterUtils;
 import com.alibaba.excel.util.StringUtils;
 
+import org.apache.commons.collections4.CollectionUtils;
+import org.apache.commons.collections4.MapUtils;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
@@ -111,6 +116,14 @@ private void dealData(AnalysisContext analysisContext) {
     }
 
     private void buildHead(AnalysisContext analysisContext, Map<Integer, ReadCellData<?>> cellDataMap) {
+        // Rule out empty head, and then take the largest column
+        if (MapUtils.isNotEmpty(cellDataMap)) {
+            cellDataMap.entrySet()
+                .stream()
+                .filter(entry -> CellDataTypeEnum.EMPTY != entry.getValue().getType())
+                .forEach(entry -> analysisContext.readSheetHolder().setMaxNotEmptyDataHeadSize(entry.getKey()));
+        }
+
         if (!HeadKindEnum.CLASS.equals(analysisContext.currentReadHolder().excelReadHeadProperty().getHeadKind())) {
             return;
         }

easyexcel-test/src/test/java/com/alibaba/easyexcel/test/core/compatibility/CompatibilityTest.java

@@ -42,4 +42,14 @@ public void t02() {
         Map<Integer, Object> row2 = list.get(2);
         Assert.assertEquals("1，2-戊二醇", row2.get(2));
     }
+
+    @Test
+    public void t03() {
+        // Exist in `sharedStrings.xml` `x:t` start tag, need to be compatible
+        List<Map<Integer, Object>> list = EasyExcel.read(TestFileUtil.getPath() + "compatibility/t03.xlsx").sheet().doReadSync();
+        log.info("data:{}", JSON.toJSONString(list));
+        Assert.assertEquals(1, list.size());
+        Map<Integer, Object> row0 = list.get(0);
+        Assert.assertEquals(12, row0.size());
+    }
 }

pom.xml

@@ -20,7 +20,7 @@
 
 
     <properties>
-        <revision>3.1.3</revision>
+        <revision>3.1.4</revision>
         <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
         <jdk.version>1.8</jdk.version>
         <gpg.skip>true</gpg.skip>

update.md

@@ -1,3 +1,7 @@
+# 3.1.4
+
+* 提高xlsx读取兼容性：在存在第一行很多空列的情况下，忽略空列
+
 # 3.1.3
 
 * 提高xlsx兼容性：兼容`sharedStrings.xml` 存在 `x:t`标签的情况
@@ -156,7 +160,9 @@
 * 发布正式版
 * 修复第一行为空不会调用`invokeHeadMap`的bug [Issue #993](https://github.com/alibaba/easyexcel/issues/993)
 *
+
 当类的属性没有按照ExcelProperty的属性index顺序排序的时候，写数据出现错乱 [Issue #1046](https://github.com/alibaba/easyexcel/issues/1046)
+
 * 新增支持自定义转换器 入参可以为空 实现`NullableObjectConverter`
   即可  [Issue #1084](https://github.com/alibaba/easyexcel/issues/1084)
 * 修复xls丢失结束标记的情况下 会漏读最后一行