优化07版超大文件读取方案

README.md

@@ -8,6 +8,11 @@ easyexcel
 
 # JAVA解析Excel工具easyexcel
 Java解析、生成Excel比较有名的框架有Apache poi、jxl。但他们都存在一个严重的问题就是非常的耗内存，poi有一套SAX模式的API可以一定程度的解决一些内存溢出的问题，但POI还是有一些缺陷，比如07版Excel解压缩以及解压后存储都是在内存中完成的，内存消耗依然很大。easyexcel重写了poi对07版Excel的解析，能够原本一个3M的excel用POI sax依然需要100M左右内存降低到几M，并且再大的excel不会出现内存溢出，03版依赖POI的sax模式。在上层做了模型转换的封装，让使用者更加简单方便
+
+## 64M内存1分钟内读取75M(46W行25列)的Excel
+当然还有急速模式能更快，但是内存占用会在100M多一点
+![img](img/readme/large.png)
+
 ## 相关文档
 * [快速使用](/quickstart.md)
 * [关于软件](/abouteasyexcel.md)

src/main/java/com/alibaba/excel/context/WriteContextImpl.java

@@ -10,6 +10,7 @@ import org.apache.poi.ss.usermodel.Sheet;
 import org.apache.poi.ss.usermodel.Workbook;
 import org.apache.poi.ss.util.CellRangeAddress;
 import org.apache.poi.xssf.streaming.SXSSFWorkbook;
+import org.apache.poi.xssf.usermodel.XSSFSheet;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
@@ -181,6 +182,16 @@ public class WriteContextImpl implements WriteContext {
     }
 
     private void initSheet() {
+        try {
+            if (writeWorkbookHolder.getXssfWorkbook() != null) {
+                writeSheetHolder
+                    .setXssfSheet(writeWorkbookHolder.getXssfWorkbook().getSheetAt(writeSheetHolder.getSheetNo()));
+            }
+        } catch (Exception e) {
+            if (LOGGER.isDebugEnabled()) {
+                LOGGER.debug("Can not find XSSFSheet:{}.", writeSheetHolder.getSheetNo());
+            }
+        }
         Sheet currentSheet;
         try {
             currentSheet = writeWorkbookHolder.getWorkbook().getSheetAt(writeSheetHolder.getSheetNo());

src/main/java/com/alibaba/excel/write/ExcelBuilderImpl.java

@@ -19,6 +19,7 @@ import org.apache.poi.ss.usermodel.Drawing;
 import org.apache.poi.ss.usermodel.Row;
 import org.apache.poi.ss.usermodel.Sheet;
 import org.apache.poi.ss.util.CellRangeAddress;
+import org.apache.poi.xssf.usermodel.XSSFSheet;
 
 import com.alibaba.excel.context.WriteContext;
 import com.alibaba.excel.context.WriteContextImpl;
@@ -129,14 +130,14 @@ public class ExcelBuilderImpl implements ExcelBuilder {
 
     private void doFill(Object data) {
         WriteSheetHolder writeSheetHolder = context.writeSheetHolder();
-        Sheet sheet = writeSheetHolder.getSheet();
+        XSSFSheet sheet = writeSheetHolder.getXssfSheet();
         Map<Integer, Integer> templateLastRowMap = context.writeWorkbookHolder().getTemplateLastRowMap();
         if (sheet == null) {
             throw new ExcelGenerateException(
                 "The corresponding table cannot be found,sheetNo:" + writeSheetHolder.getSheetNo());
         }
         List<AnalysisCell> analysisCellList = new ArrayList<AnalysisCell>();
-        for (int i = 0; i < templateLastRowMap.get(writeSheetHolder.getSheetNo()); i++) {
+        for (int i = 0; i <= sheet.getLastRowNum(); i++) {
             Row row = sheet.getRow(i);
             for (int j = 0; j < row.getLastCellNum(); j++) {
                 Cell cell = row.getCell(j);
@@ -151,7 +152,7 @@ public class ExcelBuilderImpl implements ExcelBuilder {
                     int index = 0;
                     while (matches) {
                         Matcher matcher = FILL_PATTERN.matcher(value);
-                        String variable = value.substring(matcher.start(), matcher.end());
+                        String variable = value.substring(matcher.regionStart() + 2, matcher.regionEnd() - 1);
                         variableList.add(variable);
                         value = matcher.replaceFirst("{" + index++ + "}");
                         matches = FILL_PATTERN.matcher(value).matches();

src/main/java/com/alibaba/excel/write/metadata/holder/WriteSheetHolder.java

@@ -5,7 +5,6 @@ import java.util.Map;
 
 import org.apache.poi.ss.usermodel.Sheet;
 import org.apache.poi.xssf.usermodel.XSSFSheet;
-import org.apache.poi.xssf.usermodel.XSSFWorkbook;
 
 import com.alibaba.excel.enums.HolderEnum;
 import com.alibaba.excel.enums.WriteLastRowType;

update.md

@@ -1,5 +1,4 @@
 # 2.1.0-beta1
-* 降级poi为3.1.7 兼容jdk6
 * 新增支持导入、导出支持公式
 * 新增支持读取单元格类型、写入指定单元格类型
 * 支持通过模板填充数据