姓名解析新算法

功能完善
POI升级版本

pom.xml

@@ -47,14 +47,14 @@
         <dependency>
             <groupId>org.apache.poi</groupId>
             <artifactId>poi</artifactId>
-            <version>3.9</version>
+            <version>4.1.2</version>
         </dependency>
 
         <!--xlsx(07)，07版本的-->
         <dependency>
             <groupId>org.apache.poi</groupId>
             <artifactId>poi-ooxml</artifactId>
-            <version>3.9</version>
+            <version>4.1.2</version>
         </dependency>
 
         <dependency>

src/main/java/com/gmh/controller/ExcelController.java

@@ -1,24 +1,31 @@
 package com.gmh.controller;
 
-import cn.hutool.core.util.StrUtil;
+import com.baomidou.mybatisplus.core.conditions.query.QueryWrapper;
 import com.gmh.entity.GmhUser;
 import com.gmh.entity.R;
+import com.gmh.entity.SysBaijiaxing;
+import com.gmh.entity.vo.ReadNameVo;
+import com.gmh.service.SysBaijiaxingService;
 import com.gmh.utils.POIExcelUtil;
-import org.apache.poi.ss.usermodel.Cell;
-import org.apache.poi.ss.usermodel.Row;
 import org.apache.poi.ss.usermodel.Sheet;
 import org.apache.poi.ss.usermodel.Workbook;
+import org.springframework.beans.factory.annotation.Autowired;
 import org.springframework.web.bind.annotation.RequestMapping;
 import org.springframework.web.bind.annotation.RestController;
 import org.springframework.web.multipart.MultipartFile;
 
 import java.io.IOException;
 import java.util.*;
+import java.util.stream.Collectors;
 
 @RestController
 @RequestMapping("/excel")
 public class ExcelController {
 
+    public final ThreadLocal<Workbook> workbookThreadLocal = new ThreadLocal<>();
+
+    private static final String KEY_XINGMING = "姓名";
+
     @RequestMapping("/t01")
     public R test01() {
         return R.ok("ok");
@@ -27,58 +34,139 @@ public class ExcelController {
     @RequestMapping("/readData")
     public R readSourceData(MultipartFile file) throws IOException {
         Workbook workbook = POIExcelUtil.readExcelFromInputStream(file.getInputStream(), file.getOriginalFilename());
-        Integer nameCellIndex = null;
+        List<ReadNameVo> resultList = new ArrayList<>();
-        Map<String, List<GmhUser>> sheetNameList = new LinkedHashMap<>();
         for (int i = 0; i < workbook.getNumberOfSheets(); i++) {
-            List<GmhUser> nameList = new ArrayList<>();
+            ReadNameVo readNameVo = new ReadNameVo();
             Sheet sheet = workbook.getSheetAt(i);
-            for (int r = 0; r <= sheet.getLastRowNum(); r++) {
+            List<Map<String, String>> excelMaps = POIExcelUtil.toListMap(sheet);
-                Row row = sheet.getRow(r);
+            if (excelMaps.isEmpty()) {
-                if (row == null) {
+                continue;
-                    continue;
+            }
+            Map<String, String> firstElement = excelMaps.get(0);
+            String xmKey = KEY_XINGMING;
+            for (String key : firstElement.keySet()) {
+                if (KEY_XINGMING.equals(firstElement.get(key))) {
+                    xmKey = key;
                 }
-                int lastCellNum = row.getLastCellNum();
+            }
-                for (int c = 0; c < lastCellNum; c++) {
+            if (firstElement.containsKey(xmKey)) {
-                    Cell cell = row.getCell(c);
+                if (!KEY_XINGMING.equals(xmKey)) {
-                    if (cell == null) {
+                    excelMaps.remove(firstElement);
-                        continue;
-                    }
-                    String stringCellValue = POIExcelUtil.getStringCellValExcludeBlank(cell);
-                    if (r == 0) {
-                        // 第一行有姓名列
-                        if ("姓名".equals(stringCellValue)) {
-                            nameCellIndex = c;
-                            break;
-                        } else {
-                            // 第一列没有姓名的情况
-                        }
-                    }
                 }
-                if (nameCellIndex != null) {
+                readNameVo = getReadNameDataForNameKey(excelMaps, xmKey);
-                    Cell cell = row.getCell(nameCellIndex);
+            } else {
-                    String stringCellValue = POIExcelUtil.getStringCellValExcludeBlank(cell);
+                Map<String, Integer> map = new HashMap<>();
-                    if (!"姓名".equals(stringCellValue) && StrUtil.isNotBlank(stringCellValue)) {
+                for (Map<String, String> excelMap : excelMaps) {
-                        GmhUser user = new GmhUser();
+                    for (String key : excelMap.keySet()) {
-                        user.setName(stringCellValue);
+                        String value = excelMap.get(key);
-                        nameList.add(user);
+                        if (isChinaName(value)) {
+                            Integer count = map.getOrDefault(key, 0);
+                            map.put(key, ++count);
+                        }
                     }
-                } else {
-                    System.out.println("未找到姓名列");
                 }
+                List<Map.Entry<String,Integer>> list = new ArrayList<>(map.entrySet());
+//                list.sort(Comparator.comparingInt(Map.Entry::getValue));        //升序
+                list.sort((o1, o2) -> (o2.getValue() - o1.getValue()));
+                String key = list.get(0).getKey();
+                readNameVo = getReadNameDataForNameKey(excelMaps, key);
             }
-            if (!nameList.isEmpty()) {
+            readNameVo.setSheetName(sheet.getSheetName());
-                sheetNameList.put(sheet.getSheetName(), nameList);
+            resultList.add(readNameVo);
-            }
+        }
+        return R.ok().setData(resultList);
+    }
+
+    public ReadNameVo getReadNameDataForNameKey(List<Map<String, String>> excelMaps, final String key) {
+        ReadNameVo readNameVo = new ReadNameVo();
+        // 去重前
+        List<GmhUser> oldNameList = excelMaps.stream().map(map -> new GmhUser(map.get(key))).collect(Collectors.toList());
+        // 去重后
+        LinkedHashSet<GmhUser> newNameList = new LinkedHashSet<>(oldNameList);
+        readNameVo.setNameList(newNameList);
+        // 重复检测
+        int repeatCount = oldNameList.size() - newNameList.size();
+        if (repeatCount > 0) {
+            readNameVo.setHasRepeat(true);
+            readNameVo.setRepeatCount(repeatCount);
+        } else {
+            readNameVo.setHasRepeat(false);
+            readNameVo.setRepeatCount(0);
         }
-        return R.ok().setData(sheetNameList);
+        return readNameVo;
     }
 
     @RequestMapping("/template")
     public R templateUpload(MultipartFile file) throws IOException {
         Workbook workbook = POIExcelUtil.readExcelFromInputStream(file.getInputStream(), file.getOriginalFilename());
+        List<Map<String, String>> maps = POIExcelUtil.toListMap(workbook.getSheetAt(0));
+        workbookThreadLocal.set(workbook);
+        return R.ok().setData(maps);
+    }
+
+    @Autowired
+    private SysBaijiaxingService baijiaxingService;
 
-        return R.ok();
+    private boolean isChinaName(String val) {
+        QueryWrapper<SysBaijiaxing> queryWrapper = new QueryWrapper<>();
+        queryWrapper.apply("{0} LIKE CONCAT(xingshi,'%')", val);
+        List<SysBaijiaxing> list = baijiaxingService.list(queryWrapper);
+        return !list.isEmpty();
     }
 
+    public void oleCode() {
+//        Integer nameCellIndex = null;
+//        // Sheet:nameList
+//        List<ReadNameVo> readNameList = new ArrayList<>();
+//        for (int i = 0; i < workbook.getNumberOfSheets(); i++) {
+//            List<GmhUser> nameList = new ArrayList<>();
+//            Sheet sheet = workbook.getSheetAt(i);
+//            for (int r = 0; r <= sheet.getLastRowNum(); r++) {
+//                Row row = sheet.getRow(r);
+//                if (row == null) {
+//                    continue;
+//                }
+//                List<List<String>> cellList = new ArrayList<>(row.getLastCellNum());
+//                for (int c = 0; c < row.getLastCellNum(); c++) {
+//                    Cell cell = row.getCell(c);
+//                    if (cell == null) {
+//                        continue;
+//                    }
+//                    String colString = CellReference.convertNumToColString(cell.getColumnIndex());
+//                    System.out.println(colString);
+//
+//
+//                    String stringCellValue = POIExcelUtil.getStringCellValExcludeBlank(cell);
+//                    if (r == 0) {
+//                        // 第一行有姓名列
+//                        if ("姓名".equals(stringCellValue)) {
+//                            nameCellIndex = c;
+//                            break;
+//                        } else {
+//                            // 第一行没有指明姓名列
+//                        }
+//                    }
+//                }
+//                if (nameCellIndex != null) {
+//                    Cell cell = row.getCell(nameCellIndex);
+//                    String stringCellValue = POIExcelUtil.getStringCellValExcludeBlank(cell);
+//                    if (!"姓名".equals(stringCellValue) && StrUtil.isNotBlank(stringCellValue)) {
+//                        GmhUser user = new GmhUser();
+//                        user.setName(stringCellValue);
+//                        nameList.add(user);
+//                    }
+//                } else {
+//                    System.out.println("未找到姓名列");
+//                }
+//            }
+//            if (!nameList.isEmpty()) {
+//                ReadNameVo readNameVo = new ReadNameVo();
+//                readNameVo.setSheetName(sheet.getSheetName());
+//                readNameVo.setNameList(nameList);
+//                readNameList.add(readNameVo);
+//            }
+//        }
+//        return R.ok().setData(readNameList);
+    }
 
 }

src/main/java/com/gmh/entity/GmhUser.java

@@ -15,4 +15,10 @@ public class GmhUser {
 
     private String name;
 
+    public GmhUser() {
+    }
+
+    public GmhUser(String name) {
+        this.name = name;
+    }
 }

src/main/java/com/gmh/entity/SysBaijiaxing.java

@@ -7,7 +7,7 @@ import lombok.Data;
 import static com.baomidou.mybatisplus.annotation.IdType.AUTO;
 
 @Data
-@TableName("sys_firstname")
+@TableName("sys_baijiaxing")
 public class SysBaijiaxing {
 
     @TableId(type = AUTO)

src/main/java/com/gmh/entity/vo/ReadNameVo.java

@@ -0,0 +1,20 @@
+package com.gmh.entity.vo;
+
+import com.gmh.entity.GmhUser;
+import lombok.Data;
+
+import java.util.LinkedHashSet;
+
+@Data
+public class ReadNameVo {
+
+    private String sheetName;
+
+    private LinkedHashSet<GmhUser> nameList;
+
+    // 是否有重复
+    private Boolean hasRepeat;
+
+    // 重复个数
+    private Integer repeatCount;
+}

src/main/java/com/gmh/utils/POIExcelUtil.java

@@ -1,14 +1,16 @@
 package com.gmh.utils;
 
 import cn.hutool.core.util.StrUtil;
+import cn.hutool.poi.excel.ExcelFileUtil;
+import com.sun.org.apache.bcel.internal.generic.IF_ACMPEQ;
 import org.apache.poi.hssf.usermodel.HSSFWorkbook;
-import org.apache.poi.ss.usermodel.Cell;
+import org.apache.poi.ss.usermodel.*;
-import org.apache.poi.ss.usermodel.Sheet;
+import org.apache.poi.ss.util.CellReference;
-import org.apache.poi.ss.usermodel.Workbook;
 import org.apache.poi.xssf.usermodel.XSSFWorkbook;
 
 import java.io.*;
 import java.nio.file.Files;
+import java.util.*;
 
 /**
  * POI Excel工具类
@@ -52,9 +54,9 @@ public class POIExcelUtil {
     /**
      * 复制多个sheet
      *
-     * @param workbook  workbook
+     * @param workbook workbook
-     * @param index     起始sheet
+     * @param index    起始sheet
-     * @param total     复制多个sheet
+     * @param total    复制多个sheet
      */
     public static Workbook batchCloneSheet(Workbook workbook, Integer index, Integer total) throws IOException {
         for (int i = 0; i < total; i++) {
@@ -80,18 +82,28 @@ public class POIExcelUtil {
         return readExcelFromInputStream(Files.newInputStream(file.toPath()), file.getName());
     }
 
+    public static Workbook readExcelFromInputStream(InputStream inputStream) throws IOException {
+        return readExcelFromInputStream(inputStream, null);
+    }
+
     public static Workbook readExcelFromInputStream(InputStream inputStream, String fileName) throws IOException {
-        if (StrUtil.isBlank(fileName)) {
+        if (StrUtil.isNotBlank(fileName)) {
-            throw new RuntimeException("文件格式错误");
+            String[] split = fileName.split("\\.");
-        }
+            if (split.length == 0 || split.length == 1) {
-        String[] split = fileName.split("\\.");
+                throw new RuntimeException("文件格式错误");
-        if (split.length == 0 || split.length == 1) {
+            }
-            throw new RuntimeException("文件格式错误");
+            String suffix = split[split.length - 1];
+            if ("xls".equals(suffix)) {
+                return new HSSFWorkbook(inputStream);
+            } else if ("xlsx".equals(suffix)) {
+                return new XSSFWorkbook(inputStream);
+            } else {
+                throw new RuntimeException("文件格式错误");
+            }
         }
-        String suffix = split[split.length - 1];
+        if (ExcelFileUtil.isXls(inputStream)) {
-        if ("xls".equals(suffix)) {
             return new HSSFWorkbook(inputStream);
-        } else if ("xlsx".equals(suffix)) {
+        } else if (ExcelFileUtil.isXlsx(inputStream)) {
             return new XSSFWorkbook(inputStream);
         } else {
             throw new RuntimeException("文件格式错误");
@@ -99,7 +111,101 @@ public class POIExcelUtil {
     }
 
     public static String getStringCellValExcludeBlank(Cell cell) {
-        cell.setCellType(Cell.CELL_TYPE_STRING);
+        return getStringCellVal(cell).replaceAll("\\s*", "");
-        return cell.getStringCellValue().replaceAll("\\s*", "");
+    }
+
+    public static String getStringCellVal(Cell cell) {
+        DataFormatter dataFormatter = new DataFormatter();
+        return dataFormatter.formatCellValue(cell);
+    }
+
+    /**
+     * Excel 转换为List<Map<String, String>>对象
+     *
+     * @param workbook    workbook
+     * @param sheetIndex  当前sheet页索引
+     * @return List<Map<String, String>>
+     */
+    public static List<Map<String, String>> toListMap(Workbook workbook, int sheetIndex) {
+        return toListMap(workbook, sheetIndex, -1);
+    }
+
+    /**
+     * Excel 转换为List<Map<String, String>>对象
+     *
+     * @param workbook    workbook
+     * @param sheetIndex  当前sheet页索引
+     * @param titleRowNum 标题行索引，索引值<0则代表没有标题行
+     * @return List<Map<String, String>>
+     */
+    public static List<Map<String, String>> toListMap(Workbook workbook, int sheetIndex, int titleRowNum) {
+        return toListMap(workbook.getSheetAt(sheetIndex), titleRowNum);
+    }
+
+    /**
+     * Excel 转换为List<Map<String, String>>对象
+     *
+     * @param sheet    workbook
+     * @return List<Map<String, String>>
+     */
+    public static List<Map<String, String>> toListMap(Sheet sheet) {
+        return toListMap(sheet, -1);
+    }
+
+    /**
+     * Excel 转换为List<Map<String, String>>对象
+     *
+     * @param sheet  当前sheet页对象
+     * @param titleRowNum 标题行索引，索引值<0则代表没有标题行
+     * @return List<Map<String, String>>
+     */
+    public static List<Map<String, String>> toListMap(Sheet sheet, int titleRowNum) {
+        List<Map<String, String>> list = new ArrayList<>();
+        List<String> mapKeys = new ArrayList<>();
+        if (titleRowNum >= 0) {
+            Row titleRow = sheet.getRow(titleRowNum);
+            if (titleRow == null) {
+                return list;
+            }
+            for (int i = 0; i < titleRow.getLastCellNum(); i++) {
+                Cell cell = titleRow.getCell(i);
+                if (cell == null) {
+                    break;
+                }
+                String key = getStringCellValExcludeBlank(cell);
+                mapKeys.add(key);
+            }
+        } else {
+            // 根据第一行列的数量去找下面所有的列
+            Row titleRow = sheet.getRow(0);
+            if (titleRow == null) {
+                return list;
+            }
+            for (int i = 0; i < titleRow.getLastCellNum(); i++) {
+                Cell cell = titleRow.getCell(i);
+                if (cell == null) {
+                    break;
+                }
+                String colString = CellReference.convertNumToColString(cell.getColumnIndex());
+                mapKeys.add(colString);
+            }
+        }
+
+        for (int i = 0; i <= sheet.getLastRowNum(); i++) {
+            // 指定了标题列的情况下获取数据时跳过标题列
+            if (titleRowNum >= 0 && i == titleRowNum) {
+                continue;
+            }
+            Row row = sheet.getRow(i);
+            if (row == null) {
+                continue;
+            }
+            Map<String, String> map = new LinkedHashMap<>();
+            for (int c = 0; c < mapKeys.size(); c++) {
+                map.put(mapKeys.get(c), getStringCellValExcludeBlank(row.getCell(c)));
+            }
+            list.add(map);
+        }
+        return list;
     }
 }