Java实现提取Word文档表格数据

<repositories>
    <repository>
        <id>com.e-iceblue</id>
        <name>e-iceblue</name>
        <url>https://repo.e-iceblue.com/nexus/content/groups/public/</url>
    </repository>
</repositories>
<dependency>
    <groupId>e-iceblue</groupId>
    <artifactId>spire.doc.free</artifactId>
    <version>5.3.2</version>
</dependency>

import com.spire.doc.*;
import com.spire.doc.documents.Paragraph;

import java.io.FileWriter;
import java.io.IOException;

public class ExtractWordTable {
    public static void main(String[] args) {
        // 创建一个文档对象
        Document doc = new Document();

        try {
            // 加载一个Word文档
            doc.loadFromFile("GSample.docx");

            // 遍历文档中的各节
            for (int i = 0; i < doc.getSections().getCount(); i++) {
                // 获取一节
                Section section = doc.getSections().get(i);
                // 遍历该节中的表格
                for (int j = 0; j < section.getTables().getCount(); j++) {
                    // 获取一个表格
                    Table table = section.getTables().get(j);
                    // 收集所有表格内容
                    StringBuilder tableText = new StringBuilder();
                    for (int k = 0; k < table.getRows().getCount(); k++) {
                        // 获取一行
                        TableRow row = table.getRows().get(k);
                        // 遍历行中的单元格
                        StringBuilder rowText = new StringBuilder();
                        for (int l = 0; l < row.getCells().getCount(); l++) {
                            // 获取一个单元格
                            TableCell cell = row.getCells().get(l);
                            // 遍历段落以获取单元格中的文本
                            String cellText = "";
                            for (int m = 0; m < cell.getParagraphs().getCount(); m++) {
                                Paragraph paragraph = cell.getParagraphs().get(m);
                                cellText += paragraph.getText() + " ";
                            }
                            if (l < row.getCells().getCount() - 1) {
                                rowText.append(cellText).append("\t");
                            } else {
                                rowText.append(cellText).append("\n");
                            }
                        }
                        tableText.append(rowText);
                    }

                    // 使用try-with-resources将表格文本写入文件
                    try (FileWriter writer = new FileWriter("output/Tables/Section-" + (i + 1) + "-Table-" + (j + 1) + ".txt")) {
                        writer.write(tableText.toString());
                    }
                }
            }
        } catch (IOException e) {
            throw new RuntimeException(e);
        }
    }
}

import com.spire.doc.*;
import com.spire.doc.documents.Paragraph;
import com.spire.xls.FileFormat;
import com.spire.xls.Workbook;
import com.spire.xls.Worksheet;

public class ExtractWordTableToExcel {
    public static void main(String[] args) {
        // 创建Document对象
        Document doc = new Document();

        // 创建Workbook对象
        Workbook workbook = new Workbook();
        // 删除默认的工作表
        workbook.getWorksheets().clear();

        try {
            // 加载Word文档
            doc.loadFromFile("Sample.docx");

            // 遍历文档中的各节
            for (int i = 0; i < doc.getSections().getCount(); i++) {
                // 获取一节
                Section section = doc.getSections().get(i);
                // 遍历该节中的表格
                for (int j = 0; j < section.getTables().getCount(); j++) {
                    // 获取一个表格
                    Table table = section.getTables().get(j);
                    // 为每个表格创建一个工作表
                    Worksheet sheet = workbook.getWorksheets().add("Section-" + (i + 1) + "-Table-" + (j + 1));
                    for (int k = 0; k < table.getRows().getCount(); k++) {
                        // 获取一行
                        TableRow row = table.getRows().get(k);
                        for (int l = 0; l < row.getCells().getCount(); l++) {
                            // 获取一个单元格
                            TableCell cell = row.getCells().get(l);
                            // 遍历单元格中的段落以获取文本
                            String cellText = "";
                            for (int m = 0; m < cell.getParagraphs().getCount(); m++) {
                                Paragraph paragraph = cell.getParagraphs().get(m);
                                if (m > 0 && m < cell.getParagraphs().getCount() - 1) {
                                    cellText += paragraph.getText() + "\n";
                                } else {
                                    cellText += paragraph.getText();
                                }
                                // 将单元格的文本写入对应的Excel工作表的单元格中
                                sheet.getRange().get(k + 1, l + 1).setValue(cellText);
                            }
                            // 自动调整列宽
                            sheet.autoFitColumn(l + 1);
                        }
                    }
                }
            }
        } catch (Exception e) {
            throw new RuntimeException(e);
        }
        // 保存为Excel文件
        workbook.saveToFile("output/WordTableToExcel.xlsx", FileFormat.Version2016);
        // 释放资源
        workbook.dispose();
    }
}