通过apache tika从文档(pdf、doc、docx、txt)中 提取特征数据

　　本文介绍如何通过apache tika从文档(pdf、doc、docx、txt)中 提取特征数据，比如文档中有身份证、姓名等信息。【全部是经本人实际测试过的功能】

1、需引入相关pom依赖

<!-- apache tika 包，用于解析pdf、word文本文档-->
 <dependency>
 <groupId>org.apache.tika</groupId>
 <artifactId>tika-core</artifactId>
 <version>2.8.0</version>
 </dependency>
 <!-- https://mvnrepository.com/artifact/org.apache.tika/tika-parsers-standard-package -->
 <dependency>
 <groupId>org.apache.tika</groupId>
 <artifactId>tika-parsers-standard-package</artifactId>
 <version>2.8.0</version>
 </dependency>
 <!-- tika包中缺少该xmlbeans依赖，不引入的话，tika解析word会报错 -->
 <dependency>
 <groupId>org.apache.xmlbeans</groupId>
 <artifactId>xmlbeans</artifactId>
 <version>5.1.1</version>
 </dependency>

2、编写相关代码

package org.example.wordcontent;
import org.apache.tika.Tika;
import org.apache.tika.exception.TikaException;
import java.io.File;
import java.io.IOException;
import java.io.InputStream;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
/**
 * 通过apache tika从 pdf、doc、docx、txt中提取数据
 * 核心依赖jar【tika-core 2.8.0、tika-parsers-standard-package 2.8.0(解析word时，需另外外依赖xmlbeans 5.1.1)】
 * 假定文档中的内容具有下列属性:
 * [授权人（签字）：张三
 * 身份证号码: 322025199902256056 ]
 * 待提取的内容为张三 和 322025199902256056。张三和322025199902256056的值会变
 */
public class TikaExtrator {
 public static void main(String[] args) {
 try {
 //// 替换为实际的PDF文件路径 测试例子: 如 测试.xlsx.
 InputStream input = TikaExtrator.class.getClassLoader().getResourceAsStream("综合信息查询授权书测试.docx");
 String text = extractTextFromFile(input);
 System.out.println("text: " + text);
 String name = extractName(text);
 String idNumber = extractIdNumber(text);
 System.out.println("授权人姓名: " + name);
 System.out.println("身份证号码: " + idNumber);
 } catch (IOException e) {
 e.printStackTrace();
 }
 }
 /**
 *
 * @param inputStream
 * @return
 * @throws IOException
 */
 private static String extractTextFromFile(InputStream inputStream) throws IOException {
 Tika tika = new Tika();
 try {
 return tika.parseToString(inputStream);
 } catch (TikaException e) {
 throw new RuntimeException(e);
 }
 }
 private static String extractName(String text) {
 Pattern pattern = Pattern.compile("授权人（签字）[：:]([\\u4e00-\\u9fa5]+)");
 Matcher matcher = pattern.matcher(text);
 if (matcher.find()) {
 return matcher.group(1);
 }
 return "";
 }
 private static String extractIdNumber(String text) {
 Pattern pattern = Pattern.compile("身份证号码[：:](\\d{18}|\\d{15})");
 Matcher matcher = pattern.matcher(text);
 if (matcher.find()) {
 return matcher.group(1);
 }
 return "";
 }
}

3、执行效果