Extracting Personal Information from PDF, DOC, DOCX, and TXT Files Using Apache Tika

This article explains how to extract feature data—such as a person's name and ID number—from various document formats (PDF, DOC, DOCX, TXT) using Apache Tika. The author provides a step‑by‑step guide that has been personally tested.

1. Add Maven dependencies

<!-- apache tika package for parsing pdf, word, txt -->
org.apache.tika
tika-core
2.8.0
org.apache.tika
tika-parsers-standard-package
2.8.0
org.apache.xmlbeans
xmlbeans
5.1.1

2. Write the Java code

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;

/**
 * Extract data from pdf, doc, docx, txt using Apache Tika.
 * Core jars: tika-core 2.8.0, tika-parsers-standard-package 2.8.0 (requires xmlbeans 5.1.1 for Word).
 * Assumes documents contain fields like:
 *   授权人（签字）：张三
 *   身份证号码: 322025199902256056
 */
public class TikaExtrator {
    public static void main(String[] args) {
        try {
            // Replace with actual file path; example uses a resource file.
            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();
        }
    }

    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. Execution result

Running the program prints the extracted text, the name (e.g., 张三), and the ID number (e.g., 322025199902256056). The original article includes a screenshot of the console output.