PDFToText.java

/**
 * 
 */
package FileConvert2Txt;

import java.io.ByteArrayInputStream;
import java.io.IOException;
import java.io.InputStream;
import java.nio.charset.Charset;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

import org.apache.commons.lang.StringUtils;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.text.PDFTextStripper;
import org.juno.scheduler.util.ConvertException;

/**
 * @author denghc
 * @desc PDF文件转换格式化文本
 */
public class PDFToText implements IFileToText {

	@Override
	public String fileToText(byte[] byts,int flag)throws ConvertException{
		InputStream in = null;
		PDDocument document = null;
		PDFTextStripper stripper;
		String content = null;
		in = new ByteArrayInputStream(byts);
		try {
			document = PDDocument.load(in);
			stripper = new PDFTextStripper();
			content = stripper.getText(document).trim();
		} catch (Exception e) {
			throw new ConvertException("PDF文件转换格式化文本异常"+e.getMessage());
		} finally {
			try {
				if (document != null) {
					document.close();
				}
				if (in != null) {
					in.close();
				}
			} catch (IOException e) {
				e.printStackTrace();
			}
		}
		return format(content);
	}
	private String format(String content) {
		if (content == null || content.isEmpty()) {
			return null;
		}
		String spilter = System.getProperty("line.separator");
		String blank = "    ";
		StringBuilder textStrBuffer = new StringBuilder();
		content=StringUtils.replaceEach(content,new String[]{"[　 ]","\\?",","},new String[]{"","",""});
		String[] strelines = content.trim().split(spilter);
		String buffer = blank;
		String tempStr = "";
		int pageUp = 1;//页面个数
		int pageDown = 1;//页脚个数
		for (int i = 0; i < strelines.length; i++) {
			String temp = strelines[i].trim().replaceAll("[ ]+", " ");
			if(temp.length() <= 0){
				continue;
			}
//			页眉统计
			if(!"".equals(tempStr) && temp.equals(tempStr)){
				pageUp++;
			}else if("".equals(tempStr) && (isPageNumber(temp) || temp.matches("[123456789][\\d]{0,2}"))){
				tempStr = buffer.trim();
			}
//			去除页号
			if (isPageNumber(temp) || (temp.matches("[123456789][\\d]{0,2}") && String.valueOf(pageDown).equals(temp))) {
				pageDown++;
				continue;//存在一定风险
			}
			int one = temp.indexOf("，");//逗号
			String t4 = buffer.trim();
			if (i > 4 && !(StringUtils.endsWithAny(t4,new String[]{"。","：","；","！"}))) {
//				判断上一行是否结束
				Pattern p2 = Pattern.compile("^[\u4e00-\u9fa5。；0-9]+$");// 针对标题、短句被硬性换行符分隔的情况
				Matcher m2 = p2.matcher(temp);
//				日期被分隔判断
				Pattern p3 = Pattern.compile("^[年月日]");
				Matcher m3 = p3.matcher(temp);
				Pattern p4 = Pattern.compile("\\d$");
				Matcher m4 = p4.matcher(t4);
				
				if (temp.indexOf("》") < temp.indexOf("《")
						|| (temp.indexOf("》") > 0 && temp.indexOf("《") < 0)
						|| temp.indexOf("”") < temp.indexOf("“")
						|| (temp.indexOf("”") > 0 && temp.indexOf("“") < 0)
						|| temp.indexOf("）", 3) < temp.indexOf("（", 3)
						|| (temp.indexOf("）", 3) > 0 && temp.indexOf("（", 3) < 0)
						|| (m3.find() && m4.find()) || m2.find()
						|| (temp.indexOf("。") > 0 && temp.indexOf("。") < 7)
						|| (one > 0 && one < 5 && !temp.substring(one - 1, one).matches("[\\d]"))
						|| one == 0 || temp.startsWith(".")
						|| (temp.indexOf("；") >= 0 && temp.indexOf("；") < 3)) {
					buffer = buffer.split(spilter)[0];
				}
//				本行是标题，上行一定要有换行符
				if(isKeywords(temp)){
					if(!buffer.endsWith(spilter + blank)){
						buffer = buffer + spilter + blank;
					}
				}
			}
			textStrBuffer.append(buffer);
			if (isKeywords(temp)) {// 该行为标题行
				if (temp.indexOf("）", 3) < temp.indexOf("（", 3)
						|| temp.indexOf("》") < temp.indexOf("《")
						|| temp.indexOf("”") < temp.indexOf("“")
						|| temp.endsWith("、")) {
					buffer = temp;
				} else {
					buffer = temp + spilter + blank;
				}
			} else {
				if (temp.indexOf("》") < temp.indexOf("《")
						|| temp.indexOf("”") < temp.indexOf("“")
						|| temp.indexOf("）") < temp.indexOf("（")) {
					buffer = temp;
				} else if (temp.endsWith("。") || temp.endsWith("：")
						|| temp.endsWith("；") || t4.endsWith("！")) {
					buffer = temp + spilter + blank;
				} else if ((temp.indexOf("。") >= 0 && !temp.endsWith("。"))
						|| (one > 0 && !temp.substring(one - 1, one).matches("[\\d]"))
						|| temp.indexOf("、") >= 0
						|| temp.endsWith(".")) {
					buffer = temp;
				} else {
					buffer = temp + spilter + blank;
				}
			}
		}
		textStrBuffer.append(buffer);
		if(pageUp > 4){
//			去除页眉页脚
			//匹配符号* ( ) 转义
			tempStr=StringUtils.replaceEach(tempStr,new String[]{"\\*","\\(","(","\\)",")","\\[","[","\\]","]","\\{","{","\\}","}"}, new String[]{"\\\\*","\\\\(","\\(","\\\\)","\\)","\\\\[","\\[","\\\\]","\\]","\\\\{","\\{","\\\\}","\\}"});
			return textStrBuffer.toString().replaceAll(tempStr+"[\\s]*[\\d]{0,3}", "").replaceAll(spilter + spilter, spilter).replaceAll("    " + spilter, "").replaceAll("        ", "    ");
		} else {
			return textStrBuffer.toString();
		}
	}
	
	private boolean isKeywords(String link) {		
//		以下：存在下列情况时，忽略关键字匹配
		if (link.indexOf("）") < link.indexOf("（")) {
			return false;
		}
		if (link.indexOf("》") < link.indexOf("《")) {
			return false;
		}
		if (link.endsWith("，")) {
			return false;
		}
		Pattern p2 = Pattern.compile("[<(（第]?[1234567890一二三四五六七八九十ABCDEF]{1,3}[>)、）.节章条]?");
		if(link.length() > 4){
			link = link.substring(0, 4);
		}
		Matcher m2 = p2.matcher(link);
		return m2.find();
	}
	
	public String getTextFromByte(byte[] fileContent) {
		String content = null;

		try {
			content = new String(fileContent, Charset.forName("UTF-8"));
		} catch (Exception e) {
			e.printStackTrace();
		} 
		if (content == null || content.length() <= 0) {
			return null;
		}
		// 格式排版
		StringBuffer sb = new StringBuffer();
		content = content.replaceAll("[ ]{1,}", " ");
		content = content.replaceAll("[\\?]{1,}", "");
		String spliter = System.getProperty("line.separator");
		String[] lines = content.split(spliter);
		if (lines.length <= 1) {
			spliter = "\n";
			lines = content.split(spliter);
		}
		for (int i = 0; i < lines.length; i++) {
			String line = lines[i].trim();
			if (line.length() == 0 || isPageNumber(line)) {
				continue;
			}
			line = "    " + line + spliter;
			sb.append(line);
		}

		return sb.toString();
	}
	private boolean isPageNumber(String line) {
		boolean flag = false;
		line = line.replaceAll("[\\s]+", "");
		if (line.length() < 11) {
			if (!flag) {
				Pattern p2 = Pattern.compile("-[0-9\\s]{1,}-");
				Matcher m2 = p2.matcher(line);
				flag = m2.matches();
			}
			if (!flag) {
				Pattern p3 = Pattern.compile("第[\\s0-9]{1,}页");
				Matcher m3 = p3.matcher(line);
				flag = m3.matches();
			}
			if (!flag) {
				Pattern p4 = Pattern.compile("[0-9\\s]{1,}-[0-9\\s]{1,}");
				Matcher m4 = p4.matcher(line);
				flag = m4.matches();
			}
			if(!flag){
				Pattern p4 = Pattern.compile("[0-9\\s]{1,}/[0-9\\s]{1,}");
				Matcher m4 = p4.matcher(line);
				flag = m4.matches();
			}
		}
		return flag;
	}
}