ckan2zenodo-library/src/main/java/org/gcube/data/publishing/ckan2zenodo/model/DownloadedFile.java

package org.gcube.data.publishing.ckan2zenodo.model;

import java.io.File;
import java.io.IOException;
import java.io.InputStream;
import java.net.HttpURLConnection;
import java.net.URL;
import java.nio.file.Files;
import java.nio.file.StandardCopyOption;
import java.security.DigestInputStream;
import java.security.MessageDigest;
import java.security.NoSuchAlgorithmException;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

import lombok.*;
import lombok.extern.slf4j.Slf4j;


@RequiredArgsConstructor
@Slf4j
public class DownloadedFile {

	private static final Pattern FILENAME_IN_DEPOSITION_REGEXP = Pattern.compile("(?<=filename\\=\\\").*(?=\\\")");

	@NonNull
	@Getter
	private CkanResource source;
	//private File f;

	private File f=null;
	private String MD5=null;
	private String remoteFileName=null;

	@Override
	public String toString() {
		return "DownloadedFile{" +
				"source=" + source +
				", f=" + f +
				", MD5='" + MD5 + '\'' +
				", remoteFileName='" + remoteFileName + '\'' +
				'}';
	}

	public String getToUseFileName() throws Exception {
		if (getExtension(source.getName())!=null){
			// source contains extension
			return source.getName();
		}else {
			if(remoteFileName == null){
				initRemoteFileName();
			}
			String evaluatedExtension=getExtension(remoteFileName);
			if(evaluatedExtension!=null)
				return source.getName()+evaluatedExtension;
			else return source.getName(); // No extension
		}
	}



	public File getFile() throws Exception {
		if(f==null){
			download();
		}
		return f;
	}

	public String getMD5() throws Exception {
		if(MD5==null){
			download();
		}
		return MD5;
	}





	static final String getExtension(String filename){
			int lastIndexOf = filename.lastIndexOf(".");
			if (lastIndexOf == -1) {
				return null; // no extension
			}
			return filename.substring(lastIndexOf);
	}


	/**
	 *  Actually downlaods the file INIT File, MD5 and remoteFilename.
	 */
	private void download() throws Exception {
		log.info("Downloading {} from {}",source.getName(),source.getUrl());
		URL url=new URL(source.getUrl());
		f=File.createTempFile("zenodo_", ".tmp");
		MessageDigest md = MessageDigest.getInstance("MD5");

		// Multiple tries
		InputStream is=null;
		int attempt=0;
		Exception lastException=null;


		while(is==null&&attempt<5) {
			try {
				attempt++;
				is=url.openStream();

				if(remoteFileName == null)
					remoteFileName = getFilenameFromURL(url);

			}catch(Exception e) {
				lastException=e;
				try{
					Thread.sleep(500*attempt);
				}catch(InterruptedException e1) {}
			}
		}
		if(is==null) throw new Exception("Unable to download "+source.getUrl(),lastException);

		if(remoteFileName == null)
			remoteFileName = ""; // Unable to evaluate from HEAD


		DigestInputStream dis = new DigestInputStream(is, md);
		MD5 = dis.getMessageDigest().toString();

		// Download
		long size= Files.copy(is, f.toPath(), StandardCopyOption.REPLACE_EXISTING);
		log.info("Received {} bytes for {} ",size,source.getName());
	}

	/**
	 * Performs HTTP HEAD and INIT remoteFileName
	 */
	private void initRemoteFileName() throws Exception {
		int attempt =0;
		Exception lastException=null;
		URL url=new URL(source.getUrl());
		while(remoteFileName==null&&attempt<5) {
			try {
				attempt++;
				remoteFileName = getFilenameFromURL(url);
			}catch(Exception e) {
				lastException=e;
				try{
					Thread.sleep(500*attempt);
				}catch(InterruptedException e1) {}
			}
		}
		if(remoteFileName == null) {
			remoteFileName = ""; // Unable to evaluate from HEAD
			log.warn("Unable to get remote file name from {} [resource Name {}]",source.getUrl(),source.getName(),lastException);
		}
	}


	private static final String getFilenameFromURL(URL url) throws IOException {
		HttpURLConnection con = (HttpURLConnection) url.openConnection();
		con.setRequestMethod("GET");
		String contentDisp= con.getHeaderField("Content-Disposition");

		Matcher m = FILENAME_IN_DEPOSITION_REGEXP.matcher(contentDisp);
		m.find();
		return m.group(0);
	}
}