
com.itextpdf.tool.xml.net.FileRetrieveImpl Maven / Gradle / Ivy
/*
*
* This file is part of the iText (R) project.
Copyright (c) 1998-2022 iText Group NV
* Authors: Balder Van Camp, Emiel Ackermann, et al.
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU Affero General Public License version 3
* as published by the Free Software Foundation with the addition of the
* following permission added to Section 15 as permitted in Section 7(a):
* FOR ANY PART OF THE COVERED WORK IN WHICH THE COPYRIGHT IS OWNED BY
* ITEXT GROUP. ITEXT GROUP DISCLAIMS THE WARRANTY OF NON INFRINGEMENT
* OF THIRD PARTY RIGHTS
*
* This program is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
* or FITNESS FOR A PARTICULAR PURPOSE.
* See the GNU Affero General Public License for more details.
* You should have received a copy of the GNU Affero General Public License
* along with this program; if not, see http://www.gnu.org/licenses or write to
* the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
* Boston, MA, 02110-1301 USA, or download the license from the following URL:
* http://itextpdf.com/terms-of-use/
*
* The interactive user interfaces in modified source and object code versions
* of this program must display Appropriate Legal Notices, as required under
* Section 5 of the GNU Affero General Public License.
*
* In accordance with Section 7(b) of the GNU Affero General Public License,
* a covered work must retain the producer line in every PDF that is created
* or manipulated using iText.
*
* You can be released from the requirements of the license by purchasing
* a commercial license. Buying such a license is mandatory as soon as you
* develop commercial activities involving the iText software without
* disclosing the source code of your own applications.
* These activities include: offering paid services to customers as an ASP,
* serving PDFs on the fly in a web application, shipping iText with a closed
* source product.
*
* For more information, please contact iText Software Corp. at this
* address: [email protected]
*/
package com.itextpdf.tool.xml.net;
import java.io.File;
import java.io.FileInputStream;
import java.io.IOException;
import java.io.InputStream;
import java.net.MalformedURLException;
import java.net.URL;
import java.util.List;
import java.util.concurrent.CopyOnWriteArrayList;
import com.itextpdf.text.log.Level;
import com.itextpdf.text.log.Logger;
import com.itextpdf.text.log.LoggerFactory;
import com.itextpdf.tool.xml.exceptions.LocaleMessages;
import com.itextpdf.tool.xml.exceptions.RuntimeWorkerException;
/**
* @author redlab_b
*
*/
public class FileRetrieveImpl implements FileRetrieve {
private static final Logger LOGGER = LoggerFactory.getLogger(FileRetrieveImpl.class);
private final List rootdirs;
private final List urls;
/**
*
*/
public FileRetrieveImpl() {
rootdirs = new CopyOnWriteArrayList();
urls = new CopyOnWriteArrayList();
}
/**
* Constructs a new FileRetrieveImpl with the given root url's and
* directories
*
* @param strings an array of strings, if the String starts with http or
* https it's taken as URL otherwise we check if it's a directory
* with
*
*
* File f = new File(str);
* f.isDirectory()
*
*/
public FileRetrieveImpl(final String... strings) {
this();
for (String s : strings) {
if (s.startsWith("http") || s.startsWith("https")) {
urls.add(s);
} else {
File f = new File(s);
if (f.isDirectory()) {
rootdirs.add(f);
}
}
}
}
/**
* Constructs a new FileRetrieveImpl with the given root url's and
* directories
*
* @param strings an array of strings, if the String starts with http or
* https it's taken as URL otherwise we check if it's a directory
* with
*
*
* File f = new File(str);
* f.isDirectory()
*
*/
public FileRetrieveImpl(File rootdir) {
this();
if (rootdir.isDirectory())
rootdirs.add(rootdir);
}
/**
* ProcessFromHref first tries to create an {@link URL} from the given href
,
* if that throws a {@link MalformedURLException}, it will prepend the given
* root URLs to href
until a valid URL is found.
If by then there is
* no valid url found, this method will see if the given href
is a valid file
* and can read it.
If it's not a valid file or a file that can't be read,
* the given root directories will be set as root path with the given href
as
* file path until a valid file has been found.
*/
public void processFromHref(final String href, final ReadingProcessor processor) throws IOException {
if (LOGGER.isLogging(Level.DEBUG)) {
LOGGER.debug(String.format(LocaleMessages.getInstance().getMessage("retrieve.file.from"), href));
}
URL url = null;
File f = null;
boolean isfile = false;
try {
url = new URL(href);
} catch (MalformedURLException e) {
try {
url = detectWithRootUrls(href);
} catch (MalformedURLException e1) {
// its probably a file, try to detect it.
f = new File(href);
isfile = true;
if (!(f.isFile() && f.canRead())) {
isfile = false;
for (File root : rootdirs) {
f = new File(root, href);
if (f.isFile() && f.canRead()) {
isfile = true;
break;
}
}
}
}
}
InputStream in = null;
if (null != url) {
in = url.openStream();
} else if (isfile) {
in = new FileInputStream(f);
} else {
throw new IOException(LocaleMessages.getInstance().getMessage("retrieve.file.from.nothing"));
}
read(processor, in);
}
/**
* @param href the reference
* @throws MalformedURLException if no valid URL could be found.
*/
private URL detectWithRootUrls(final String href) throws MalformedURLException {
for (String root : urls) {
try {
return new URL(root + href);
} catch (MalformedURLException e) {
}
}
throw new MalformedURLException();
}
/* (non-Javadoc)
* @see com.itextpdf.tool.xml.net.FileRetrieve#processFromStream(java.io.InputStream, com.itextpdf.tool.xml.net.ReadingProcessor)
*/
public void processFromStream(final InputStream in, final ReadingProcessor processor) throws IOException {
read(processor, in);
}
/**
* @param processor
* @param in
* @throws IOException
*/
private void read(final ReadingProcessor processor, final InputStream in) throws IOException {
try {
int inbit = -1;
while ((inbit = in.read()) != -1) {
processor.process(inbit);
}
} catch (IOException e) {
throw e;
} finally {
try {
if (null != in) {
in.close();
}
} catch (IOException e) {
throw new RuntimeWorkerException(e);
}
}
}
/**
* Add a root directory.
* @param dir the root directory
*/
public void addRootDir(final File dir) {
rootdirs.add(dir);
}
/**
* Add a root URL.
* @param url the URL
*/
public void addURL(final String url) {
urls.add(url);
}
}
© 2015 - 2025 Weber Informatics LLC | Privacy Policy