54ba62a2c6e74332ffc742cb23faf21615b5d39fLubos Kosco * CDDL HEADER START
54ba62a2c6e74332ffc742cb23faf21615b5d39fLubos Kosco * The contents of this file are subject to the terms of the
54ba62a2c6e74332ffc742cb23faf21615b5d39fLubos Kosco * Common Development and Distribution License (the "License").
54ba62a2c6e74332ffc742cb23faf21615b5d39fLubos Kosco * You may not use this file except in compliance with the License.
54ba62a2c6e74332ffc742cb23faf21615b5d39fLubos Kosco * See LICENSE.txt included in this distribution for the specific
54ba62a2c6e74332ffc742cb23faf21615b5d39fLubos Kosco * language governing permissions and limitations under the License.
54ba62a2c6e74332ffc742cb23faf21615b5d39fLubos Kosco * When distributing Covered Code, include this CDDL HEADER in each
54ba62a2c6e74332ffc742cb23faf21615b5d39fLubos Kosco * file and include the License file at LICENSE.txt.
54ba62a2c6e74332ffc742cb23faf21615b5d39fLubos Kosco * If applicable, add the following below this CDDL HEADER, with the
54ba62a2c6e74332ffc742cb23faf21615b5d39fLubos Kosco * fields enclosed by brackets "[]" replaced with your own identifying
54ba62a2c6e74332ffc742cb23faf21615b5d39fLubos Kosco * information: Portions Copyright [yyyy] [name of copyright owner]
54ba62a2c6e74332ffc742cb23faf21615b5d39fLubos Kosco * CDDL HEADER END
eb1776903fd1f998009e97470a65fba8a499a0d9Lubos Kosco * Copyright (c) 2010, 2015, Oracle and/or its affiliates. All rights reserved.
eb1776903fd1f998009e97470a65fba8a499a0d9Lubos Kosco private final String defaultEncoding = new InputStreamReader(new ByteArrayInputStream(new byte[0])).getEncoding();
a39bcfe2e58183496eab6572675e2896e5045fa7Knut Anders Hatlen private static StreamSource getStreamSource(final byte[] bytes) {
a39bcfe2e58183496eab6572675e2896e5045fa7Knut Anders Hatlen public InputStream getStream() throws IOException {
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen public void defaultEncoding() throws IOException {
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen new TestableTextAnalyzer().analyze(new Document(),
a39bcfe2e58183496eab6572675e2896e5045fa7Knut Anders Hatlen getStreamSource("hello".getBytes()), null);
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen public void resetsStreamOnShortInput() throws IOException {
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen new TestableTextAnalyzer().analyze(new Document(),
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen public void utf8WithBOM() throws IOException {
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen byte[] buffer = new byte[]{(byte) 239, (byte) 187, (byte) 191, 'h', 'e', 'l', 'l', 'o'};
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen new TestableTextAnalyzer().analyze(new Document(),
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen public void utf16WithBOM() throws IOException {
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen final ByteBuffer utf16str = Charset.forName("UTF-16").encode("hello");
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen byte[] bytes = new byte[utf16str.remaining()];
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen new TestableTextAnalyzer().analyze(new Document(),
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen public void utf16WithBOMAlternate() throws IOException {
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen final ByteBuffer utf16str = Charset.forName("UTF-16").encode("hello");
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen byte[] bytes = new byte[utf16str.remaining()];
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen for (int i = 0; i < bytes.length; i += 2) {
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen new TestableTextAnalyzer().analyze(new Document(),
3733e9d930124c0050f9d5f27ee7a2f1b1b0bb2eKnut Anders Hatlen public class TestableTextAnalyzer extends TextAnalyzer {
a39bcfe2e58183496eab6572675e2896e5045fa7Knut Anders Hatlen public void analyze(Document doc, StreamSource src, Writer xrefOut) throws IOException {
a39bcfe2e58183496eab6572675e2896e5045fa7Knut Anders Hatlen try (Reader r = getReader(src.getStream())) {