001/**
002 * Licensed to the Apache Software Foundation (ASF) under one or more
003 * contributor license agreements.  See the NOTICE file distributed with
004 * this work for additional information regarding copyright ownership.
005 * The ASF licenses this file to You under the Apache License, Version 2.0
006 * (the "License"); you may not use this file except in compliance with
007 * the License.  You may obtain a copy of the License at
008 *
009 *      http://www.apache.org/licenses/LICENSE-2.0
010 *
011 * Unless required by applicable law or agreed to in writing, software
012 * distributed under the License is distributed on an "AS IS" BASIS,
013 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
014 * See the License for the specific language governing permissions and
015 * limitations under the License.
016 */
017package org.apache.camel.util;
018
019import java.io.Closeable;
020import java.io.File;
021import java.io.FileInputStream;
022import java.io.FileNotFoundException;
023import java.io.IOException;
024import java.io.InputStream;
025import java.io.InputStreamReader;
026import java.io.StringReader;
027import java.nio.CharBuffer;
028import java.nio.channels.Channels;
029import java.nio.channels.ReadableByteChannel;
030import java.nio.charset.Charset;
031import java.nio.charset.CharsetDecoder;
032import java.nio.charset.IllegalCharsetNameException;
033import java.nio.charset.UnsupportedCharsetException;
034import java.util.InputMismatchException;
035import java.util.Iterator;
036import java.util.LinkedHashMap;
037import java.util.Map;
038import java.util.Map.Entry;
039import java.util.NoSuchElementException;
040import java.util.Objects;
041import java.util.regex.Matcher;
042import java.util.regex.Pattern;
043
044public final class Scanner implements Iterator<String>, Closeable {
045
046    private static final Map<String, Pattern> CACHE = new LinkedHashMap<String, Pattern>() {
047        @Override
048        protected boolean removeEldestEntry(Entry<String, Pattern> eldest) {
049            return size() >= 7;
050        }
051    };
052
053    private static final String WHITESPACE_PATTERN = "\\s+";
054
055    private static final String FIND_ANY_PATTERN = "(?s).*";
056
057    private static final int BUFFER_SIZE = 1024;
058
059    private Readable source;
060    private Pattern delimPattern;
061    private Matcher matcher;
062    private CharBuffer buf;
063    private int position;
064    private boolean inputExhausted;
065    private boolean needInput;
066    private boolean skipped;
067    private int savedPosition = -1;
068    private boolean closed;
069    private IOException lastIOException;
070
071    public Scanner(InputStream source, String charsetName, String pattern) {
072        this(new InputStreamReader(Objects.requireNonNull(source, "source"), toDecoder(charsetName)), cachePattern(pattern));
073    }
074
075    public Scanner(File source, String charsetName, String pattern) throws FileNotFoundException {
076        this(new FileInputStream(Objects.requireNonNull(source, "source")).getChannel(), charsetName, pattern);
077    }
078
079    public Scanner(String source, String pattern) {
080        this(new StringReader(Objects.requireNonNull(source, "source")), cachePattern(pattern));
081    }
082
083    public Scanner(ReadableByteChannel source, String charsetName, String pattern) {
084        this(Channels.newReader(Objects.requireNonNull(source, "source"), toDecoder(charsetName), -1), cachePattern(pattern));
085    }
086
087    public Scanner(Readable source, String pattern) {
088        this(Objects.requireNonNull(source, "source"), cachePattern(pattern));
089    }
090
091    private Scanner(Readable source, Pattern pattern) {
092        this.source = source;
093        delimPattern = pattern != null ? pattern : cachePattern(WHITESPACE_PATTERN);
094        buf = CharBuffer.allocate(BUFFER_SIZE);
095        buf.limit(0);
096        matcher = delimPattern.matcher(buf);
097        matcher.useTransparentBounds(true);
098        matcher.useAnchoringBounds(false);
099    }
100
101    private static CharsetDecoder toDecoder(String charsetName) {
102        try {
103            Charset cs = charsetName != null ? Charset.forName(charsetName) : Charset.defaultCharset();
104            return cs.newDecoder();
105        } catch (IllegalCharsetNameException | UnsupportedCharsetException e) {
106            throw new IllegalArgumentException(e);
107        }
108    }
109
110    public boolean hasNext() {
111        checkClosed();
112        saveState();
113        while (!inputExhausted) {
114            if (hasTokenInBuffer()) {
115                revertState();
116                return true;
117            }
118            readMore();
119        }
120        boolean result = hasTokenInBuffer();
121        revertState();
122        return result;
123    }
124
125    public String next() {
126        checkClosed();
127        while (true) {
128            String token = getCompleteTokenInBuffer();
129            if (token != null) {
130                skipped = false;
131                return token;
132            }
133            if (needInput) {
134                readMore();
135            } else {
136                throwFor();
137            }
138        }
139    }
140
141    private void saveState() {
142        savedPosition = position;
143    }
144
145    private void revertState() {
146        position = savedPosition;
147        savedPosition = -1;
148        skipped = false;
149    }
150
151    private void readMore() {
152        if (buf.limit() == buf.capacity()) {
153            expandBuffer();
154        }
155        int p = buf.position();
156        buf.position(buf.limit());
157        buf.limit(buf.capacity());
158        int n;
159        try {
160            n = source.read(buf);
161        } catch (IOException ioe) {
162            lastIOException = ioe;
163            n = -1;
164        }
165        if (n == -1) {
166            inputExhausted = true;
167            needInput = false;
168        } else if (n > 0) {
169            needInput = false;
170        }
171        buf.limit(buf.position());
172        buf.position(p);
173    }
174
175    private void expandBuffer() {
176        int offset = savedPosition == -1 ? position : savedPosition;
177        buf.position(offset);
178        if (offset > 0) {
179            buf.compact();
180            translateSavedIndexes(offset);
181            position -= offset;
182            buf.flip();
183        } else {
184            int newSize = buf.capacity() * 2;
185            CharBuffer newBuf = CharBuffer.allocate(newSize);
186            newBuf.put(buf);
187            newBuf.flip();
188            translateSavedIndexes(offset);
189            position -= offset;
190            buf = newBuf;
191            matcher.reset(buf);
192        }
193    }
194
195    private void translateSavedIndexes(int offset) {
196        if (savedPosition != -1) {
197            savedPosition -= offset;
198        }
199    }
200
201    private void throwFor() {
202        skipped = false;
203        if (inputExhausted && position == buf.limit()) {
204            throw new NoSuchElementException();
205        } else {
206            throw new InputMismatchException();
207        }
208    }
209
210    private boolean hasTokenInBuffer() {
211        matcher.usePattern(delimPattern);
212        matcher.region(position, buf.limit());
213        if (matcher.lookingAt()) {
214            position = matcher.end();
215        }
216        return position != buf.limit();
217    }
218
219    private String getCompleteTokenInBuffer() {
220        matcher.usePattern(delimPattern);
221        if (!skipped) {
222            matcher.region(position, buf.limit());
223            if (matcher.lookingAt()) {
224                if (matcher.hitEnd() && !inputExhausted) {
225                    needInput = true;
226                    return null;
227                }
228                skipped = true;
229                position = matcher.end();
230            }
231        }
232        if (position == buf.limit()) {
233            if (inputExhausted) {
234                return null;
235            }
236            needInput = true;
237            return null;
238        }
239        matcher.region(position, buf.limit());
240        boolean foundNextDelim = matcher.find();
241        if (foundNextDelim && (matcher.end() == position)) {
242            foundNextDelim = matcher.find();
243        }
244        if (foundNextDelim) {
245            if (matcher.requireEnd() && !inputExhausted) {
246                needInput = true;
247                return null;
248            }
249            int tokenEnd = matcher.start();
250            matcher.usePattern(cachePattern(FIND_ANY_PATTERN));
251            matcher.region(position, tokenEnd);
252            if (matcher.matches()) {
253                String s = matcher.group();
254                position = matcher.end();
255                return s;
256            } else {
257                return null;
258            }
259        }
260        if (inputExhausted) {
261            matcher.usePattern(cachePattern(FIND_ANY_PATTERN));
262            matcher.region(position, buf.limit());
263            if (matcher.matches()) {
264                String s = matcher.group();
265                position = matcher.end();
266                return s;
267            }
268            return null;
269        }
270        needInput = true;
271        return null;
272    }
273
274    private void checkClosed() {
275        if (closed) {
276            throw new IllegalStateException();
277        }
278    }
279
280    public void close() throws IOException {
281        if (!closed) {
282            closed = true;
283            if (source instanceof Closeable) {
284                try {
285                    ((Closeable) source).close();
286                } catch (IOException e) {
287                    lastIOException = e;
288                }
289            }
290        }
291        if (lastIOException != null) {
292            throw lastIOException;
293        }
294    }
295
296    private static Pattern cachePattern(String pattern) {
297        if (pattern == null) {
298            return null;
299        }
300        return CACHE.computeIfAbsent(pattern, Pattern::compile);
301    }
302
303}