001: /*
002: * gnu/regexp/REFilterInputStream.java
003: * Copyright (C) 1998-2001 Wes Biggs
004: *
005: * This library is free software; you can redistribute it and/or modify
006: * it under the terms of the GNU Lesser General Public License as published
007: * by the Free Software Foundation; either version 2.1 of the License, or
008: * (at your option) any later version.
009: *
010: * This library is distributed in the hope that it will be useful,
011: * but WITHOUT ANY WARRANTY; without even the implied warranty of
012: * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
013: * GNU Lesser General Public License for more details.
014: *
015: * You should have received a copy of the GNU Lesser General Public License
016: * along with this program; if not, write to the Free Software
017: * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
018: */
019:
020: package gnu.regexp;
021:
022: import java.io.FilterInputStream;
023: import java.io.InputStream;
024:
025: /**
026: * Replaces instances of a given RE found within an InputStream
027: * with replacement text. The replacements are interpolated into the
028: * stream when a match is found.
029: *
030: * @author <A HREF="mailto:wes@cacas.org">Wes Biggs</A>
031: * @deprecated This class cannot properly handle all character
032: * encodings. For proper handling, use the REFilterReader
033: * class instead.
034: */
035:
036: public class REFilterInputStream extends FilterInputStream {
037:
038: private RE expr;
039: private String replace;
040: private String buffer;
041: private int bufpos;
042: private int offset;
043: private CharIndexedInputStream stream;
044:
045: /**
046: * Creates an REFilterInputStream. When reading from this stream,
047: * occurrences of patterns matching the supplied regular expression
048: * will be replaced with the supplied replacement text (the
049: * metacharacters $0 through $9 may be used to refer to the full
050: * match or subexpression matches).
051: *
052: * @param stream The InputStream to be filtered.
053: * @param expr The regular expression to search for.
054: * @param replace The text pattern to replace matches with.
055: */
056: public REFilterInputStream(InputStream stream, RE expr,
057: String replace) {
058: super (stream);
059: this .stream = new CharIndexedInputStream(stream, 0);
060: this .expr = expr;
061: this .replace = replace;
062: }
063:
064: /**
065: * Reads the next byte from the stream per the general contract of
066: * InputStream.read(). Returns -1 on error or end of stream.
067: */
068: public int read() {
069: // If we have buffered replace data, use it.
070: if ((buffer != null) && (bufpos < buffer.length())) {
071: return (int) buffer.charAt(bufpos++);
072: }
073:
074: // check if input is at a valid position
075: if (!stream.isValid())
076: return -1;
077:
078: REMatch mymatch = new REMatch(expr.getNumSubs(), offset, 0);
079: if (expr.match(stream, mymatch)) {
080: mymatch.end[0] = mymatch.index;
081: mymatch.finish(stream);
082: stream.move(mymatch.toString().length());
083: offset += mymatch.toString().length();
084: buffer = mymatch.substituteInto(replace);
085: bufpos = 1;
086:
087: // This is prone to infinite loops if replace string turns out empty.
088: if (buffer.length() > 0) {
089: return buffer.charAt(0);
090: }
091: }
092: char ch = stream.charAt(0);
093: if (ch == CharIndexed.OUT_OF_BOUNDS)
094: return -1;
095: stream.move(1);
096: offset++;
097: return ch;
098: }
099:
100: /**
101: * Returns false. REFilterInputStream does not support mark() and
102: * reset() methods.
103: */
104: public boolean markSupported() {
105: return false;
106: }
107:
108: /** Reads from the stream into the provided array. */
109: public int read(byte[] b, int off, int len) {
110: int i;
111: int ok = 0;
112: while (len-- > 0) {
113: i = read();
114: if (i == -1)
115: return (ok == 0) ? -1 : ok;
116: b[off++] = (byte) i;
117: ok++;
118: }
119: return ok;
120: }
121:
122: /** Reads from the stream into the provided array. */
123: public int read(byte[] b) {
124: return read(b, 0, b.length);
125: }
126: }
|