blob: 708623524967bd2689830a4d3b5e5117c890e7ac [file] [log] [blame]
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.jasper.compiler;
import java.io.CharArrayWriter;
import java.io.FileNotFoundException;
import java.io.IOException;
import java.io.InputStreamReader;
import org.apache.jasper.JasperException;
import org.apache.jasper.JspCompilationContext;
import org.apache.jasper.runtime.ExceptionUtils;
import org.apache.juli.logging.Log;
import org.apache.juli.logging.LogFactory;
import org.apache.tomcat.Jar;
/**
* JspReader is an input buffer for the JSP parser. It should allow
* unlimited lookahead and pushback. It also has a bunch of parsing
* utility methods for understanding htmlesque thingies.
*
* @author Anil K. Vijendran
* @author Anselm Baird-Smith
* @author Harish Prabandham
* @author Rajiv Mordani
* @author Mandar Raje
* @author Danno Ferrin
* @author Kin-man Chung
* @author Shawn Bayern
* @author Mark Roth
*/
class JspReader {
/**
* Logger.
*/
private final Log log = LogFactory.getLog(JspReader.class);
/**
* The current spot in the file.
*/
private Mark current;
/**
* The compilation context.
*/
private final JspCompilationContext context;
/**
* The Jasper error dispatcher.
*/
private final ErrorDispatcher err;
/**
* Constructor.
*
* @param ctxt The compilation context
* @param fname The file name
* @param encoding The file encoding
* @param jar ?
* @param err The error dispatcher
* @throws JasperException If a Jasper-internal error occurs
* @throws FileNotFoundException If the JSP file is not found (or is unreadable)
* @throws IOException If an IO-level error occurs, e.g. reading the file
*/
public JspReader(JspCompilationContext ctxt,
String fname,
String encoding,
Jar jar,
ErrorDispatcher err)
throws JasperException, FileNotFoundException, IOException {
this(ctxt, fname, JspUtil.getReader(fname, encoding, jar, ctxt, err),
err);
}
/**
* Constructor: same as above constructor but with initialized reader
* to the file given.
*
* @param ctxt The compilation context
* @param fname The file name
* @param reader A reader for the JSP source file
* @param err The error dispatcher
*
* @throws JasperException If an error occurs parsing the JSP file
*/
public JspReader(JspCompilationContext ctxt,
String fname,
InputStreamReader reader,
ErrorDispatcher err)
throws JasperException {
this.context = ctxt;
this.err = err;
try {
CharArrayWriter caw = new CharArrayWriter();
char buf[] = new char[1024];
for (int i = 0 ; (i = reader.read(buf)) != -1 ;)
caw.write(buf, 0, i);
caw.close();
current = new Mark(this, caw.toCharArray(), fname);
} catch (Throwable ex) {
ExceptionUtils.handleThrowable(ex);
log.error("Exception parsing file ", ex);
err.jspError("jsp.error.file.cannot.read", fname);
} finally {
if (reader != null) {
try {
reader.close();
} catch (Exception any) {
if(log.isDebugEnabled()) {
log.debug("Exception closing reader: ", any);
}
}
}
}
}
/**
* @return JSP compilation context with which this JspReader is
* associated
*/
JspCompilationContext getJspCompilationContext() {
return context;
}
/**
* Checks if the current file has more input.
*
* @return True if more reading is possible
*/
boolean hasMoreInput() {
return current.cursor < current.stream.length;
}
int nextChar() {
if (!hasMoreInput())
return -1;
int ch = current.stream[current.cursor];
current.cursor++;
if (ch == '\n') {
current.line++;
current.col = 0;
} else {
current.col++;
}
return ch;
}
/**
* A faster approach than calling {@link #mark()} & {@link #nextChar()}.
* However, this approach is only safe if the mark is only used within the
* JspReader.
*/
private int nextChar(Mark mark) {
if (!hasMoreInput()) {
return -1;
}
int ch = current.stream[current.cursor];
mark.init(current, true);
current.cursor++;
if (ch == '\n') {
current.line++;
current.col = 0;
} else {
current.col++;
}
return ch;
}
/**
* Search the given character, If it was found, then mark the current cursor
* and the cursor point to next character.
*/
private Boolean indexOf(char c, Mark mark) {
if (!hasMoreInput())
return null;
int end = current.stream.length;
int ch;
int line = current.line;
int col = current.col;
int i = current.cursor;
for(; i < end; i ++) {
ch = current.stream[i];
if (ch == c) {
mark.update(i, line, col);
}
if (ch == '\n') {
line++;
col = 0;
} else {
col++;
}
if (ch == c) {
current.update(i+1, line, col);
return Boolean.TRUE;
}
}
current.update(i, line, col);
return Boolean.FALSE;
}
/**
* Back up the current cursor by one char, assumes current.cursor > 0,
* and that the char to be pushed back is not '\n'.
*/
void pushChar() {
current.cursor--;
current.col--;
}
String getText(Mark start, Mark stop) {
Mark oldstart = mark();
reset(start);
CharArrayWriter caw = new CharArrayWriter();
while (!markEquals(stop)) {
caw.write(nextChar());
}
caw.close();
setCurrent(oldstart);
return caw.toString();
}
/**
* Read ahead one character without moving the cursor.
*
* @return The next character or -1 if no further input is available
*/
int peekChar() {
return peekChar(0);
}
/**
* Read ahead the given number of characters without moving the cursor.
*
* @param readAhead The number of characters to read ahead. NOTE: This is
* zero based.
*
* @return The requested character or -1 if the end of the input is reached
* first
*/
int peekChar(int readAhead) {
int target = current.cursor + readAhead;
if (target < current.stream.length) {
return current.stream[target];
}
return -1;
}
Mark mark() {
return new Mark(current);
}
/**
* This method avoids a call to {@link #mark()} when doing comparison.
*/
private boolean markEquals(Mark another) {
return another.equals(current);
}
void reset(Mark mark) {
current = new Mark(mark);
}
/**
* Similar to {@link #reset(Mark)} but no new Mark will be created.
* Therefore, the parameter mark must NOT be used in other places.
*/
private void setCurrent(Mark mark) {
current = mark;
}
/**
* search the stream for a match to a string
* @param string The string to match
* @return <strong>true</strong> is one is found, the current position
* in stream is positioned after the search string, <strong>
* false</strong> otherwise, position in stream unchanged.
*/
boolean matches(String string) {
int len = string.length();
int cursor = current.cursor;
int streamSize = current.stream.length;
if (cursor + len < streamSize) { //Try to scan in memory
int line = current.line;
int col = current.col;
int ch;
int i = 0;
for(; i < len; i ++) {
ch = current.stream[i+cursor];
if (string.charAt(i) != ch) {
return false;
}
if (ch == '\n') {
line ++;
col = 0;
} else {
col++;
}
}
current.update(i+cursor, line, col);
} else {
Mark mark = mark();
int ch = 0;
int i = 0;
do {
ch = nextChar();
if (((char) ch) != string.charAt(i++)) {
setCurrent(mark);
return false;
}
} while (i < len);
}
return true;
}
boolean matchesETag(String tagName) {
Mark mark = mark();
if (!matches("</" + tagName))
return false;
skipSpaces();
if (nextChar() == '>')
return true;
setCurrent(mark);
return false;
}
boolean matchesETagWithoutLessThan(String tagName) {
Mark mark = mark();
if (!matches("/" + tagName))
return false;
skipSpaces();
if (nextChar() == '>')
return true;
setCurrent(mark);
return false;
}
/**
* Looks ahead to see if there are optional spaces followed by
* the given String. If so, true is returned and those spaces and
* characters are skipped. If not, false is returned and the
* position is restored to where we were before.
*/
boolean matchesOptionalSpacesFollowedBy(String s) {
Mark mark = mark();
skipSpaces();
boolean result = matches( s );
if( !result ) {
setCurrent(mark);
}
return result;
}
int skipSpaces() {
int i = 0;
while (hasMoreInput() && isSpace()) {
i++;
nextChar();
}
return i;
}
/**
* Skip until the given string is matched in the stream.
* When returned, the context is positioned past the end of the match.
*
* @param s The String to match.
* @return A non-null <code>Mark</code> instance (positioned immediately
* before the search string) if found, <strong>null</strong>
* otherwise.
*/
Mark skipUntil(String limit) {
Mark ret = mark();
int limlen = limit.length();
char firstChar = limit.charAt(0);
Boolean result = null;
Mark restart = null;
skip:
while((result = indexOf(firstChar, ret)) != null) {
if (result.booleanValue()) {
if (restart != null) {
restart.init(current, true);
} else {
restart = mark();
}
for (int i = 1 ; i < limlen ; i++) {
if (peekChar() == limit.charAt(i)) {
nextChar();
} else {
current.init(restart, true);
continue skip;
}
}
return ret;
}
}
return null;
}
/**
* Skip until the given string is matched in the stream, but ignoring
* chars initially escaped by a '\' and any EL expressions.
* When returned, the context is positioned past the end of the match.
*
* @param s The String to match.
* @param ignoreEL <code>true</code> if something that looks like EL should
* not be treated as EL.
* @return A non-null <code>Mark</code> instance (positioned immediately
* before the search string) if found, <strong>null</strong>
* otherwise.
*/
Mark skipUntilIgnoreEsc(String limit, boolean ignoreEL) {
Mark ret = mark();
int limlen = limit.length();
int ch;
int prev = 'x'; // Doesn't matter
char firstChar = limit.charAt(0);
skip:
for (ch = nextChar(ret) ; ch != -1 ; prev = ch, ch = nextChar(ret)) {
if (ch == '\\' && prev == '\\') {
ch = 0; // Double \ is not an escape char anymore
} else if (prev == '\\') {
continue;
} else if (!ignoreEL && (ch == '$' || ch == '#') && peekChar() == '{' ) {
// Move beyond the '{'
nextChar();
skipELExpression();
} else if (ch == firstChar) {
for (int i = 1 ; i < limlen ; i++) {
if (peekChar() == limit.charAt(i))
nextChar();
else
continue skip;
}
return ret;
}
}
return null;
}
/**
* Skip until the given end tag is matched in the stream.
* When returned, the context is positioned past the end of the tag.
*
* @param tag The name of the tag whose ETag (</tag>) to match.
* @return A non-null <code>Mark</code> instance (positioned immediately
* before the ETag) if found, <strong>null</strong> otherwise.
*/
Mark skipUntilETag(String tag) {
Mark ret = skipUntil("</" + tag);
if (ret != null) {
skipSpaces();
if (nextChar() != '>')
ret = null;
}
return ret;
}
/**
* Parse ELExpressionBody that is a body of ${} or #{} expression. Initial
* reader position is expected to be just after '${' or '#{' characters.
* <p>
* In case of success, this method returns <code>Mark</code> for the last
* character before the terminating '}' and reader is positioned just after
* the '}' character. If no terminating '}' is encountered, this method
* returns <code>null</code>.
* <p>
* Starting with EL 3.0, nested paired {}s are supported.
*
* @return Mark for the last character of EL expression or <code>null</code>
*/
Mark skipELExpression() {
// ELExpressionBody.
// Starts with "#{" or "${". Ends with "}".
// May contain quoted "{", "}", '{', or '}' and nested "{...}"
Mark last = mark();
boolean singleQuoted = false;
boolean doubleQuoted = false;
int nesting = 0;
int currentChar;
do {
currentChar = nextChar(last);
while (currentChar == '\\' && (singleQuoted || doubleQuoted)) {
// skip character following '\' within quotes
// No need to update 'last', as neither of these characters
// can be the closing '}'.
nextChar();
currentChar = nextChar();
}
if (currentChar == -1) {
return null;
}
if (currentChar == '"' && !singleQuoted) {
doubleQuoted = !doubleQuoted;
} else if (currentChar == '\'' && !doubleQuoted) {
singleQuoted = !singleQuoted;
} else if (currentChar == '{' && !doubleQuoted && !singleQuoted) {
nesting++;
} else if (currentChar =='}' && !doubleQuoted && !singleQuoted) {
// Note: This also matches the terminating '}' at which point
// nesting will be set to -1 - hence the test for
// while (currentChar != '}' || nesting > -1 ||...) below
// to continue the loop until the final '}' is detected
nesting--;
}
} while (currentChar != '}' || singleQuoted || doubleQuoted || nesting > -1);
return last;
}
final boolean isSpace() {
// Note: If this logic changes, also update Node.TemplateText.rtrim()
return peekChar() <= ' ';
}
/**
* Parse a space delimited token.
* If quoted the token will consume all characters up to a matching quote,
* otherwise, it consumes up to the first delimiter character.
*
* @param quoted If <strong>true</strong> accept quoted strings.
*/
String parseToken(boolean quoted) throws JasperException {
StringBuilder StringBuilder = new StringBuilder();
skipSpaces();
StringBuilder.setLength(0);
if (!hasMoreInput()) {
return "";
}
int ch = peekChar();
if (quoted) {
if (ch == '"' || ch == '\'') {
char endQuote = ch == '"' ? '"' : '\'';
// Consume the open quote:
ch = nextChar();
for (ch = nextChar(); ch != -1 && ch != endQuote;
ch = nextChar()) {
if (ch == '\\')
ch = nextChar();
StringBuilder.append((char) ch);
}
// Check end of quote, skip closing quote:
if (ch == -1) {
err.jspError(mark(), "jsp.error.quotes.unterminated");
}
} else {
err.jspError(mark(), "jsp.error.attr.quoted");
}
} else {
if (!isDelimiter()) {
// Read value until delimiter is found:
do {
ch = nextChar();
// Take care of the quoting here.
if (ch == '\\') {
if (peekChar() == '"' || peekChar() == '\'' ||
peekChar() == '>' || peekChar() == '%')
ch = nextChar();
}
StringBuilder.append((char) ch);
} while (!isDelimiter());
}
}
return StringBuilder.toString();
}
/**
* Parse utils - Is current character a token delimiter ?
* Delimiters are currently defined to be =, &gt;, &lt;, ", and ' or any
* any space character as defined by <code>isSpace</code>.
*
* @return A boolean.
*/
private boolean isDelimiter() {
if (! isSpace()) {
int ch = peekChar();
// Look for a single-char work delimiter:
if (ch == '=' || ch == '>' || ch == '"' || ch == '\''
|| ch == '/') {
return true;
}
// Look for an end-of-comment or end-of-tag:
if (ch == '-') {
Mark mark = mark();
if (((ch = nextChar()) == '>')
|| ((ch == '-') && (nextChar() == '>'))) {
setCurrent(mark);
return true;
} else {
setCurrent(mark);
return false;
}
}
return false;
} else {
return true;
}
}
}