/*
* Copyright 2008 The Closure Compiler Authors.
*
* Licensed 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 com.google.javascript.jscomp.deps;
import com.google.common.annotations.GwtIncompatible;
import com.google.common.base.CharMatcher;
import com.google.javascript.jscomp.CheckLevel;
import com.google.javascript.jscomp.DiagnosticType;
import com.google.javascript.jscomp.ErrorManager;
import com.google.javascript.jscomp.JSError;
import java.io.BufferedReader;
import java.io.IOException;
import java.io.Reader;
import java.util.ArrayList;
import java.util.LinkedHashMap;
import java.util.List;
import java.util.Map;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
/**
* Base class for classes that parse JavaScript sources on a line-by-line basis. Strips comments
* from files and records all parsing errors.
*
* @author agrieve@google.com (Andrew Grieve)
*/
@GwtIncompatible("java.io")
public abstract class JsFileLineParser {
static final DiagnosticType PARSE_WARNING = DiagnosticType.warning(
"DEPS_PARSE_WARNING", "{0}\n{1}");
public static final DiagnosticType PARSE_ERROR = DiagnosticType.error(
"DEPS_PARSE_ERROR", "{0}\n{1}");
boolean shortcutMode = false;
/**
* Thrown by base classes to signify a problem parsing a line.
*/
static class ParseException extends Exception {
public static final long serialVersionUID = 1L;
private final boolean fatal;
/**
* Constructor.
*
* @param message A description of what caused the exception.
* @param fatal Whether the exception is recoverable.
*/
public ParseException(String message, boolean fatal) {
super(message);
this.fatal = fatal;
}
public boolean isFatal() {
return fatal;
}
}
/** Pattern for matching JavaScript string literals. */
private static final Pattern STRING_LITERAL_PATTERN = Pattern.compile(
"\\s*(?:'((?:\\\\'|[^'])*?)'|\"((?:\\\\\"|[^\"])*?)\")\\s*");
/** Matcher used in the parsing string literals. */
private final Matcher valueMatcher = STRING_LITERAL_PATTERN.matcher("");
/** Path of the file currently being parsed. */
String filePath;
/** The line number of the line currently being parsed. */
int lineNum;
/** Handles error messages. */
ErrorManager errorManager;
/** Did our parse succeed. */
boolean parseSucceeded;
/**
* Constructor.
*
* @param errorManager Parse error handler.
*/
public JsFileLineParser(ErrorManager errorManager) {
this.errorManager = errorManager;
}
/**
* In shortcut mode, the file line parser can stop reading early if
* it thinks it found enough information.
*
* For example, many parsers assume that dependency information never
* shows up after "real" code.
*/
public void setShortcutMode(boolean mode) {
this.shortcutMode = mode;
}
public boolean didParseSucceed() {
return parseSucceeded;
}
/**
* Performs the line-by-line parsing of the given fileContents. This method
* strips out JavaScript comments and then uses the abstract parseLine()
* method to do the line parsing.
*
* @param filePath The path to the file being parsed. Used for reporting parse
* exceptions.
* @param fileContents A reader for the contents of the file.
*/
void doParse(String filePath, Reader fileContents) {
this.filePath = filePath;
parseSucceeded = true;
BufferedReader lineBuffer = new BufferedReader(fileContents);
// Parse all lines.
String line = null;
lineNum = 0;
boolean inMultilineComment = false;
try {
while (null != (line = lineBuffer.readLine())) {
++lineNum;
try {
String revisedLine = line;
if (inMultilineComment) {
int endOfComment = revisedLine.indexOf("*/");
if (endOfComment != -1) {
revisedLine = revisedLine.substring(endOfComment + 2);
inMultilineComment = false;
} else {
revisedLine = "";
}
}
if (!inMultilineComment) {
while (true) {
int startOfLineComment = revisedLine.indexOf("//");
int startOfMultilineComment = revisedLine.indexOf("/*");
if (startOfLineComment != -1 &&
(startOfMultilineComment == -1 ||
startOfLineComment < startOfMultilineComment)) {
revisedLine = revisedLine.substring(0, startOfLineComment);
break;
} else if (startOfMultilineComment != -1) {
int endOfMultilineComment = revisedLine.indexOf("*/",
startOfMultilineComment + 2);
if (endOfMultilineComment == -1) {
revisedLine = revisedLine.substring(
0, startOfMultilineComment);
inMultilineComment = true;
break;
} else {
revisedLine =
revisedLine.substring(0, startOfMultilineComment) +
revisedLine.substring(endOfMultilineComment + 2);
}
} else {
break;
}
}
}
if (!revisedLine.isEmpty()) {
// This check for shortcut mode should be redundant, but
// it's done for safety reasons.
if (!parseLine(revisedLine) && shortcutMode) {
break;
}
}
} catch (ParseException e) {
// Inform the error handler of the exception.
errorManager.report(
e.isFatal() ? CheckLevel.ERROR : CheckLevel.WARNING,
JSError.make(filePath, lineNum, 0 /* char offset */,
e.isFatal() ? PARSE_ERROR : PARSE_WARNING,
e.getMessage(), line));
parseSucceeded = parseSucceeded && !e.isFatal();
}
}
} catch (IOException e) {
errorManager.report(CheckLevel.ERROR,
JSError.make(filePath, 0, 0 /* char offset */,
PARSE_ERROR, "Error reading file: " + filePath));
parseSucceeded = false;
}
}
/**
* Called for each line of the file being parsed.
*
* @param line The line to parse.
* @return true to keep going, false otherwise.
* @throws ParseException Should be thrown to signify a problem with the line.
*/
abstract boolean parseLine(String line) throws ParseException;
/**
* Parses a JS string literal.
*
* @param jsStringLiteral The literal. Must look like "asdf" or 'asdf'
* @throws ParseException Thrown if there is a string literal that cannot be
* parsed.
*/
String parseJsString(String jsStringLiteral) throws ParseException {
valueMatcher.reset(jsStringLiteral);
if (!valueMatcher.matches()) {
throw new ParseException("Syntax error in JS String literal", true /* fatal */);
}
return valueMatcher.group(1) != null ? valueMatcher.group(1) : valueMatcher.group(2);
}
/**
* Parses a JavaScript array of string literals. (eg: ['a', 'b', "c"]).
* @param input A string containing a JavaScript array of string literals.
* @return A list of parsed string literals.
* @throws ParseException Thrown if there is a syntax error with the input.
*/
List<String> parseJsStringArray(String input)
throws ParseException {
List<String> results = new ArrayList<>();
int indexStart = input.indexOf('[');
int indexEnd = input.lastIndexOf(']');
if ((indexStart == -1) || (indexEnd == -1)) {
throw new ParseException("Syntax error when parsing JS array", true /* fatal */);
}
String innerValues = input.substring(indexStart + 1, indexEnd);
if (!innerValues.trim().isEmpty()) {
valueMatcher.reset(innerValues);
for (;;) {
// Parse the current string literal.
if (!valueMatcher.lookingAt()) {
throw new ParseException("Syntax error in JS String literal", true /* fatal */);
}
// Add it to the results.
results.add(valueMatcher.group(1) != null ?
valueMatcher.group(1) : valueMatcher.group(2));
if (valueMatcher.hitEnd()) {
break;
}
// Ensure there is a comma after the value.
if (innerValues.charAt(valueMatcher.end()) != ',') {
throw new ParseException("Missing comma in string array", true /* fatal */);
}
// Move to the next value.
valueMatcher.region(valueMatcher.end() + 1, valueMatcher.regionEnd());
}
}
return results;
}
// TODO(sdh): Consider simplifying this by reusing the parser or a separate JSON library.
/**
* Parses a JavaScript map of string literals. (eg: {'a': 'b', "c": "d"}).
* @param input A string containing a JavaScript map of string literals.
* @return A map of parsed string literals.
* @throws ParseException Thrown if there is a syntax error with the input.
*/
Map<String, String> parseJsStringMap(String input) throws ParseException {
input = CharMatcher.whitespace().trimFrom(input);
check(
!input.isEmpty() && input.charAt(0) == '{' && input.charAt(input.length() - 1) == '}',
"Syntax error when parsing JS object");
input = input.substring(1, input.length() - 1).trim();
Map<String, String> results = new LinkedHashMap<>();
boolean done = input.isEmpty();
valueMatcher.reset(input);
while (!done) {
// Parse the next key (TODO(sdh): need to support non-quoted keys?).
check(valueMatcher.lookingAt(), "Bad key in JS object literal");
String key = valueMatcher.group(1) != null ? valueMatcher.group(1) : valueMatcher.group(2);
check(!valueMatcher.hitEnd(), "Missing value in JS object literal");
check(input.charAt(valueMatcher.end()) == ':', "Missing colon in JS object literal");
valueMatcher.region(valueMatcher.end() + 1, valueMatcher.regionEnd());
// Parse the corresponding value.
check(valueMatcher.lookingAt(), "Bad value in JS object literal");
String val = valueMatcher.group(1) != null ? valueMatcher.group(1) : valueMatcher.group(2);
results.put(key, val);
if (!valueMatcher.hitEnd()) {
check(input.charAt(valueMatcher.end()) == ',', "Missing comma in JS object literal");
valueMatcher.region(valueMatcher.end() + 1, valueMatcher.regionEnd());
} else {
done = true;
}
}
return results;
}
private static void check(boolean condition, String message) throws ParseException {
if (!condition) {
throw new ParseException(message, true /* fatal */);
}
}
}