org.apache.juneau.json.JsonParserSession Maven / Gradle / Ivy
// ***************************************************************************************************************************
// * 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.juneau.json;
import static org.apache.juneau.internal.StringUtils.*;
import java.io.*;
import java.lang.reflect.*;
import java.util.*;
import org.apache.juneau.*;
import org.apache.juneau.internal.*;
import org.apache.juneau.parser.*;
import org.apache.juneau.transform.*;
/**
* Session object that lives for the duration of a single use of {@link JsonParser}.
*
*
* This class is NOT thread safe.
* It is typically discarded after one-time use although it can be reused against multiple inputs.
*/
@SuppressWarnings({ "unchecked", "rawtypes" })
public final class JsonParserSession extends ReaderParserSession {
private static final AsciiSet decChars = new AsciiSet("0123456789");
/**
* Create a new session using properties specified in the context.
*
* @param ctx
* The context creating this session object.
* The context contains all the configuration settings for this object.
* @param args
* Runtime session arguments.
*/
protected JsonParserSession(JsonParserContext ctx, ParserSessionArgs args) {
super(ctx, args);
}
/**
* Returns true if the specified character is whitespace.
*
*
* The definition of whitespace is different for strict vs lax mode.
* Strict mode only interprets 0x20 (space), 0x09 (tab), 0x0A (line feed) and 0x0D (carriage return) as whitespace.
* Lax mode uses {@link Character#isWhitespace(int)} to make the determination.
*
* @param cp The codepoint.
* @return true if the specified character is whitespace.
*/
protected final boolean isWhitespace(int cp) {
if (isStrict())
return cp <= 0x20 && (cp == 0x09 || cp == 0x0A || cp == 0x0D || cp == 0x20);
return Character.isWhitespace(cp);
}
/**
* Returns true if the specified character is whitespace or '/'.
*
* @param cp The codepoint.
* @return true if the specified character is whitespace or '/'.
*/
protected final boolean isCommentOrWhitespace(int cp) {
if (cp == '/')
return true;
if (isStrict())
return cp <= 0x20 && (cp == 0x09 || cp == 0x0A || cp == 0x0D || cp == 0x20);
return Character.isWhitespace(cp);
}
@Override /* ParserSession */
protected T doParse(ParserPipe pipe, ClassMeta type) throws Exception {
ParserReader r = pipe.getParserReader();
if (r == null)
return null;
T o = parseAnything(type, r, getOuter(), null);
validateEnd(r);
return o;
}
@Override /* ReaderParserSession */
protected Map doParseIntoMap(ParserPipe pipe, Map m, Type keyType, Type valueType) throws Exception {
ParserReader r = pipe.getParserReader();
m = parseIntoMap2(r, m, (ClassMeta)getClassMeta(keyType), (ClassMeta)getClassMeta(valueType), null);
validateEnd(r);
return m;
}
@Override /* ReaderParserSession */
protected Collection doParseIntoCollection(ParserPipe pipe, Collection c, Type elementType) throws Exception {
ParserReader r = pipe.getParserReader();
c = parseIntoCollection2(r, c, getClassMeta(elementType), null);
validateEnd(r);
return c;
}
private T parseAnything(ClassMeta eType, ParserReader r, Object outer, BeanPropertyMeta pMeta) throws Exception {
if (eType == null)
eType = (ClassMeta)object();
PojoSwap swap = (PojoSwap)eType.getPojoSwap(this);
ClassMeta sType = swap == null ? eType : swap.getSwapClassMeta(this);
setCurrentClass(sType);
String wrapperAttr = sType.getExtendedMeta(JsonClassMeta.class).getWrapperAttr();
Object o = null;
skipCommentsAndSpace(r);
if (wrapperAttr != null)
skipWrapperAttrStart(r, wrapperAttr);
int c = r.peek();
if (c == -1) {
if (isStrict())
throw new ParseException(r.getLocation(this), "Empty input.");
// Let o be null.
} else if ((c == ',' || c == '}' || c == ']')) {
if (isStrict())
throw new ParseException(r.getLocation(this), "Missing value detected.");
// Handle bug in Cognos 10.2.1 that can product non-existent values.
// Let o be null;
} else if (c == 'n') {
parseKeyword("null", r);
} else if (sType.isObject()) {
if (c == '{') {
ObjectMap m2 = new ObjectMap(this);
parseIntoMap2(r, m2, string(), object(), pMeta);
o = cast(m2, pMeta, eType);
} else if (c == '[') {
o = parseIntoCollection2(r, new ObjectList(this), object(), pMeta);
} else if (c == '\'' || c == '"') {
o = parseString(r);
if (sType.isChar())
o = o.toString().charAt(0);
} else if (c >= '0' && c <= '9' || c == '-' || c == '.') {
o = parseNumber(r, null);
} else if (c == 't') {
parseKeyword("true", r);
o = Boolean.TRUE;
} else {
parseKeyword("false", r);
o = Boolean.FALSE;
}
} else if (sType.isBoolean()) {
o = parseBoolean(r);
} else if (sType.isCharSequence()) {
o = parseString(r);
} else if (sType.isChar()) {
o = parseString(r).charAt(0);
} else if (sType.isNumber()) {
o = parseNumber(r, (Class)sType.getInnerClass());
} else if (sType.isMap()) {
Map m = (sType.canCreateNewInstance(outer) ? (Map)sType.newInstance(outer) : new ObjectMap(this));
o = parseIntoMap2(r, m, sType.getKeyType(), sType.getValueType(), pMeta);
} else if (sType.isCollection()) {
if (c == '{') {
ObjectMap m = new ObjectMap(this);
parseIntoMap2(r, m, string(), object(), pMeta);
o = cast(m, pMeta, eType);
} else {
Collection l = (sType.canCreateNewInstance(outer) ? (Collection)sType.newInstance() : new ObjectList(this));
o = parseIntoCollection2(r, l, sType, pMeta);
}
} else if (sType.canCreateNewBean(outer)) {
BeanMap m = newBeanMap(outer, sType.getInnerClass());
o = parseIntoBeanMap2(r, m).getBean();
} else if (sType.canCreateNewInstanceFromString(outer) && (c == '\'' || c == '"')) {
o = sType.newInstanceFromString(outer, parseString(r));
} else if (sType.canCreateNewInstanceFromNumber(outer) && isFirstNumberChar((char)c)) {
o = sType.newInstanceFromNumber(this, outer, parseNumber(r, sType.getNewInstanceFromNumberClass()));
} else if (sType.isArray() || sType.isArgs()) {
if (c == '{') {
ObjectMap m = new ObjectMap(this);
parseIntoMap2(r, m, string(), object(), pMeta);
o = cast(m, pMeta, eType);
} else {
ArrayList l = (ArrayList)parseIntoCollection2(r, new ArrayList(), sType, pMeta);
o = toArray(sType, l);
}
} else if (c == '{') {
Map m = new ObjectMap(this);
parseIntoMap2(r, m, sType.getKeyType(), sType.getValueType(), pMeta);
if (m.containsKey(getBeanTypePropertyName(eType)))
o = cast((ObjectMap)m, pMeta, eType);
else
throw new ParseException(r.getLocation(this), "Class ''{0}'' could not be instantiated. Reason: ''{1}''",
sType.getInnerClass().getName(), sType.getNotABeanReason());
} else if (sType.canCreateNewInstanceFromString(outer) && ! isStrict()) {
o = sType.newInstanceFromString(outer, parseString(r));
} else {
throw new ParseException(r.getLocation(this), "Unrecognized syntax for class type ''{0}'', starting character ''{1}''",
sType, (char)c);
}
if (wrapperAttr != null)
skipWrapperAttrEnd(r);
if (swap != null && o != null)
o = swap.unswap(this, o, eType);
if (outer != null)
setParent(eType, o, outer);
return (T)o;
}
private Number parseNumber(ParserReader r, Class type) throws Exception {
int c = r.peek();
if (c == '\'' || c == '"')
return parseNumber(r, parseString(r), type);
return parseNumber(r, parseNumberString(r), type);
}
private Number parseNumber(ParserReader r, String s, Class type) throws Exception {
// JSON has slightly different number rules from Java.
// Strict mode enforces these different rules, lax does not.
if (isStrict()) {
// Lax allows blank strings to represent 0.
// Strict does not allow blank strings.
if (s.length() == 0)
throw new ParseException(r.getLocation(this), "Invalid JSON number: ''{0}''", s);
// Need to weed out octal and hexadecimal formats: 0123,-0123,0x123,-0x123.
// Don't weed out 0 or -0.
boolean isNegative = false;
char c = s.charAt(0);
if (c == '-') {
isNegative = true;
c = (s.length() == 1 ? 'x' : s.charAt(1));
}
// JSON doesn't allow '.123' and '-.123'.
if (c == '.')
throw new ParseException(loc(r), "Invalid JSON number: ''{0}''", s);
// '01' is not a valid number, but '0.1', '0e1', '0e+1' are valid.
if (c == '0' && s.length() > (isNegative ? 2 : 1)) {
char c2 = s.charAt((isNegative ? 2 : 1));
if (c2 != '.' && c2 != 'e' && c2 != 'E')
throw new ParseException(loc(r), "Invalid JSON number: ''{0}''", s);
}
// JSON doesn't allow '1.' or '0.e1'.
int i = s.indexOf('.');
if (i != -1 && (s.length() == (i+1) || ! decChars.contains(s.charAt(i+1))))
throw new ParseException(loc(r), "Invalid JSON number: ''{0}''", s);
}
return StringUtils.parseNumber(s, type);
}
private Boolean parseBoolean(ParserReader r) throws Exception {
int c = r.peek();
if (c == '\'' || c == '"')
return Boolean.valueOf(parseString(r));
if (c == 't') {
parseKeyword("true", r);
return Boolean.TRUE;
}
parseKeyword("false", r);
return Boolean.FALSE;
}
private Map parseIntoMap2(ParserReader r, Map m, ClassMeta keyType,
ClassMeta valueType, BeanPropertyMeta pMeta) throws Exception {
if (keyType == null)
keyType = (ClassMeta)string();
int S0=0; // Looking for outer {
int S1=1; // Looking for attrName start.
int S3=3; // Found attrName end, looking for :.
int S4=4; // Found :, looking for valStart: { [ " ' LITERAL.
int S5=5; // Looking for , or }
int S6=6; // Found , looking for attr start.
int state = S0;
String currAttr = null;
int c = 0;
while (c != -1) {
c = r.read();
if (state == S0) {
if (c == '{')
state = S1;
} else if (state == S1) {
if (c == '}') {
return m;
} else if (isCommentOrWhitespace(c)) {
skipCommentsAndSpace(r.unread());
} else {
currAttr = parseFieldName(r.unread());
state = S3;
}
} else if (state == S3) {
if (c == ':')
state = S4;
} else if (state == S4) {
if (isCommentOrWhitespace(c)) {
skipCommentsAndSpace(r.unread());
} else {
K key = convertAttrToType(m, currAttr, keyType);
V value = parseAnything(valueType, r.unread(), m, pMeta);
setName(valueType, value, key);
m.put(key, value);
state = S5;
}
} else if (state == S5) {
if (c == ',')
state = S6;
else if (isCommentOrWhitespace(c))
skipCommentsAndSpace(r.unread());
else if (c == '}') {
return m;
} else {
break;
}
} else if (state == S6) {
if (c == '}') {
break;
} else if (isCommentOrWhitespace(c)) {
skipCommentsAndSpace(r.unread());
} else {
currAttr = parseFieldName(r.unread());
state = S3;
}
}
}
if (state == S0)
throw new ParseException(loc(r), "Expected '{' at beginning of JSON object.");
if (state == S1)
throw new ParseException(loc(r), "Could not find attribute name on JSON object.");
if (state == S3)
throw new ParseException(loc(r), "Could not find ':' following attribute name on JSON object.");
if (state == S4)
throw new ParseException(loc(r), "Expected one of the following characters: {,[,',\",LITERAL.");
if (state == S5)
throw new ParseException(loc(r), "Could not find '}' marking end of JSON object.");
if (state == S6)
throw new ParseException(loc(r), "Unexpected '}' found in JSON object.");
return null; // Unreachable.
}
/*
* Parse a JSON attribute from the character array at the specified position, then
* set the position marker to the last character in the field name.
*/
private String parseFieldName(ParserReader r) throws Exception {
int c = r.peek();
if (c == '\'' || c == '"')
return parseString(r);
if (isStrict())
throw new ParseException(loc(r), "Unquoted attribute detected.");
r.mark();
// Look for whitespace.
while (c != -1) {
c = r.read();
if (c == ':' || isWhitespace(c) || c == '/') {
r.unread();
String s = r.getMarked().intern();
return s.equals("null") ? null : s;
}
}
throw new ParseException(loc(r), "Could not find the end of the field name.");
}
private Collection parseIntoCollection2(ParserReader r, Collection l,
ClassMeta type, BeanPropertyMeta pMeta) throws Exception {
int S0=0; // Looking for outermost [
int S1=1; // Looking for starting [ or { or " or ' or LITERAL or ]
int S2=2; // Looking for , or ]
int S3=3; // Looking for starting [ or { or " or ' or LITERAL
int argIndex = 0;
int state = S0;
int c = 0;
while (c != -1) {
c = r.read();
if (state == S0) {
if (c == '[')
state = S1;
} else if (state == S1) {
if (c == ']') {
return l;
} else if (isCommentOrWhitespace(c)) {
skipCommentsAndSpace(r.unread());
} else if (c != -1) {
l.add((E)parseAnything(type.isArgs() ? type.getArg(argIndex++) : type.getElementType(), r.unread(), l, pMeta));
state = S2;
}
} else if (state == S2) {
if (c == ',') {
state = S3;
} else if (isCommentOrWhitespace(c)) {
skipCommentsAndSpace(r.unread());
} else if (c == ']') {
return l;
} else {
break; // Invalid character found.
}
} else if (state == S3) {
if (isCommentOrWhitespace(c)) {
skipCommentsAndSpace(r.unread());
} else if (c == ']') {
break;
} else if (c != -1) {
l.add((E)parseAnything(type.isArgs() ? type.getArg(argIndex++) : type.getElementType(), r.unread(), l, pMeta));
state = S2;
}
}
}
if (state == S0)
throw new ParseException(loc(r), "Expected '[' at beginning of JSON array.");
if (state == S1)
throw new ParseException(loc(r), "Expected one of the following characters: {,[,',\",LITERAL.");
if (state == S2)
throw new ParseException(loc(r), "Expected ',' or ']'.");
if (state == S3)
throw new ParseException(loc(r), "Unexpected trailing comma in array.");
return null; // Unreachable.
}
private BeanMap parseIntoBeanMap2(ParserReader r, BeanMap m) throws Exception {
int S0=0; // Looking for outer {
int S1=1; // Looking for attrName start.
int S3=3; // Found attrName end, looking for :.
int S4=4; // Found :, looking for valStart: { [ " ' LITERAL.
int S5=5; // Looking for , or }
int state = S0;
String currAttr = "";
int c = 0;
int currAttrLine = -1, currAttrCol = -1;
while (c != -1) {
c = r.read();
if (state == S0) {
if (c == '{')
state = S1;
} else if (state == S1) {
if (c == '}') {
return m;
} else if (isCommentOrWhitespace(c)) {
skipCommentsAndSpace(r.unread());
} else {
r.unread();
currAttrLine= r.getLine();
currAttrCol = r.getColumn();
currAttr = parseFieldName(r);
state = S3;
}
} else if (state == S3) {
if (c == ':')
state = S4;
} else if (state == S4) {
if (isCommentOrWhitespace(c)) {
skipCommentsAndSpace(r.unread());
} else {
if (! currAttr.equals(getBeanTypePropertyName(m.getClassMeta()))) {
BeanPropertyMeta pMeta = m.getPropertyMeta(currAttr);
setCurrentProperty(pMeta);
if (pMeta == null) {
onUnknownProperty(r.getPipe(), currAttr, m, currAttrLine, currAttrCol);
parseAnything(object(), r.unread(), m.getBean(false), null); // Read content anyway to ignore it
} else {
ClassMeta cm = pMeta.getClassMeta();
Object value = parseAnything(cm, r.unread(), m.getBean(false), pMeta);
setName(cm, value, currAttr);
pMeta.set(m, currAttr, value);
}
setCurrentProperty(null);
}
state = S5;
}
} else if (state == S5) {
if (c == ',')
state = S1;
else if (isCommentOrWhitespace(c))
skipCommentsAndSpace(r.unread());
else if (c == '}') {
return m;
}
}
}
if (state == S0)
throw new ParseException(loc(r), "Expected '{' at beginning of JSON object.");
if (state == S1)
throw new ParseException(loc(r), "Could not find attribute name on JSON object.");
if (state == S3)
throw new ParseException(loc(r), "Could not find ':' following attribute name on JSON object.");
if (state == S4)
throw new ParseException(loc(r), "Expected one of the following characters: {,[,',\",LITERAL.");
if (state == S5)
throw new ParseException(loc(r), "Could not find '}' marking end of JSON object.");
return null; // Unreachable.
}
/*
* Starting from the specified position in the character array, returns the
* position of the character " or '.
* If the string consists of a concatenation of strings (e.g. 'AAA' + "BBB"), this method
* will automatically concatenate the strings and return the result.
*/
private String parseString(ParserReader r) throws Exception {
r.mark();
int qc = r.read(); // The quote character being used (" or ')
if (qc != '"' && isStrict()) {
String msg = (
qc == '\''
? "Invalid quote character \"{0}\" being used."
: "Did not find quote character marking beginning of string. Character=\"{0}\""
);
throw new ParseException(loc(r), msg, (char)qc);
}
final boolean isQuoted = (qc == '\'' || qc == '"');
String s = null;
boolean isInEscape = false;
int c = 0;
while (c != -1) {
c = r.read();
// Strict syntax requires that all control characters be escaped.
if (isStrict() && c <= 0x1F)
throw new ParseException("Unescaped control character encountered: ''0x{0}''", String.format("%04X", c));
if (isInEscape) {
switch (c) {
case 'n': r.replace('\n'); break;
case 'r': r.replace('\r'); break;
case 't': r.replace('\t'); break;
case 'f': r.replace('\f'); break;
case 'b': r.replace('\b'); break;
case '\\': r.replace('\\'); break;
case '/': r.replace('/'); break;
case '\'': r.replace('\''); break;
case '"': r.replace('"'); break;
case 'u': {
String n = r.read(4);
try {
r.replace(Integer.parseInt(n, 16), 6);
} catch (NumberFormatException e) {
throw new ParseException(loc(r), "Invalid Unicode escape sequence in string.");
}
break;
}
default:
throw new ParseException(loc(r), "Invalid escape sequence in string.");
}
isInEscape = false;
} else {
if (c == '\\') {
isInEscape = true;
r.delete();
} else if (isQuoted) {
if (c == qc) {
s = r.getMarked(1, -1);
break;
}
} else {
if (c == ',' || c == '}' || c == ']' || isWhitespace(c)) {
s = r.getMarked(0, -1);
r.unread();
break;
} else if (c == -1) {
s = r.getMarked(0, 0);
break;
}
}
}
}
if (s == null)
throw new ParseException(loc(r), "Could not find expected end character ''{0}''.", (char)qc);
// Look for concatenated string (i.e. whitespace followed by +).
skipCommentsAndSpace(r);
if (r.peek() == '+') {
if (isStrict())
throw new ParseException(loc(r), "String concatenation detected.");
r.read(); // Skip past '+'
skipCommentsAndSpace(r);
s += parseString(r);
}
return trim(s); // End of input reached.
}
/*
* Looks for the keywords true, false, or null.
* Throws an exception if any of these keywords are not found at the specified position.
*/
private void parseKeyword(String keyword, ParserReader r) throws Exception {
try {
String s = r.read(keyword.length());
if (s.equals(keyword))
return;
throw new ParseException(loc(r), "Unrecognized syntax.");
} catch (IndexOutOfBoundsException e) {
throw new ParseException(loc(r), "Unrecognized syntax.");
}
}
/*
* Doesn't actually parse anything, but moves the position beyond any whitespace or comments.
* If positionOnNext is 'true', then the cursor will be set to the point immediately after
* the comments and whitespace. Otherwise, the cursor will be set to the last position of
* the comments and whitespace.
*/
private void skipCommentsAndSpace(ParserReader r) throws Exception {
int c = 0;
while ((c = r.read()) != -1) {
if (! isWhitespace(c)) {
if (c == '/') {
if (isStrict())
throw new ParseException(loc(r), "Javascript comment detected.");
skipComments(r);
} else {
r.unread();
return;
}
}
}
}
/*
* Doesn't actually parse anything, but moves the position beyond the construct "{wrapperAttr:" when
* the @Json.wrapperAttr() annotation is used on a class.
*/
private void skipWrapperAttrStart(ParserReader r, String wrapperAttr) throws Exception {
int S0=0; // Looking for outer {
int S1=1; // Looking for attrName start.
int S3=3; // Found attrName end, looking for :.
int S4=4; // Found :, looking for valStart: { [ " ' LITERAL.
int state = S0;
String currAttr = null;
int c = 0;
while (c != -1) {
c = r.read();
if (state == S0) {
if (c == '{')
state = S1;
} else if (state == S1) {
if (isCommentOrWhitespace(c)) {
skipCommentsAndSpace(r.unread());
} else {
currAttr = parseFieldName(r.unread());
if (! currAttr.equals(wrapperAttr))
throw new ParseException(loc(r),
"Expected to find wrapper attribute ''{0}'' but found attribute ''{1}''", wrapperAttr, currAttr);
state = S3;
}
} else if (state == S3) {
if (c == ':')
state = S4;
} else if (state == S4) {
if (isCommentOrWhitespace(c)) {
skipCommentsAndSpace(r.unread());
} else {
r.unread();
return;
}
}
}
if (state == S0)
throw new ParseException(loc(r), "Expected '{' at beginning of JSON object.");
if (state == S1)
throw new ParseException(loc(r), "Could not find attribute name on JSON object.");
if (state == S3)
throw new ParseException(loc(r), "Could not find ':' following attribute name on JSON object.");
if (state == S4)
throw new ParseException(loc(r), "Expected one of the following characters: {,[,',\",LITERAL.");
}
/*
* Doesn't actually parse anything, but moves the position beyond the construct "}" when
* the @Json.wrapperAttr() annotation is used on a class.
*/
private void skipWrapperAttrEnd(ParserReader r) throws ParseException, IOException {
int c = 0;
while ((c = r.read()) != -1) {
if (! isWhitespace(c)) {
if (c == '/') {
if (isStrict())
throw new ParseException(loc(r), "Javascript comment detected.");
skipComments(r);
} else if (c == '}') {
return;
} else {
throw new ParseException(loc(r), "Could not find '}' at the end of JSON wrapper object.");
}
}
}
}
/*
* Doesn't actually parse anything, but when positioned at the beginning of comment,
* it will move the pointer to the last character in the comment.
*/
private void skipComments(ParserReader r) throws ParseException, IOException {
int c = r.read();
// "/* */" style comments
if (c == '*') {
while (c != -1)
if ((c = r.read()) == '*')
if ((c = r.read()) == '/')
return;
// "//" style comments
} else if (c == '/') {
while (c != -1) {
c = r.read();
if (c == -1 || c == '\n')
return;
}
}
throw new ParseException(loc(r), "Open ended comment.");
}
/*
* Call this method after you've finished a parsing a string to make sure that if there's any
* remainder in the input, that it consists only of whitespace and comments.
*/
private void validateEnd(ParserReader r) throws Exception {
skipCommentsAndSpace(r);
int c = r.read();
if (c != -1 && c != ';') // var x = {...}; expressions can end with a semicolon.
throw new ParseException(loc(r), "Remainder after parse: ''{0}''.", (char)c);
}
private ObjectMap loc(ParserReader r) {
return getLastLocation().append("line", r.getLine()).append("column", r.getColumn());
}
}