Many resources are needed to download a project. Please understand that we have to compensate our server costs. Thank you in advance.
Project price only 1 $
You can buy this project and download/modify it how often you want.
net.sourceforge.plantuml.jasic.Jasic Maven / Gradle / Ivy
PlantUML is a component that allows to quickly write :
* sequence diagram,
* use case diagram,
* class diagram,
* activity diagram,
* component diagram,
* state diagram
* object diagram
package net.sourceforge.plantuml.jasic;
import java.io.BufferedReader;
import java.io.IOException;
import java.io.InputStreamReader;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
/*
Jasic uses the MIT License:
Copyright (c) 2010 Robert Nystrom
Permission is hereby granted, free of charge, to any person obtaining a copy of
this software and associated documentation files (the "Software"), to deal in
the Software without restriction, including without limitation the rights to
use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of
the Software, and to permit persons to whom the Software is furnished to do so,
subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS
FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR
COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER
IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
*
*/
/*
* This defines a single class that contains an entire interpreter for a language very similar to the original BASIC.
* Everything is here (albeit in very simplified form): tokenizing, parsing, and interpretation. The file is organized
* in phases, with each appearing roughly in the order that they occur when a program is run. You should be able to read
* this top-down to walk through the entire process of loading and running a program.
*
* Jasic language syntax ---------------------
*
* Comments start with ' and proceed to the end of the line:
*
* print "hi there" ' this is a comment
*
* Numbers and strings are supported. Strings should be in "double quotes", and only positive integers can be parsed
* (though numbers are double internally).
*
* Variables are identified by name which must start with a letter and can contain letters or numbers. Case is
* significant for names and keywords.
*
* Each statement is on its own line. Optionally, a line may have a label before the statement. A label is a name that
* ends with a colon:
*
* foo:
*
*
* The following statements are supported:
*
* = Evaluates the expression and assigns the result to the given named variable. All variables are
* globally scoped.
*
* pi = (314159 / 10000)
*
* print Evaluates the expression and prints the result.
*
* print "hello, " + "world"
*
* input Reads in a line of input from the user and stores it in the variable with the given name.
*
* input guess
*
* goto Jumps to the statement after the label with the given name.
*
* goto loop
*
* if then Evaluates the expression. If it evaluates to a non-zero number, then jumps to the
* statement after the given label.
*
* if a < b then dosomething
*
*
* The following expressions are supported:
*
* = Evaluates to 1 if the two expressions are equal, 0 otherwise.
*
* + If the left-hand expression is a number, then adds the two expressions, otherwise
* concatenates the two strings.
*
* - * / <
* > You can figure it out.
*
* A name in an expression simply returns the value of the variable with that name. If the variable was never
* set, it defaults to 0.
*
* All binary operators have the same precedence. Sorry, I had to cut corners somewhere.
*
* To keep things simple, I've omitted some stuff or hacked things a bit. When possible, I'll leave a "HACK" note there
* explaining what and why. If you make your own interpreter, you'll want to address those.
*
* @author Bob Nystrom
*/
public class Jasic {
// Tokenizing (lexing) -----------------------------------------------------
/**
* This function takes a script as a string of characters and chunks it into a sequence of tokens. Each token is a
* meaningful unit of program, like a variable name, a number, a string, or an operator.
*/
private static List tokenize(String source) {
List tokens = new ArrayList<>();
String token = "";
TokenizeState state = TokenizeState.DEFAULT;
// Many tokens are a single character, like operators and ().
String charTokens = "\n=+-*/<>()";
TokenType[] tokenTypes = { TokenType.LINE, TokenType.EQUALS, TokenType.OPERATOR, TokenType.OPERATOR,
TokenType.OPERATOR, TokenType.OPERATOR, TokenType.OPERATOR, TokenType.OPERATOR, TokenType.LEFT_PAREN,
TokenType.RIGHT_PAREN };
// Scan through the code one character at a time, building up the list
// of tokens.
for (int i = 0; i < source.length(); i++) {
char c = source.charAt(i);
switch (state) {
case DEFAULT:
if (charTokens.indexOf(c) != -1) {
tokens.add(new Token(Character.toString(c), tokenTypes[charTokens.indexOf(c)]));
} else if (Character.isLetter(c)) {
token += c;
state = TokenizeState.WORD;
} else if (Character.isDigit(c)) {
token += c;
state = TokenizeState.NUMBER;
} else if (c == '"') {
state = TokenizeState.STRING;
} else if (c == '\'') {
state = TokenizeState.COMMENT;
}
break;
case WORD:
if (Character.isLetterOrDigit(c)) {
token += c;
} else if (c == ':') {
tokens.add(new Token(token, TokenType.LABEL));
token = "";
state = TokenizeState.DEFAULT;
} else {
tokens.add(new Token(token, TokenType.WORD));
token = "";
state = TokenizeState.DEFAULT;
i--; // Reprocess this character in the default state.
}
break;
case NUMBER:
// HACK: Negative numbers and floating points aren't supported.
// To get a negative number, just do 0 - .
// To get a floating point, divide.
if (Character.isDigit(c)) {
token += c;
} else {
tokens.add(new Token(token, TokenType.NUMBER));
token = "";
state = TokenizeState.DEFAULT;
i--; // Reprocess this character in the default state.
}
break;
case STRING:
if (c == '"') {
tokens.add(new Token(token, TokenType.STRING));
token = "";
state = TokenizeState.DEFAULT;
} else {
token += c;
}
break;
case COMMENT:
if (c == '\n') {
state = TokenizeState.DEFAULT;
}
break;
}
}
// HACK: Silently ignore any in-progress token when we run out of
// characters. This means that, for example, if a script has a string
// that's missing the closing ", it will just ditch it.
return tokens;
}
// Token data --------------------------------------------------------------
/**
* This defines the different kinds of tokens or meaningful chunks of code that the parser knows how to consume.
* These let us distinguish, for example, between a string "foo" and a variable named "foo".
*
* HACK: A typical tokenizer would actually have unique token types for each keyword (print, goto, etc.) so that the
* parser doesn't have to look at the names, but Jasic is a little more crude.
*/
private enum TokenType {
WORD, NUMBER, STRING, LABEL, LINE, EQUALS, OPERATOR, LEFT_PAREN, RIGHT_PAREN, EOF
}
/**
* This is a single meaningful chunk of code. It is created by the tokenizer and consumed by the parser.
*/
private static class Token {
public Token(String text, TokenType type) {
this.text = text;
this.type = type;
}
public final String text;
public final TokenType type;
}
/**
* This defines the different states the tokenizer can be in while it's scanning through the source code. Tokenizers
* are state machines, which means the only data they need to store is where they are in the source code and this
* one "state" or mode value.
*
* One of the main differences between tokenizing and parsing is this regularity. Because the tokenizer stores only
* this one state value, it can't handle nesting (which would require also storing a number to identify how deeply
* nested you are). The parser is able to handle that.
*/
private enum TokenizeState {
DEFAULT, WORD, NUMBER, STRING, COMMENT
}
// Parsing -----------------------------------------------------------------
/**
* This defines the Jasic parser. The parser takes in a sequence of tokens and generates an abstract syntax tree.
* This is the nested data structure that represents the series of statements, and the expressions (which can nest
* arbitrarily deeply) that they evaluate. In technical terms, what we have is a recursive descent parser, the
* simplest kind to hand-write.
*
* As a side-effect, this phase also stores off the line numbers for each label in the program. It's a bit gross,
* but it works.
*/
private class Parser {
public Parser(List tokens) {
this.tokens = tokens;
position = 0;
}
/**
* The top-level function to start parsing. This will keep consuming tokens and routing to the other parse
* functions for the different grammar syntax until we run out of code to parse.
*
* @param labels
* A map of label names to statement indexes. The parser will fill this in as it scans the code.
* @return The list of parsed statements.
*/
public List parse(Map labels) {
List statements = new ArrayList<>();
while (true) {
// Ignore empty lines.
while (match(TokenType.LINE))
;
if (match(TokenType.LABEL)) {
// Mark the index of the statement after the label.
labels.put(last(1).text, statements.size());
} else if (match(TokenType.WORD, TokenType.EQUALS)) {
String name = last(2).text;
Expression value = expression();
statements.add(new AssignStatement(name, value));
} else if (match("print")) {
statements.add(new PrintStatement(expression()));
} else if (match("input")) {
statements.add(new InputStatement(consume(TokenType.WORD).text));
} else if (match("goto")) {
statements.add(new GotoStatement(consume(TokenType.WORD).text));
} else if (match("if")) {
Expression condition = expression();
consume("then");
String label = consume(TokenType.WORD).text;
statements.add(new IfThenStatement(condition, label));
} else
break; // Unexpected token (likely EOF), so end.
}
return statements;
}
// The following functions each represent one grammatical part of the
// language. If this parsed English, these functions would be named like
// noun() and verb().
/**
* Parses a single expression. Recursive descent parsers start with the lowest-precedent term and moves towards
* higher precedence. For Jasic, binary operators (+, -, etc.) are the lowest.
*
* @return The parsed expression.
*/
private Expression expression() {
return operator();
}
/**
* Parses a series of binary operator expressions into a single expression. In Jasic, all operators have the
* same predecence and associate left-to-right. That means it will interpret: 1 + 2 * 3 - 4 / 5 like: ((((1 + 2)
* * 3) - 4) / 5)
*
* It works by building the expression tree one at a time. So, given this code: 1 + 2 * 3, this will:
*
* 1. Parse (1) as an atomic expression. 2. See the (+) and start a new operator expression. 3. Parse (2) as an
* atomic expression. 4. Build a (1 + 2) expression and replace (1) with it. 5. See the (*) and start a new
* operator expression. 6. Parse (3) as an atomic expression. 7. Build a ((1 + 2) * 3) expression and replace (1
* + 2) with it. 8. Return the last expression built.
*
* @return The parsed expression.
*/
private Expression operator() {
Expression expression = atomic();
// Keep building operator expressions as long as we have operators.
while (match(TokenType.OPERATOR) || match(TokenType.EQUALS)) {
char operator = last(1).text.charAt(0);
Expression right = atomic();
expression = new OperatorExpression(expression, operator, right);
}
return expression;
}
/**
* Parses an "atomic" expression. This is the highest level of precedence and contains single literal tokens
* like 123 and "foo", as well as parenthesized expressions.
*
* @return The parsed expression.
*/
private Expression atomic() {
if (match(TokenType.WORD)) {
// A word is a reference to a variable.
return new VariableExpression(last(1).text);
} else if (match(TokenType.NUMBER)) {
return new NumberValue(Double.parseDouble(last(1).text));
} else if (match(TokenType.STRING)) {
return new StringValue(last(1).text);
} else if (match(TokenType.LEFT_PAREN)) {
// The contents of a parenthesized expression can be any
// expression. This lets us "restart" the precedence cascade
// so that you can have a lower precedence expression inside
// the parentheses.
Expression expression = expression();
consume(TokenType.RIGHT_PAREN);
return expression;
}
throw new Error("Couldn't parse :(");
}
// The following functions are the core low-level operations that the
// grammar parser is built in terms of. They match and consume tokens in
// the token stream.
/**
* Consumes the next two tokens if they are the given type (in order). Consumes no tokens if either check fais.
*
* @param type1
* Expected type of the next token.
* @param type2
* Expected type of the subsequent token.
* @return True if tokens were consumed.
*/
private boolean match(TokenType type1, TokenType type2) {
if (get(0).type != type1)
return false;
if (get(1).type != type2)
return false;
position += 2;
return true;
}
/**
* Consumes the next token if it's the given type.
*
* @param type
* Expected type of the next token.
* @return True if the token was consumed.
*/
private boolean match(TokenType type) {
if (get(0).type != type)
return false;
position++;
return true;
}
/**
* Consumes the next token if it's a word token with the given name.
*
* @param name
* Expected name of the next word token.
* @return True if the token was consumed.
*/
private boolean match(String name) {
if (get(0).type != TokenType.WORD)
return false;
if (!get(0).text.equals(name))
return false;
position++;
return true;
}
/**
* Consumes the next token if it's the given type. If not, throws an exception. This is for cases where the
* parser demands a token of a certain type in a certain position, for example a matching ) after an opening (.
*
* @param type
* Expected type of the next token.
* @return The consumed token.
*/
private Token consume(TokenType type) {
if (get(0).type != type)
throw new Error("Expected " + type + ".");
return tokens.get(position++);
}
/**
* Consumes the next token if it's a word with the given name. If not, throws an exception.
*
* @param name
* Expected name of the next word token.
* @return The consumed token.
*/
private Token consume(String name) {
if (!match(name))
throw new Error("Expected " + name + ".");
return last(1);
}
/**
* Gets a previously consumed token, indexing backwards. last(1) will be the token just consumed, last(2) the
* one before that, etc.
*
* @param offset
* How far back in the token stream to look.
* @return The consumed token.
*/
private Token last(int offset) {
return tokens.get(position - offset);
}
/**
* Gets an unconsumed token, indexing forward. get(0) will be the next token to be consumed, get(1) the one
* after that, etc.
*
* @param offset
* How far forward in the token stream to look.
* @return The yet-to-be-consumed token.
*/
private Token get(int offset) {
if (position + offset >= tokens.size()) {
return new Token("", TokenType.EOF);
}
return tokens.get(position + offset);
}
private final List tokens;
private int position;
}
// Abstract syntax tree (AST) ----------------------------------------------
// These classes define the syntax tree data structures. This is how code is
// represented internally in a way that's easy for the interpreter to
// understand.
//
// HACK: Unlike most real compilers or interpreters, the logic to execute
// the code is baked directly into these classes. Typically, it would be
// separated out so that the AST us just a static data structure.
/**
* Base interface for a Jasic statement. The different supported statement types like "print" and "goto" implement
* this.
*/
public interface Statement {
/**
* Statements implement this to actually perform whatever behavior the statement causes. "print" statements will
* display text here, "goto" statements will change the current statement, etc.
*/
void execute();
}
/**
* Base interface for an expression. An expression is like a statement except that it also returns a value when
* executed. Expressions do not appear at the top level in Jasic programs, but are used in many statements. For
* example, the value printed by a "print" statement is an expression. Unlike statements, expressions can nest.
*/
public interface Expression {
/**
* Expression classes implement this to evaluate the expression and return the value.
*
* @return The value of the calculated expression.
*/
Value evaluate();
}
/**
* A "print" statement evaluates an expression, converts the result to a string, and displays it to the user.
*/
public class PrintStatement implements Statement {
public PrintStatement(Expression expression) {
this.expression = expression;
}
public void execute() {
System.out.println(expression.evaluate().toString());
}
private final Expression expression;
}
/**
* An "input" statement reads input from the user and stores it in a variable.
*/
public class InputStatement implements Statement {
public InputStatement(String name) {
this.name = name;
}
public void execute() {
try {
String input = lineIn.readLine();
// Store it as a number if possible, otherwise use a string.
try {
double value = Double.parseDouble(input);
variables.put(name, new NumberValue(value));
} catch (NumberFormatException e) {
variables.put(name, new StringValue(input));
}
} catch (IOException e1) {
// HACK: Just ignore the problem.
}
}
private final String name;
}
/**
* An assignment statement evaluates an expression and stores the result in a variable.
*/
public class AssignStatement implements Statement {
public AssignStatement(String name, Expression value) {
this.name = name;
this.value = value;
}
public void execute() {
variables.put(name, value.evaluate());
}
private final String name;
private final Expression value;
}
/**
* A "goto" statement jumps execution to another place in the program.
*/
public class GotoStatement implements Statement {
public GotoStatement(String label) {
this.label = label;
}
public void execute() {
if (labels.containsKey(label)) {
currentStatement = labels.get(label).intValue();
}
}
private final String label;
}
/**
* An if then statement jumps execution to another place in the program, but only if an expression evaluates to
* something other than 0.
*/
public class IfThenStatement implements Statement {
public IfThenStatement(Expression condition, String label) {
this.condition = condition;
this.label = label;
}
public void execute() {
if (labels.containsKey(label)) {
double value = condition.evaluate().toNumber();
if (value != 0) {
currentStatement = labels.get(label).intValue();
}
}
}
private final Expression condition;
private final String label;
}
/**
* A variable expression evaluates to the current value stored in that variable.
*/
public class VariableExpression implements Expression {
public VariableExpression(String name) {
this.name = name;
}
public Value evaluate() {
if (variables.containsKey(name)) {
return variables.get(name);
}
return new NumberValue(0);
}
private final String name;
}
/**
* An operator expression evaluates two expressions and then performs some arithmetic operation on the results.
*/
public class OperatorExpression implements Expression {
public OperatorExpression(Expression left, char operator, Expression right) {
this.left = left;
this.operator = operator;
this.right = right;
}
public Value evaluate() {
Value leftVal = left.evaluate();
Value rightVal = right.evaluate();
switch (operator) {
case '=':
// Coerce to the left argument's type, then compare.
if (leftVal instanceof NumberValue) {
return new NumberValue((leftVal.toNumber() == rightVal.toNumber()) ? 1 : 0);
} else {
return new NumberValue(leftVal.toString().equals(rightVal.toString()) ? 1 : 0);
}
case '+':
// Addition if the left argument is a number, otherwise do
// string concatenation.
if (leftVal instanceof NumberValue) {
return new NumberValue(leftVal.toNumber() + rightVal.toNumber());
} else {
return new StringValue(leftVal.toString() + rightVal.toString());
}
case '-':
return new NumberValue(leftVal.toNumber() - rightVal.toNumber());
case '*':
return new NumberValue(leftVal.toNumber() * rightVal.toNumber());
case '/':
return new NumberValue(leftVal.toNumber() / rightVal.toNumber());
case '<':
// Coerce to the left argument's type, then compare.
if (leftVal instanceof NumberValue) {
return new NumberValue((leftVal.toNumber() < rightVal.toNumber()) ? 1 : 0);
} else {
return new NumberValue((leftVal.toString().compareTo(rightVal.toString()) < 0) ? 1 : 0);
}
case '>':
// Coerce to the left argument's type, then compare.
if (leftVal instanceof NumberValue) {
return new NumberValue((leftVal.toNumber() > rightVal.toNumber()) ? 1 : 0);
} else {
return new NumberValue((leftVal.toString().compareTo(rightVal.toString()) > 0) ? 1 : 0);
}
}
throw new Error("Unknown operator.");
}
private final Expression left;
private final char operator;
private final Expression right;
}
// Value types -------------------------------------------------------------
/**
* This is the base interface for a value. Values are the data that the interpreter processes. They are what gets
* stored in variables, printed, and operated on.
*
* There is an implementation of this interface for each of the different primitive types (really just double and
* string) that Jasic supports. Wrapping them in a single Value interface lets Jasic be dynamically-typed and
* convert between different representations as needed.
*
* Note that Value extends Expression. This is a bit of a hack, but it lets us use values (which are typically only
* ever seen by the interpreter and not the parser) as both runtime values, and as object representing literals in
* code.
*/
public interface Value extends Expression {
/**
* Value types override this to convert themselves to a string representation.
*/
String toString();
/**
* Value types override this to convert themselves to a numeric representation.
*/
double toNumber();
}
/**
* A numeric value. Jasic uses doubles internally for all numbers.
*/
public class NumberValue implements Value {
public NumberValue(double value) {
this.value = value;
}
@Override
public String toString() {
return Double.toString(value);
}
public double toNumber() {
return value;
}
public Value evaluate() {
return this;
}
private final double value;
}
/**
* A string value.
*/
public class StringValue implements Value {
public StringValue(String value) {
this.value = value;
}
@Override
public String toString() {
return value;
}
public double toNumber() {
return Double.parseDouble(value);
}
public Value evaluate() {
return this;
}
private final String value;
}
// Interpreter -------------------------------------------------------------
/**
* Constructs a new Jasic instance. The instance stores the global state of the interpreter such as the values of
* all of the variables and the current statement.
*/
public Jasic() {
variables = new HashMap();
labels = new HashMap();
InputStreamReader converter = new InputStreamReader(System.in);
lineIn = new BufferedReader(converter);
}
/**
* This is where the magic happens. This runs the code through the parsing pipeline to generate the AST. Then it
* executes each statement. It keeps track of the current line in a member variable that the statement objects have
* access to. This lets "goto" and "if then" do flow control by simply setting the index of the current statement.
*
* In an interpreter that didn't mix the interpretation logic in with the AST node classes, this would be doing a
* lot more work.
*
* @param source
* A string containing the source code of a .jas script to interpret.
*/
public void interpret(String source) {
// Tokenize.
List tokens = tokenize(source);
// Parse.
Parser parser = new Parser(tokens);
List statements = parser.parse(labels);
// Interpret until we're done.
currentStatement = 0;
while (currentStatement < statements.size()) {
int thisStatement = currentStatement;
currentStatement++;
statements.get(thisStatement).execute();
}
}
private final Map variables;
private final Map labels;
private final BufferedReader lineIn;
private int currentStatement;
}
