Package org.python.pydev.parser.fastparser

Source Code of org.python.pydev.parser.fastparser.FastDefinitionsParser

/**
* Copyright (c) 2005-2011 by Appcelerator, Inc. All Rights Reserved.
* Licensed under the terms of the Eclipse Public License (EPL).
* Please see the license.txt included with this distribution for details.
* Any modifications to this file must keep this entire header intact.
*/
package org.python.pydev.parser.fastparser;

import java.util.ArrayList;
import java.util.List;

import org.python.pydev.core.ObjectsPool;
import org.python.pydev.core.ObjectsPool.ObjectsPoolMap;
import org.python.pydev.core.docutils.ParsingUtils;
import org.python.pydev.core.docutils.StringUtils;
import org.python.pydev.core.docutils.SyntaxErrorException;
import org.python.pydev.core.log.Log;
import org.python.pydev.core.structure.FastStack;
import org.python.pydev.parser.jython.SimpleNode;
import org.python.pydev.parser.jython.ast.Assign;
import org.python.pydev.parser.jython.ast.Attribute;
import org.python.pydev.parser.jython.ast.ClassDef;
import org.python.pydev.parser.jython.ast.FunctionDef;
import org.python.pydev.parser.jython.ast.Module;
import org.python.pydev.parser.jython.ast.Name;
import org.python.pydev.parser.jython.ast.NameTok;
import org.python.pydev.parser.jython.ast.exprType;
import org.python.pydev.parser.jython.ast.stmtType;

import com.aptana.shared_core.callbacks.ICallback;
import com.aptana.shared_core.string.FastStringBuffer;
import com.aptana.shared_core.structure.Tuple;

/**
* @note: Unfinished
*
* This class should be able to gather the definitions found in a module in a very fast way.
*
* The target is having a performance around 5x faster than doing a regular parse, focusing on getting
* the name tokens for:
*
* classes, functions, class attributes, instance attributes -- basically the tokens that provide a
* definition that can be 'globally' accessed.
*
* @author Fabio
*/
public final class FastDefinitionsParser {

    /**
     * Set and kept in the constructor
     */

    /**
     * The chars we should iterate through.
     */
    final private char[] cs;

    /**
     * The length of the buffer we're iterating.
     */
    final private int length;

    /**
     * Current iteration index
     */
    private int currIndex = 0;

    /**
     * The current column
     */
    private int col;

    /**
     * The current row
     */
    private int row = 0;

    /**
     * The column where the 1st char was found
     */
    private int firstCharCol = 1;

    /**
     * Holds things added to the 'global' module
     */
    private final ArrayList<stmtType> body = new ArrayList<stmtType>(16);

    /**
     * Holds a stack of classes so that we create a new one in each new scope to be filled and when the scope is ended,
     * it should have its body filled with the stackBody contents related to each
     */
    private final FastStack<SimpleNode> stack = new FastStack<SimpleNode>(20);

    /**
     * For each item in the stack, there's a stackBody that has the contents to be added later to that class.
     */
    private final FastStack<List<stmtType>> stackBody = new FastStack<List<stmtType>>(20);

    /**
     * Buffer with the contents of a line.
     */
    private final FastStringBuffer lineBuffer = new FastStringBuffer();

    /**
     * Should we debug?
     */
    private final static boolean DEBUG = false;

    private FastDefinitionsParser(char[] cs) {
        this(cs, cs.length);
    }

    /**
     * Constructor
     *
     * @param cs array of chars that should be considered.
     * @param len the number of chars to be used (usually cs.length).
     */
    private FastDefinitionsParser(char[] cs, int len) {
        this.cs = cs;
        this.length = len;
    }

    /**
     * This is the method that actually extracts things from the passed buffer.
     * @throws SyntaxErrorException
     */
    private void extractBody() throws SyntaxErrorException {
        ParsingUtils parsingUtils = ParsingUtils.create(cs, false, length);

        if (currIndex < length) {
            handleNewLine(parsingUtils);
        }
        //in the 1st attempt to handle the 1st line, if it had nothing we could actually go backward 1 char
        if (currIndex < 0) {
            currIndex = 0;
        }

        for (; currIndex < length; currIndex++, col++) {
            char c = cs[currIndex];

            switch (c) {

                case '\'':
                case '"':
                    if (DEBUG) {
                        System.out.println("literal");
                    }
                    //go to the end of the literal
                    int initialIndex = currIndex;
                    currIndex = parsingUtils.getLiteralEnd(currIndex, c);

                    //keep the row count correct
                    updateCountRow(initialIndex, currIndex);
                    break;

                case '#':
                    if (DEBUG) {
                        System.out.println("comment");
                    }
                    //go to the end of the comment
                    while (currIndex < length) {
                        c = cs[currIndex];
                        if (c == '\r' || c == '\n') {
                            currIndex--;
                            break;
                        }
                        currIndex++;
                    }

                    break;

                case '{':
                case '[':
                case '(':
                    //starting some call, dict, list, tuple... those don't count on getting some actual definition
                    initialIndex = currIndex;
                    currIndex = parsingUtils.eatPar(currIndex, null, c);

                    //keep the row count correct
                    updateCountRow(initialIndex, currIndex);
                    break;

                case '\r':
                    if (currIndex < length - 1 && cs[currIndex + 1] == '\n') {
                        currIndex++;
                    }
                    /*FALLTHROUGH**/
                case '\n':
                    currIndex++;
                    handleNewLine(parsingUtils);
                    if (currIndex < length) {
                        c = cs[currIndex];
                    }

                    break;

                case '=':
                    if (currIndex < length - 1 && cs[currIndex + 1] != '=') {
                        //should not be ==
                        //other cases such as !=, +=, -= are already treated because they don't constitute valid
                        //chars for an identifier.

                        if (DEBUG) {
                            System.out.println("Found possible attribute:" + lineBuffer + " col:" + firstCharCol);
                        }

                        //if we've an '=', let's get the whole line contents to analyze...
                        //Note: should have stopped just before the new line (so, as we'll do currIndex++ in the
                        //next loop, that's ok).
                        initialIndex = currIndex;
                        currIndex = parsingUtils.getFullFlattenedLine(currIndex, lineBuffer);

                        //keep the row count correct
                        updateCountRow(initialIndex, currIndex);

                        String equalsLine = lineBuffer.toString().trim();
                        lineBuffer.clear();

                        final List<String> splitted = StringUtils.split(equalsLine, '=');
                        final int splittedLen = splitted.size();
                        ArrayList<exprType> targets = new ArrayList<exprType>(2);

                        for (int j = 0; j < splittedLen - 1 || (splittedLen == 1 && j == 0); j++) { //we don't want to get the last one.
                            String lineContents = splitted.get(j).trim();
                            if (lineContents.length() == 0) {
                                continue;
                            }
                            boolean add = true;
                            for (int i = 0; i < lineContents.length(); i++) {
                                char lineC = lineContents.charAt(i);
                                //can only be made of valid java chars (no spaces or similar things)
                                if (lineC != '.' && !Character.isJavaIdentifierPart(lineC)) {
                                    add = false;
                                    break;
                                }
                            }
                            if (add) {
                                //only add if it was something valid
                                if (lineContents.indexOf('.') != -1) {
                                    List<String> dotSplit = StringUtils.dotSplit(lineContents);
                                    if (dotSplit.size() == 2 && dotSplit.get(0).equals("self")) {
                                        Attribute attribute = new Attribute(new Name("self", Name.Load, false),
                                                new NameTok(dotSplit.get(1), NameTok.Attrib), Attribute.Load);
                                        targets.add(attribute);
                                    }

                                } else {
                                    Name name = new Name(lineContents, Name.Store, false);
                                    targets.add(name);
                                }
                            }
                        }

                        if (targets.size() > 0) {
                            Assign assign = new Assign(targets.toArray(new exprType[targets.size()]), null);
                            assign.beginColumn = this.firstCharCol;
                            assign.beginLine = this.row;
                            addToPertinentScope(assign);
                        }
                    }
                    //No default
            }
            lineBuffer.append(c);
        }

        endScopesInStack();
    }

    public void updateCountRow(int initialIndex, int currIndex) {
        char c;
        int len = length;
        for (int k = initialIndex; k < len && k <= currIndex; k++) {
            c = cs[k];
            switch (c) {
                case '\n':
                    row += 1;
                    break;

                case '\r':
                    row += 1;
                    if (k < len - 1 && k <= currIndex - 1) {
                        if (cs[k + 1] == '\n') {
                            k++; //skip the \n after the \r
                        }
                    }
                    break;
            }
        }
    }

    /**
     * Called when a new line is found. Tries to make the match of function and class definitions.
     * @throws SyntaxErrorException
     */
    private void handleNewLine(ParsingUtils parsingUtils) throws SyntaxErrorException {
        if (currIndex >= length - 1) {
            return;
        }

        col = 1;
        row++;
        if (DEBUG) {
            System.out.println("Handling new line:" + row);
        }

        lineBuffer.clear();
        char c = cs[currIndex];

        while (currIndex < length - 1 && Character.isWhitespace(c) && c != '\r' && c != '\n') {
            currIndex++;
            col++;
            c = cs[currIndex];
        }

        if (c == 'c' && matchClass()) {
            int startClassCol = col;
            currIndex += 6;
            col += 6;

            startClass(getNextIdentifier(c), row, startClassCol);

        } else if (c == 'd' && matchFunction()) {
            int startMethodCol = col;
            currIndex += 4;
            col += 4;

            startMethod(getNextIdentifier(c), row, startMethodCol);
        }
        firstCharCol = col;
        if (currIndex < length) {

            //starting some call, dict, list, tuple... those don't count on getting some actual definition
            int initialIndex = currIndex;

            int tempIndex = skipWhitespaces(currIndex);

            if (tempIndex >= length) {
                return;
            }
            c = cs[tempIndex];

            boolean updateIndex = false;
            switch (c) {
                case '(':
                    tempIndex = parsingUtils.eatPar(tempIndex, null, c);

                    if (tempIndex < length) {
                        tempIndex = skipWhitespaces(tempIndex);

                        c = cs[tempIndex];
                        if (c == ')') {
                            tempIndex++;
                        }
                    }

                    if (tempIndex < length) {
                        tempIndex = skipWhitespaces(tempIndex);

                        c = cs[tempIndex];
                        if (c == ':') {
                            tempIndex++;

                            if (tempIndex < length) {
                                c = cs[tempIndex];
                                if (c != '\r' && c != '\n') {
                                    updateIndex = true;
                                }
                            }
                        }
                    }

                    if (updateIndex) {
                        tempIndex = skipWhitespaces(tempIndex);
                        currIndex = tempIndex;
                        //keep the row count correct
                        updateCountRow(initialIndex, currIndex);

                        //now, update the first char col to be the char after the ':' in "def m2(self):", in a line as
                        //def m2(self): self.a = 10 (all in a single line)
                        int i = tempIndex;
                        while (i > 0 && i < length) {
                            c = cs[i];
                            if (c == '\r' || c == '\n') {
                                break;
                            }
                            i--;
                        }
                        firstCharCol = tempIndex - i;
                    } else {
                        currIndex--;
                    }

                    break;

                default:
                    currIndex--;
                    break;

            }
        }
    }

    /**
     * Note that it'll only skip whitespaces (not newlines)
     */
    private int skipWhitespaces(int tempIndex) {
        char c;
        while (tempIndex < length) {
            c = cs[tempIndex];
            if (c == ' ' || c == '\t') {
                tempIndex++;
            } else {
                break;
            }
        }
        return tempIndex;
    }

    /**
     * Get the next identifier available.
     * @param c the current char
     * @return the identifier found
     */
    private String getNextIdentifier(char c) {
        c = this.cs[currIndex];

        while (currIndex < length && Character.isWhitespace(c)) {
            currIndex++;
            c = this.cs[currIndex];
        }

        int currClassNameCol = currIndex;
        while (Character.isJavaIdentifierPart(c)) {
            currIndex++;
            if (currIndex >= length) {
                break;
            }
            c = this.cs[currIndex];
        }
        return ObjectsPool.internLocal(interned, new String(this.cs, currClassNameCol, currIndex - currClassNameCol));
    }

    private final ObjectsPoolMap interned = new ObjectsPoolMap();

    /**
     * Start a new method scope with the given row and column.
     * @param startMethodRow the row where the scope should start
     * @param startMethodCol the column where the scope should start
     */
    private void startMethod(String name, int startMethodRow, int startMethodCol) {
        if (startMethodCol == 1) {
            endScopesInStack();
        }
        NameTok nameTok = new NameTok(name, NameTok.ClassName);
        FunctionDef functionDef = new FunctionDef(nameTok, null, null, null, null);
        functionDef.beginLine = startMethodRow;
        functionDef.beginColumn = startMethodCol;

        addToPertinentScope(functionDef);
        if (stack.size() == 0) {
            stack.push(functionDef);
        }
    }

    /**
     * Start a new class scope with the given row and column.
     * @param startClassRow the row where the scope should start
     * @param startClassCol the column where the scope should start
     */
    private void startClass(String name, int startClassRow, int startClassCol) {
        if (startClassCol == 1) {
            endScopesInStack();
        }
        NameTok nameTok = new NameTok(name, NameTok.ClassName);
        ClassDef classDef = new ClassDef(nameTok, null, null, null, null, null, null);

        classDef.beginLine = startClassRow;
        classDef.beginColumn = startClassCol;

        stack.push(classDef);
        stackBody.push(new ArrayList<stmtType>(10));
    }

    private void endScopesInStack() {
        while (stack.size() > 0) {
            endScope();
        }
    }

    /**
     * Finish the current scope in the stack.
     *
     * May close many scopes in a single call depending on where the class should be added to.
     */
    private void endScope() {
        SimpleNode pop = stack.pop();
        if (!(pop instanceof ClassDef)) {
            return;
        }
        ClassDef def = (ClassDef) pop;
        List<stmtType> body = stackBody.pop();
        def.body = body.toArray(new stmtType[body.size()]);
        addToPertinentScope(def);
    }

    /**
     * This is the definition to be added to a given scope.
     *
     * It'll find a correct scope based on the column it has to be added to.
     *
     * @param newStmt the definition to be added
     */
    private void addToPertinentScope(stmtType newStmt) {
        //see where it should be added (global or class scope)
        while (stack.size() > 0) {
            SimpleNode parent = stack.peek();
            if (parent.beginColumn < newStmt.beginColumn) {
                if (parent instanceof FunctionDef) {
                    return;
                }
                List<stmtType> peek = stackBody.peek();

                if (newStmt instanceof FunctionDef) {
                    int size = peek.size();
                    if (size > 0) {
                        stmtType existing = peek.get(size - 1);
                        if (existing.beginColumn < newStmt.beginColumn) {
                            //we don't want to add a method inside a method at this point.
                            //all the items added should have the same column.
                            return;
                        }
                    }
                } else if (newStmt instanceof Assign) {
                    Assign assign = (Assign) newStmt;
                    exprType target = assign.targets[0];

                    //an assign could be in a method or in a class depending on where we're right now...
                    int size = peek.size();
                    if (size > 0) {
                        stmtType existing = peek.get(size - 1);
                        if (existing.beginColumn < assign.beginColumn) {
                            //add the assign to the correct place
                            if (existing instanceof FunctionDef) {
                                FunctionDef functionDef = (FunctionDef) existing;

                                if (target instanceof Attribute) {
                                    addAssignToFunctionDef(assign, functionDef);
                                }
                                return;
                            }
                        }
                    }

                    //if it still hasn't returned and it's a name, add it to the global scope.
                    if (target instanceof Name) {

                    }
                }
                peek.add(newStmt);
                return;
            } else {
                endScope();
            }
        }
        //if it still hasn't returned, add it to the global
        this.body.add(newStmt);
    }

    /**
     * Adds an assign statement to the given function definition.
     *
     * @param assign the assign to be added
     * @param functionDef the function definition where it should be added
     */
    private void addAssignToFunctionDef(Assign assign, FunctionDef functionDef) {
        //if it's an attribute at this point, it'll always start with self!
        if (functionDef.body == null) {
            if (functionDef.specialsAfter == null) {
                functionDef.specialsAfter = new ArrayList<Object>(3);
            }
            functionDef.body = new stmtType[10];
            functionDef.body[0] = assign;
            functionDef.specialsAfter.add(1); //real len
        } else {
            //already exists... let's add it... as it's an array, we may have to reallocate it
            Integer currLen = (Integer) functionDef.specialsAfter.get(0);
            currLen += 1;
            functionDef.specialsAfter.set(0, currLen);
            if (functionDef.body.length < currLen) {
                stmtType[] newBody = new stmtType[functionDef.body.length * 2];
                System.arraycopy(functionDef.body, 0, newBody, 0, functionDef.body.length);
                functionDef.body = newBody;
            }
            functionDef.body[currLen - 1] = assign;
        }
    }

    /**
     * @return true if we have a match for 'class' in the current index (the 'c' must be already matched at this point)
     */
    private boolean matchClass() {
        if (currIndex + 5 > this.length) {
            return false;
        }
        return (this.cs[currIndex + 1] == 'l' && this.cs[currIndex + 2] == 'a' && this.cs[currIndex + 3] == 's'
                && this.cs[currIndex + 4] == 's' && Character.isWhitespace(this.cs[currIndex + 5]));
    }

    /**
     * @return true if we have a match for 'def' in the current index (the 'd' must be already matched at this point)
     */
    private boolean matchFunction() {
        if (currIndex + 3 > this.length) {
            return false;
        }
        return (this.cs[currIndex + 1] == 'e' && this.cs[currIndex + 2] == 'f' && Character
                .isWhitespace(this.cs[currIndex + 3]));
    }

    /**
     * Callbacks called just before returning a parsed object. Used for tests
     */
    public static List<ICallback<Object, Tuple<String, SimpleNode>>> parseCallbacks = new ArrayList<ICallback<Object, Tuple<String, SimpleNode>>>();

    /**
     * Convenience method for parse(s.toCharArray())
     * @param s the string to be parsed
     * @return a Module node with the structure found
     */
    public static SimpleNode parse(String s, String moduleName) {
        return parse(s.toCharArray(), moduleName);
    }

    /**
     * This method will parse the char array passed and will build a structure with the contents of the file.
     * @param cs the char array to be parsed
     * @return a Module node with the structure found
     */
    public static SimpleNode parse(char[] cs, String moduleName) {
        return parse(cs, moduleName, cs.length);
    }

    public static SimpleNode parse(char[] cs, String moduleName, int len) {
        FastDefinitionsParser parser = new FastDefinitionsParser(cs, len);
        try {
            parser.extractBody();
        } catch (SyntaxErrorException e) {
            throw new RuntimeException(e);
        } catch (StackOverflowError e) {
            RuntimeException runtimeException = new RuntimeException(e);
            Log.log("Error parsing: " + moduleName + "\nContents:\n" + new String(cs, 0, len > 1000 ? 1000 : len),
                    runtimeException); //report at most 1000 chars...
            throw runtimeException;
        }
        List<stmtType> body = parser.body;
        Module ret = new Module(body.toArray(new stmtType[body.size()]));
        if (parseCallbacks.size() > 0) {
            Tuple<String, SimpleNode> arg = new Tuple<String, SimpleNode>(moduleName, ret);
            for (ICallback<Object, Tuple<String, SimpleNode>> c : parseCallbacks) {
                c.call(arg);
            }
        }
        return ret;
    }

    public static SimpleNode parse(String s) {
        return parse(s.toCharArray(), null);
    }

}
TOP

Related Classes of org.python.pydev.parser.fastparser.FastDefinitionsParser

TOP
Copyright © 2018 www.massapi.com. All rights reserved.
All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.