com.google.javascript.jscomp.MaybeReachingVariableUse Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of closure-compiler-linter Show documentation
Show all versions of closure-compiler-linter Show documentation
Closure Compiler is a JavaScript optimizing compiler. It parses your
JavaScript, analyzes it, removes dead code and rewrites and minimizes
what's left. It also checks syntax, variable references, and types, and
warns about common JavaScript pitfalls. It is used in many of Google's
JavaScript apps, including Gmail, Google Web Search, Google Maps, and
Google Docs.
This binary checks for style issues such as incorrect or missing JSDoc
usage, and missing goog.require() statements. It does not do more advanced
checks such as typechecking.
/*
* Copyright 2009 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;
import static com.google.common.base.Preconditions.checkNotNull;
import static com.google.common.base.Preconditions.checkState;
import com.google.common.collect.HashMultimap;
import com.google.common.collect.Multimap;
import com.google.common.collect.MultimapBuilder;
import com.google.javascript.jscomp.ControlFlowGraph.Branch;
import com.google.javascript.jscomp.graph.DiGraph.DiGraphEdge;
import com.google.javascript.jscomp.graph.GraphNode;
import com.google.javascript.jscomp.graph.LatticeElement;
import com.google.javascript.rhino.Node;
import java.util.ArrayList;
import java.util.Collection;
import java.util.HashMap;
import java.util.HashSet;
import java.util.List;
import java.util.Map;
import java.util.Set;
/**
* Computes "may be" reaching use for all definitions of each variables.
*
* A use of {@code A} in {@code alert(A)} is a "may be" reaching use of
* the definition of {@code A} at {@code A = foo()} if at least one path from
* the use node reaches that definition and it is the last definition before
* the use on that path.
*/
class MaybeReachingVariableUse extends
DataFlowAnalysis {
// The scope of the function that we are analyzing.
private final Set escaped;
private final Map allVarsInFn;
private final List orderedVars;
MaybeReachingVariableUse(
ControlFlowGraph cfg,
Scope jsScope,
AbstractCompiler compiler,
SyntacticScopeCreator scopeCreator) {
super(cfg, new ReachingUsesJoinOp());
this.escaped = new HashSet<>();
this.allVarsInFn = new HashMap<>();
this.orderedVars = new ArrayList<>();
// TODO(user): Maybe compute it somewhere else and re-use the escape
// local set here.
computeEscaped(jsScope.getParent(), escaped, compiler, scopeCreator);
NodeUtil.getAllVarsDeclaredInFunction(
allVarsInFn, orderedVars, compiler, scopeCreator, jsScope.getParent());
}
/**
* May use definition lattice representation. It captures a product
* lattice for each local (non-escaped) variable. The sub-lattice is
* a n + 2 power set element lattice with all the Nodes in the program,
* TOP and BOTTOM. This is better explained with an example:
*
* Consider: A sub-lattice element representing the variable A represented
* by { N_4, N_5} where N_x is a Node in the program. This implies at
* that particular point in the program the content of A is "upward exposed"
* at point N_4 and N_5.
*
* Example:
*
* A = 1;
* ...
* N_3:
* N_4: print(A);
* N_5: y = A;
* N_6: A = 1;
* N_7: print(A);
*
* At N_3, reads of A in {N_4, N_5} are said to be upward exposed.
*/
static final class ReachingUses implements LatticeElement {
final Multimap mayUseMap;
public ReachingUses() {
mayUseMap = HashMultimap.create();
}
/**
* Copy constructor.
*
* @param other The constructed object is a replicated copy of this element.
*/
public ReachingUses(ReachingUses other) {
mayUseMap = MultimapBuilder.hashKeys().hashSetValues().build(other.mayUseMap);
}
@Override
public boolean equals(Object other) {
return (other instanceof ReachingUses)
&& ((ReachingUses) other).mayUseMap.equals(this.mayUseMap);
}
@Override
public int hashCode() {
return mayUseMap.hashCode();
}
}
/**
* The join is a simple union because of the "may be" nature of the analysis.
*
* Consider: A = 1; if (x) { A = 2 }; alert(A);
*
* The read of A "may be" exposed to A = 1 in the beginning.
*/
private static class ReachingUsesJoinOp implements JoinOp {
@Override
public ReachingUses apply(List from) {
ReachingUses result = new ReachingUses();
for (ReachingUses uses : from) {
result.mayUseMap.putAll(uses.mayUseMap);
}
return result;
}
}
@Override
boolean isForward() {
return false;
}
@Override
ReachingUses createEntryLattice() {
return new ReachingUses();
}
@Override
ReachingUses createInitialEstimateLattice() {
return new ReachingUses();
}
@Override
ReachingUses flowThrough(Node n, ReachingUses input) {
ReachingUses output = new ReachingUses(input);
// If there's an ON_EX edge, this cfgNode may or may not get executed.
// We can express this concisely by just pretending this happens in
// a conditional.
boolean conditional = hasExceptionHandler(n);
computeMayUse(n, n, output, conditional);
return output;
}
private boolean hasExceptionHandler(Node cfgNode) {
List> branchEdges = getCfg().getOutEdges(cfgNode);
for (DiGraphEdge edge : branchEdges) {
if (edge.getValue() == Branch.ON_EX) {
return true;
}
}
return false;
}
/**
* @param conditional Whether {@code n} is only conditionally evaluated given that {@code cfgNode}
* is evaluated. Do not remove conditionally redefined variables from the reaching uses set.
*/
private void computeMayUse(Node n, Node cfgNode, ReachingUses output, boolean conditional) {
switch (n.getToken()) {
case BLOCK:
case ROOT:
case FUNCTION:
return;
case NAME:
if (NodeUtil.isLhsByDestructuring(n)) {
if (!conditional) {
removeFromUseIfLocal(n.getString(), output);
}
} else {
addToUseIfLocal(n.getString(), cfgNode, output);
}
return;
case WHILE:
case DO:
case IF:
case FOR:
computeMayUse(NodeUtil.getConditionExpression(n), cfgNode, output, conditional);
return;
case FOR_IN:
case FOR_OF:
case FOR_AWAIT_OF:
// for(x in y) {...}
Node lhs = n.getFirstChild();
Node rhs = lhs.getNext();
if (NodeUtil.isNameDeclaration(lhs)) {
lhs = lhs.getLastChild(); // for(var x in y) {...}
if (lhs.isDestructuringLhs()) {
lhs = lhs.getFirstChild(); // for (let [x] of obj) {...}
}
}
if (lhs.isName() && !conditional) {
removeFromUseIfLocal(lhs.getString(), output);
} else if (lhs.isDestructuringPattern()) {
computeMayUse(lhs, cfgNode, output, true);
}
computeMayUse(rhs, cfgNode, output, conditional);
return;
case AND:
case OR:
computeMayUse(n.getLastChild(), cfgNode, output, true);
computeMayUse(n.getFirstChild(), cfgNode, output, conditional);
return;
case HOOK:
computeMayUse(n.getLastChild(), cfgNode, output, true);
computeMayUse(n.getSecondChild(), cfgNode, output, true);
computeMayUse(n.getFirstChild(), cfgNode, output, conditional);
return;
case VAR:
case LET:
case CONST:
Node varName = n.getFirstChild();
checkState(n.hasChildren(), "AST should be normalized", n);
if (varName.isDestructuringLhs()) {
// Note: since destructuring is evaluated in reverse AST order, we traverse the first
// child before the second in order to do our backwards data flow analysis.
computeMayUse(varName.getFirstChild(), cfgNode, output, conditional);
computeMayUse(varName.getSecondChild(), cfgNode, output, conditional);
} else if (varName.hasChildren()) {
computeMayUse(varName.getFirstChild(), cfgNode, output, conditional);
if (!conditional) {
removeFromUseIfLocal(varName.getString(), output);
}
} // else var name declaration with no initial value
return;
case DEFAULT_VALUE:
if (n.getFirstChild().isDestructuringPattern()) {
computeMayUse(n.getFirstChild(), cfgNode, output, conditional);
computeMayUse(n.getSecondChild(), cfgNode, output, true);
} else if (n.getFirstChild().isName()) {
// assigning to the name occurs after evaluating the default value
if (!conditional) {
removeFromUseIfLocal(n.getFirstChild().getString(), output);
}
computeMayUse(n.getSecondChild(), cfgNode, output, true);
} else {
computeMayUse(n.getSecondChild(), cfgNode, output, true);
computeMayUse(n.getFirstChild(), cfgNode, output, conditional);
}
break;
default:
if (NodeUtil.isAssignmentOp(n) && n.getFirstChild().isName()) {
Node name = n.getFirstChild();
if (!conditional) {
removeFromUseIfLocal(name.getString(), output);
}
// In case of a += "Hello". There is a read of a.
if (!n.isAssign()) {
addToUseIfLocal(name.getString(), cfgNode, output);
}
computeMayUse(name.getNext(), cfgNode, output, conditional);
} else if (n.isAssign() && n.getFirstChild().isDestructuringPattern()) {
// Note: the rhs of destructuring is evaluated before the lhs
computeMayUse(n.getFirstChild(), cfgNode, output, conditional);
computeMayUse(n.getSecondChild(), cfgNode, output, conditional);
} else {
/*
* We want to traverse in reverse order because we want the LAST
* definition in the sub-tree.
*/
for (Node c = n.getLastChild(); c != null; c = c.getPrevious()) {
computeMayUse(c, cfgNode, output, conditional);
}
}
}
}
/**
* Sets the variable for the given name to the node value in the upward
* exposed lattice. Do nothing if the variable name is one of the escaped
* variable.
*/
private void addToUseIfLocal(String name, Node node, ReachingUses use) {
Var var = allVarsInFn.get(name);
if (var == null) {
return;
}
if (!escaped.contains(var)) {
use.mayUseMap.put(var, node);
}
}
/**
* Removes the variable for the given name from the node value in the upward
* exposed lattice. Do nothing if the variable name is one of the escaped
* variable.
*/
private void removeFromUseIfLocal(String name, ReachingUses use) {
Var var = allVarsInFn.get(name);
if (var == null) {
return;
}
if (!escaped.contains(var)) {
use.mayUseMap.removeAll(var);
}
}
/**
* Gets a list of nodes that may be using the value assigned to {@code name}
* in {@code defNode}. {@code defNode} must be one of the control flow graph
* nodes.
*
* @param name name of the variable. It can only be names of local variable
* that are not function parameters, escaped variables or variables
* declared in catch.
* @param defNode The list of upward exposed use for the variable.
*/
Collection getUses(String name, Node defNode) {
GraphNode n = getCfg().getNode(defNode);
checkNotNull(n);
FlowState state = n.getAnnotation();
return state.getOut().mayUseMap.get(allVarsInFn.get(name));
}
}