com.google.javascript.jscomp.AbstractPeepholeOptimization Maven / Gradle / Ivy
Show all versions of closure-compiler Show documentation
/*
* Copyright 2010 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 com.google.javascript.jscomp.base.Tri;
import com.google.javascript.jscomp.parsing.parser.FeatureSet;
import com.google.javascript.rhino.Node;
import java.math.BigInteger;
/**
* An abstract class whose implementations run peephole optimizations:
* optimizations that look at a small section of code and either remove
* that code (if it is not needed) or replaces it with smaller code.
*/
abstract class AbstractPeepholeOptimization {
/** Intentionally not exposed to subclasses */
private AbstractCompiler compiler;
/** Intentionally not exposed to subclasses */
private AstAnalyzer astAnalyzer;
/**
* Given a node to optimize and a traversal, optimize the node. Subclasses
* should override to provide their own peephole optimization.
*
* @param subtree The subtree that will be optimized.
* @return The new version of the subtree (or null if the subtree or one of
* its parents was removed from the AST). If the subtree has not changed,
* this method must return {@code subtree}.
*/
abstract Node optimizeSubtree(Node subtree);
/**
* Helper method for reporting an error to the compiler when applying a
* peephole optimization.
*
* @param diagnostic The error type
* @param n The node for which the error should be reported
*/
protected void report(DiagnosticType diagnostic, Node n) {
JSError error = JSError.make(n, diagnostic, n.toString());
compiler.report(error);
}
/**
* Are the nodes equal for the purpose of inlining?
* If type aware optimizations are on, type equality is checked.
*/
protected boolean areNodesEqualForInlining(Node n1, Node n2) {
/* Our implementation delegates to the compiler. We provide this
* method because we don't want to expose Compiler to PeepholeOptimizations.
*/
checkNotNull(compiler);
return compiler.areNodesEqualForInlining(n1, n2);
}
/**
* Is the current AST normalized? (e.g. has the Normalize pass been run
* and has the Denormalize pass not yet been run?)
*/
protected boolean isASTNormalized() {
checkNotNull(compiler);
return compiler.getLifeCycleStage().isNormalized();
}
/** Informs the optimization that a traversal will begin. */
void beginTraversal(AbstractCompiler compiler) {
this.compiler = checkNotNull(compiler);
astAnalyzer = compiler.getAstAnalyzer();
}
/** Returns whether the node may create new mutable state, or change existing state. */
protected boolean mayEffectMutableState(Node n) {
return astAnalyzer.mayEffectMutableState(n);
}
/** Returns whether the node may have side effects when executed. */
protected boolean mayHaveSideEffects(Node n) {
return astAnalyzer.mayHaveSideEffects(n);
}
/**
* Returns the number value of the node if it has one and it cannot have side effects.
*
* Returns {@code null} otherwise.
*/
protected Double getSideEffectFreeNumberValue(Node n) {
Double value = NodeUtil.getNumberValue(n);
// Calculating the number value, if any, is likely to be faster than calculating side effects,
// and there are only a very few cases where we can compute a number value, but there could
// also be side effects. e.g. `void doSomething()` has value NaN, regardless of the behavior
// of `doSomething()`
if (value != null && astAnalyzer.mayHaveSideEffects(n)) {
value = null;
}
return value;
}
/**
* Returns the bigint value of the node if it has one and it cannot have side effects.
*
*
Returns {@code null} otherwise.
*/
protected BigInteger getSideEffectFreeBigIntValue(Node n) {
BigInteger value = NodeUtil.getBigIntValue(n);
// Calculating the bigint value, if any, is likely to be faster than calculating side effects,
// and there are only a very few cases where we can compute a bigint value, but there could
// also be side effects. e.g. `void doSomething()` has value NaN, regardless of the behavior
// of `doSomething()`
if (value != null && astAnalyzer.mayHaveSideEffects(n)) {
value = null;
}
return value;
}
/**
* Gets the value of a node as a String, or {@code null} if it cannot be converted.
*
*
This method effectively emulates the String() JavaScript cast function when
* possible and the node has no side effects. Otherwise, it returns {@code null}.
*/
protected String getSideEffectFreeStringValue(Node n) {
String value = NodeUtil.getStringValue(n);
// Calculating the string value, if any, is likely to be faster than calculating side effects,
// and there are only a very few cases where we can compute a string value, but there could
// also be side effects. e.g. `void doSomething()` has value 'undefined', regardless of the
// behavior of `doSomething()`
if (value != null && astAnalyzer.mayHaveSideEffects(n)) {
value = null;
}
return value;
}
/**
* Calculate the known boolean value for a node if possible and if it has no side effects.
*
*
Returns {@link Tri#UNKNOWN} if the node has side effects or its value cannot be statically
* determined.
*/
protected Tri getSideEffectFreeBooleanValue(Node n) {
Tri value = NodeUtil.getBooleanValue(n);
// Calculating the boolean value, if any, is likely to be faster than calculating side effects,
// and there are only a very few cases where we can compute a boolean value, but there could
// also be side effects. e.g. `void doSomething()` has value `false`, regardless of the
// behavior of `doSomething()`
if (value != Tri.UNKNOWN && astAnalyzer.mayHaveSideEffects(n)) {
value = Tri.UNKNOWN;
}
return value;
}
/**
* Returns true if the current node's type implies side effects.
*
*
This is a non-recursive version of the may have side effects check; used to check wherever
* the current node's type is one of the reason's why a subtree has side effects.
*/
protected boolean nodeTypeMayHaveSideEffects(Node n) {
return astAnalyzer.nodeTypeMayHaveSideEffects(n);
}
/**
* Returns whether the output language is ECMAScript 5 or later. Workarounds for quirks in
* browsers that do not support ES5 can be ignored when this is true.
*/
protected boolean isEcmaScript5OrGreater() {
return compiler != null
&& compiler.getOptions().getOutputFeatureSet().contains(FeatureSet.ES5);
}
/** Returns the current coding convention. */
protected CodingConvention getCodingConvention() {
// Note: this assumes a thread safe coding convention object.
return compiler.getCodingConvention();
}
protected final void reportChangeToEnclosingScope(Node n) {
compiler.reportChangeToEnclosingScope(n);
}
/** Calls {@link NodeUtil#deleteNode(Node, AbstractCompiler)} */
protected final void deleteNode(Node property) {
checkNotNull(compiler);
NodeUtil.deleteNode(property, compiler);
}
/** Calls {@link NodeUtil#markFunctionsDeleted(Node, AbstractCompiler)} */
protected final void markFunctionsDeleted(Node function) {
checkNotNull(compiler);
NodeUtil.markFunctionsDeleted(function, compiler);
}
/** Calls {@link NodeUtil#markNewScopesChanged(Node, AbstractCompiler)} */
protected final void markNewScopesChanged(Node n) {
checkNotNull(compiler);
NodeUtil.markNewScopesChanged(n, compiler);
}
}