hiifong/pmd
1
0
Fork 0
forked from phoedos/pmd
Code Pull Requests Activity

Prevent double counting time

Browse Source 
- First off, all LPS are benchmarked in the AstProcessor, having this here was inconsistent
 - Since these passes can be triggered by the symbol table LPS phase, we ended up double counting total time.
 - This however puts all disambiguation passes done for symbol table as symbol table cost, which although accurate,
may not help to identify speed up opportunities as clearly, but the benchmark is not a profiling tool.
This commit is contained in:
Juan Martín Sotuyo Dodero
2023-01-25 17:52:53 -03:00
parent c5d410367a
commit 521e020100
1 changed files with 1 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
pmd-java/src/main/java/net/sourceforge/pmd/lang/java/ast/AstDisambiguationPass.java
View File

@ -12,7 +12,6 @@ import java.util.Iterator;
import org.checkerframework.checker.nullness.qual.NonNull;
import org.checkerframework.checker.nullness.qual.Nullable;
import net.sourceforge.pmd.benchmark.TimeTracker;
import net.sourceforge.pmd.lang.ast.Node;
import net.sourceforge.pmd.lang.ast.NodeStream;
import net.sourceforge.pmd.lang.ast.impl.javacc.JavaccToken;
@ -56,7 +55,7 @@ final class AstDisambiguationPass {
*/
public static void disambigWithCtx(NodeStream<? extends JavaNode> nodes, ReferenceCtx ctx) {
assert ctx != null : "Null context";
TimeTracker.bench("AST disambiguation", () -> nodes.forEach(it -> it.acceptVisitor(DisambigVisitor.INSTANCE, ctx)));
nodes.forEach(it -> it.acceptVisitor(DisambigVisitor.INSTANCE, ctx));
}