From 005a6a19ee0af3e284637abe9ecea5501993fc4c Mon Sep 17 00:00:00 2001
From: Scott Wolchok <swolchok@fb.com>
Date: Thu, 7 Jan 2016 15:16:52 -0800
Subject: [PATCH] Avoid copying cached RegExes for scalars

This improves performance on the test.yaml attached to #158 by about
25% on my machine as compared to the previous commit (0.25s -> 0.20s),
as measured by `time build/util/parse < test.yaml > /dev/null`.
---
 src/exp.h          | 4 ++++
 src/scanscalar.cpp | 8 ++++++--
 src/scanscalar.h   | 6 ++++--
 src/scantoken.cpp  | 5 +++--
 4 files changed, 17 insertions(+), 6 deletions(-)

diff --git a/src/exp.h b/src/exp.h
index ba82874..50b0220 100644
--- a/src/exp.h
+++ b/src/exp.h
@@ -20,6 +20,10 @@ namespace YAML {
 
 namespace Exp {
 // misc
+inline const RegEx& Empty() {
+  static const RegEx e;
+  return e;
+}
 inline const RegEx& Space() {
   static const RegEx e = RegEx(' ');
   return e;
diff --git a/src/scanscalar.cpp b/src/scanscalar.cpp
index 8253b8d..62ac875 100644
--- a/src/scanscalar.cpp
+++ b/src/scanscalar.cpp
@@ -28,13 +28,17 @@ std::string ScanScalar(Stream& INPUT, ScanScalarParams& params) {
   std::string scalar;
   params.leadingSpaces = false;
 
+  if (!params.end) {
+    params.end = &Exp::Empty();
+  }
+
   while (INPUT) {
     // ********************************
     // Phase #1: scan until line ending
 
     std::size_t lastNonWhitespaceChar = scalar.size();
     bool escapedNewline = false;
-    while (!params.end.Matches(INPUT) && !Exp::Break().Matches(INPUT)) {
+    while (!params.end->Matches(INPUT) && !Exp::Break().Matches(INPUT)) {
       if (!INPUT)
         break;
 
@@ -87,7 +91,7 @@ std::string ScanScalar(Stream& INPUT, ScanScalarParams& params) {
       break;
 
     // are we done via character match?
-    int n = params.end.Match(INPUT);
+    int n = params.end->Match(INPUT);
     if (n >= 0) {
       if (params.eatEnd)
         INPUT.eat(n);
diff --git a/src/scanscalar.h b/src/scanscalar.h
index 62da13c..c3a574a 100644
--- a/src/scanscalar.h
+++ b/src/scanscalar.h
@@ -19,7 +19,8 @@ enum FOLD { DONT_FOLD, FOLD_BLOCK, FOLD_FLOW };
 
 struct ScanScalarParams {
   ScanScalarParams()
-      : eatEnd(false),
+      : end(nullptr),
+        eatEnd(false),
         indent(0),
         detectIndent(false),
         eatLeadingWhitespace(0),
@@ -32,7 +33,8 @@ struct ScanScalarParams {
         leadingSpaces(false) {}
 
   // input:
-  RegEx end;          // what condition ends this scalar?
+  const RegEx* end;   // what condition ends this scalar?
+                      // unowned.
   bool eatEnd;        // should we eat that condition when we see it?
   int indent;         // what level of indentation should be eaten and ignored?
   bool detectIndent;  // should we try to autodetect the indent?
diff --git a/src/scantoken.cpp b/src/scantoken.cpp
index 79e7682..fd8758d 100644
--- a/src/scantoken.cpp
+++ b/src/scantoken.cpp
@@ -298,7 +298,7 @@ void Scanner::ScanPlainScalar() {
   // set up the scanning parameters
   ScanScalarParams params;
   params.end =
-      (InFlowContext() ? Exp::ScanScalarEndInFlow() : Exp::ScanScalarEnd());
+      (InFlowContext() ? &Exp::ScanScalarEndInFlow() : &Exp::ScanScalarEnd());
   params.eatEnd = false;
   params.indent = (InFlowContext() ? 0 : GetTopIndent() + 1);
   params.fold = FOLD_FLOW;
@@ -338,7 +338,8 @@ void Scanner::ScanQuotedScalar() {
 
   // setup the scanning parameters
   ScanScalarParams params;
-  params.end = (single ? RegEx(quote) && !Exp::EscSingleQuote() : RegEx(quote));
+  RegEx end = (single ? RegEx(quote) && !Exp::EscSingleQuote() : RegEx(quote));
+  params.end = &end;
   params.eatEnd = true;
   params.escape = (single ? '\'' : '\\');
   params.indent = 0;