BeeEngine/glslang-zig
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Preprocessor: 1) Rationalize the "fixed atom" scheme, 2) remove redundant lookups when the text is already available.

Browse source 
This simplification is a prelude to eliminating what I appear unnecessary
symbol inserts into tables when tokenizing in the preprecessor, which
show up as taking notable time. (Performance issue.)  It also simply makes
the preprocessor easier to understand, which it is badly in need of.
This commit is contained in:
John Kessenich 2015-07-19 15:49:03 -06:00
parent 2f273369e4
commit 6ab3d582d6
10 changed files with 362 additions and 412 deletions
Download patch file Download diff file Expand all files Collapse all files

2
glslang/MachineIndependent/ParseHelper.h
View file

@ -306,7 +306,7 @@ protected:
TInputScanner* currentScanner;
int numErrors; // number of compile-time errors encountered
bool parsingBuiltins; // true if parsing built-in symbols/functions
std::unordered_map<TString, TExtensionBehavior> extensionBehavior; // for each extension string, what its current behavior is set to
TUnorderedMap<TString, TExtensionBehavior> extensionBehavior; // for each extension string, what its current behavior is set to
static const int maxSamplerIndex = EsdNumDims * (EbtNumTypes * (2 * 2 * 2)); // see computeSamplerTypeIndex()
TPrecisionQualifier defaultSamplerPrecision[maxSamplerIndex];
bool afterEOF;