From 173d1fd1c3fef385f73accc4b2bbb1b6f92ac3f5 Mon Sep 17 00:00:00 2001
From: Paul Guyot <pguyot@kallisys.net>
Date: Sun, 4 Jul 2010 15:08:47 +0200
Subject: beam_asm: Simplify string table generation for beams

The code for generating the string table (which is now
only used for bit syntax matching) in a BEAM file is quite
complicated and potentially expensive when compiling modules
with many thousands of clauses doing bit syntax matching.

Simplify and optimize the code using bit syntax and
binary:match/2 instead of the list operations in the
original code.
---
 lib/compiler/src/beam_dict.erl | 30 ++++++++++++++----------------
 1 file changed, 14 insertions(+), 16 deletions(-)

(limited to 'lib/compiler/src/beam_dict.erl')

diff --git a/lib/compiler/src/beam_dict.erl b/lib/compiler/src/beam_dict.erl
index 9164259a94..a1f994dfbd 100644
--- a/lib/compiler/src/beam_dict.erl
+++ b/lib/compiler/src/beam_dict.erl
@@ -33,7 +33,7 @@
 	 exports = []		    :: [{label(), arity(), label()}],
 	 locals = []		    :: [{label(), arity(), label()}],
 	 imports = gb_trees:empty() :: gb_tree(),      	%{{M,F,A},Index}
-	 strings = []		    :: string(),	%String pool
+	 strings = <<>>		    :: binary(),	%String pool
 	 lambdas = [],				%[{...}]
 	 literals = dict:new()	    :: dict(),	%Format: {Literal,Number}
 	 next_atom = 1		    :: pos_integer(),
@@ -119,10 +119,11 @@ import(Mod0, Name0, Arity, #asm{imports=Imp0,next_import=NextIndex}=D0)
 
 string(Str, Dict) when is_list(Str) ->
     #asm{strings=Strings,string_offset=NextOffset} = Dict,
-    case old_string(Str, Strings) of
+    StrBin = list_to_binary(Str),
+    case old_string(StrBin, Strings) of
 	none ->
-	    NewDict = Dict#asm{strings=Strings++Str,
-			       string_offset=NextOffset+length(Str)},
+	    NewDict = Dict#asm{strings = <<Strings/binary,StrBin/binary>>,
+			       string_offset=NextOffset+byte_size(StrBin)},
 	    {NextOffset,NewDict};
 	Offset when is_integer(Offset) ->
 	    {NextOffset-Offset,Dict}
@@ -187,7 +188,7 @@ import_table(#asm{imports=Imp,next_import=NumImports}) ->
     ImpTab = [MFA || {MFA,_} <- Sorted],
     {NumImports,ImpTab}.
 
--spec string_table(bdict()) -> {non_neg_integer(), [string()]}.
+-spec string_table(bdict()) -> {non_neg_integer(), binary()}.
 
 string_table(#asm{strings=Strings,string_offset=Size}) ->
     {Size,Strings}.
@@ -217,15 +218,12 @@ literal_table(#asm{literals=Tab,next_literal=NumLiterals}) ->
 my_term_to_binary(Term) ->
     term_to_binary(Term, [{minor_version,1}]).
 
-%% Search for string Str in the string pool Pool.
+%% Search for binary string Str in the binary string pool Pool.
 %%    old_string(Str, Pool) -> none | Index
--spec old_string(string(), string()) -> 'none' | pos_integer().
-
-old_string([C|Str]=Str0, [C|Pool]) ->
-    case lists:prefix(Str, Pool) of
-	true -> length(Pool)+1;
-	false -> old_string(Str0, Pool)
-    end;
-old_string([_|_]=Str, [_|Pool]) ->
-    old_string(Str, Pool);
-old_string([_|_], []) -> none.
+-spec old_string(binary(), binary()) -> 'none' | pos_integer().
+
+old_string(Str, Pool) ->
+    case binary:match(Pool, Str) of
+        nomatch -> none;
+        {Start,_Length} -> byte_size(Pool) - Start
+    end.
-- 
cgit v1.2.3