From 5549706443f70ab6dd57421006c09b4fd03d16f5 Mon Sep 17 00:00:00 2001
From: Bill <sunyinqi0508@gmail.com>
Date: Sun, 23 Oct 2022 05:47:53 +0800
Subject: [PATCH] fixed issue for user module

---
 .gitignore             |  1 +
 Makefile               |  5 +++--
 reconstruct/ast.py     |  6 +++---
 sdk/Makefile           |  8 +++++++-
 sdk/irf.cpp            | 33 +++++++++++++++++++++++----------
 server/table.h         | 17 ++++++++++++++++-
 server/vector_type.hpp |  3 +++
 tests/dt.a             | 40 ++++++++++++++++++++--------------------
 tests/dt2.a            | 36 ++++++++++++++++--------------------
 tests/q1.sql           |  2 --
 10 files changed, 92 insertions(+), 59 deletions(-)

diff --git a/.gitignore b/.gitignore
index 644be8b..508685f 100644
--- a/.gitignore
+++ b/.gitignore
@@ -51,6 +51,7 @@ k
 **/Debug
 **/Release
 test*.c*
+data/benchmark
 *.csv
 !test.csv
 !test2.csv
diff --git a/Makefile b/Makefile
index dd7747e..1707240 100644
--- a/Makefile
+++ b/Makefile
@@ -4,11 +4,12 @@ MonetDB_INC =
 Threading = 
 CXXFLAGS = --std=c++1z
 ifeq ($(AQ_DEBUG), 1)
-	OPTFLAGS = -g3 
+	OPTFLAGS = -g3 -fsanitize=address -fsanitize=leak
+	LINKFLAGS = 
 else
 	OPTFLAGS = -O3 -DNDEBUG -fno-stack-protector 
+	LINKFLAGS = -flto
 endif
-LINKFLAGS = -flto # + $(AQ_LINK_FLAG)
 SHAREDFLAGS = -shared  
 FPIC = -fPIC
 COMPILER = $(shell $(CXX) --version | grep -q clang && echo clang|| echo gcc) 
diff --git a/reconstruct/ast.py b/reconstruct/ast.py
index d82ebce..173399b 100644
--- a/reconstruct/ast.py
+++ b/reconstruct/ast.py
@@ -343,7 +343,7 @@ class projection(ast_node):
                     )
                 else:
                     # for funcs evaluate f_i(x, ...)
-                    self.context.emitc(f'{self.out_table.contextname_cpp}->get_col<{key}>() = {val[1]};')
+                    self.context.emitc(f'{self.out_table.contextname_cpp}->get_col<{key}>().initfrom({val[1]}, "{cols[i].name}");')
         # print out col_is
         if 'into' not in node:
             self.context.emitc(f'print(*{self.out_table.contextname_cpp});')
@@ -990,7 +990,7 @@ class load(ast_node):
                 self.context.queries.append(f'F{fname}')
                 ret_type = VoidT
                 if 'ret_type' in f:
-                    ret_type = Types.decode(f['ret_type'])
+                    ret_type = Types.decode(f['ret_type'], vector_type='vector_type')
                 nargs = 0
                 arglist = ''
                 if 'vars' in f:
@@ -1000,7 +1000,7 @@ class load(ast_node):
                     nargs = len(arglist)
                     arglist = ', '.join(arglist)
                 # create c++ stub 
-                cpp_stub = f'{ret_type.cname} (*{fname})({arglist}) = nullptr;'
+                cpp_stub = f'{"vectortype_cstorage" if isinstance(ret_type, VectorT) else ret_type.cname} (*{fname})({arglist}) = nullptr;'
                 self.context.module_stubs += cpp_stub + '\n'
                 self.context.module_map[fname] = cpp_stub
                 #registration for parser
diff --git a/sdk/Makefile b/sdk/Makefile
index 7bd5c8c..b146a81 100644
--- a/sdk/Makefile
+++ b/sdk/Makefile
@@ -1,5 +1,11 @@
+OPT_FLASG = 
+ifneq ($(DEBUG), 1)
+	OPT_FLAGS = -Ofast -march=native -flto -DNDEBUG 
+else 
+	OPT_FLAGS = -g3 -D_DEBUG -fsanitize=leak -fsanitize=address
+endif
 example:
 	$(CXX) -shared -fPIC example.cpp aquery_mem.cpp -fno-semantic-interposition -Ofast -march=native -flto --std=c++1z -o ../test.so
 irf:
-	$(CXX) -shared -fPIC RF.cpp irf.cpp incrementalDecisionTree.cpp aquery_mem.cpp Evaluation.cpp -fno-semantic-interposition -Ofast -march=native -flto --std=c++1z -o ../libirf.so
+	$(CXX) -shared -fPIC RF.cpp irf.cpp incrementalDecisionTree.cpp aquery_mem.cpp Evaluation.cpp -fno-semantic-interposition $(OPT_FLAGS) --std=c++1z -o ../libirf.so
 all: example
diff --git a/sdk/irf.cpp b/sdk/irf.cpp
index 8433c95..36cf4c2 100644
--- a/sdk/irf.cpp
+++ b/sdk/irf.cpp
@@ -36,19 +36,32 @@ __AQEXPORT__(bool) additem(ColRef<double>X, long y, long size){
 	pt ++;
 	return 1;
 }
-__AQEXPORT__(bool) fit(){
-	if(pt<=0)return 0;
-	dt->fit(data, result, pt);
-	return 1;
+__AQEXPORT__(bool) fit(vector_type<vector_type<double>> v, vector_type<long> res){
+	double** data = (double**)malloc(v.size*sizeof(double*));
+	for(int i = 0; i < v.size; ++i)
+		data[i] = v.container[i].container;
+	dt->fit(data, res.container, v.size);
+	return true;
 }
 
-__AQEXPORT__(ColRef_storage) predict(){
-	int* result = (int*)malloc(pt*sizeof(int));
-	for(long i=0; i<pt; i++){
-		result[i]=dt->Test(data[i], dt->DTree);
-	}
+__AQEXPORT__(vectortype_cstorage) predict(vector_type<vector_type<double>> v){
+	int* result = (int*)malloc(v.size*sizeof(int));
 	
-	return ColRef_storage(new ColRef_storage(result, pt, 0, "prediction", 0), 1, 0, "prediction", 0);
+	for(long i=0; i<v.size; i++){
+		result[i]=dt->Test(v.container[i].container, dt->DTree);
+		//printf("%d ", result[i]);
+	}
+	auto container = (vector_type<int>*)malloc(sizeof(vector_type<int>));
+	container->size = v.size;
+	container->capacity = 0;
+	container->container = result;
+	// container->out(10);
+	// ColRef<vector_type<int>>* col = (ColRef<vector_type<int>>*)malloc(sizeof(ColRef<vector_type<int>>));
+	auto ret = vectortype_cstorage{.container = container, .size = 1, .capacity = 0};
+	// col->initfrom(ret, "sibal");
+	// print(*col);
+	return ret;
+	//return true;
 }
 
 
diff --git a/server/table.h b/server/table.h
index 56c7a4b..f3911af 100644
--- a/server/table.h
+++ b/server/table.h
@@ -74,7 +74,7 @@ public:
 		this->container = (_Ty*)container;
 		this->name = name;
 	}
-	template<template <typename ...> class VT, typename T>
+	template<template <typename> class VT, typename T>
 	void initfrom(const VT<T>& v, const char* name = "") {
 		ty = types::Types<_Ty>::getType();
 		this->size = v.size;
@@ -82,6 +82,21 @@ public:
 		this->container = (_Ty*)(v.container);
 		this->name = name;
 	}
+	void initfrom(vectortype_cstorage v, const char* name = "") {
+		ty = types::Types<_Ty>::getType();
+		this->size = v.size;
+		this->capacity = v.capacity;
+		this->container = (_Ty*)v.container;
+		this->name = name;
+	}
+	template<typename T>
+	void initfrom(const T& v, const char* name = "") {
+		ty = types::Types<_Ty>::getType();
+		this->size = 0;
+		this->capacity = 0;
+		this->emplace_back(v);
+		this->name = name;
+	}
 	template <class T>
 	ColRef<_Ty>& operator =(ColRef<T>&& vt) {
 		this->container = (_Ty*)vt.container;
diff --git a/server/vector_type.hpp b/server/vector_type.hpp
index 620e14d..98d79f5 100644
--- a/server/vector_type.hpp
+++ b/server/vector_type.hpp
@@ -77,6 +77,9 @@ public:
 	constexpr vector_type(vector_type<_Ty>&& vt) noexcept : capacity(0) {
 		_move(std::move(vt));
 	}
+	vector_type(vectortype_cstorage vt) noexcept : capacity(vt.capacity), size(vt.size), container((_Ty*)vt.container) {
+		out(10);
+	};
 	// size >= capacity ==> readonly vector
 	constexpr vector_type(const uint32_t size, void* data) : 
 		size(size), capacity(0), container(static_cast<_Ty*>(data)) {}
diff --git a/tests/dt.a b/tests/dt.a
index 5a52ac1..9ac773b 100644
--- a/tests/dt.a
+++ b/tests/dt.a
@@ -1,21 +1,21 @@
 LOAD MODULE FROM "./libirf.so"
-FUNCTIONS (
-    newtree(height:int, f:int64, sparse:vecint, forget:double, maxf:int64, noclasses:int64, e:int, r:int64, rb:int64) -> bool,
-    additem(X:vecdouble, y:int64, size:int64) -> bool,
-    fit() -> bool,
-    predict() -> vecint
-);
-create table tb(x int);
-create table tb2(x double, y double, z double);
-insert into tb values (0);
-insert into tb values (0);
-insert into tb values (0);
-select newtree(5, 3, tb.x, 0, 3, 2, 0, 100, 1) from tb;
-insert into tb2 values (1, 0, 1);
-insert into tb2 values (0, 1, 1);
-insert into tb2 values (1, 1, 1);
-select additem(tb2.x, 1, 3) from tb2;
-select additem(tb2.y, 0, -1) from tb2;
-select additem(tb2.z, 1, -1) from tb2;
-select fit();
-select predict();
+ FUNCTIONS (
+     newtree(height:int, f:int64, sparse:vecint, forget:double, maxf:int64, noclasses:int64, e:int, r:int64, rb:int64) -> bool,
+     additem(X:vecdouble, y:int64, size:int64) -> bool,
+     fit() -> bool,
+     predict() -> vecint
+ );
+ create table tb(x int);
+ create table tb2(x double, y double, z double);
+ insert into tb values (0);
+ insert into tb values (0);
+ insert into tb values (0);
+ select newtree(5, 3, tb.x, 0, 3, 2, 0, 100, 1) from tb;
+ insert into tb2 values (1, 0, 1);
+ insert into tb2 values (0, 1, 1);
+ insert into tb2 values (1, 1, 1);
+ select additem(tb2.x, 1, 3) from tb2;
+ select additem(tb2.y, 0, -1) from tb2;
+ select additem(tb2.z, 1, -1) from tb2;
+ select fit();
+ select predict();
\ No newline at end of file
diff --git a/tests/dt2.a b/tests/dt2.a
index 0f9dc7f..bd2bde7 100644
--- a/tests/dt2.a
+++ b/tests/dt2.a
@@ -1,26 +1,22 @@
 LOAD MODULE FROM "./libirf.so"
-FUNCTIONS (
-    newtree(height:int, f:int64, sparse:vecint64, forget:double, maxf:int64, noclasses:int64, e:int, r:int64, rb:int64) -> bool,
-    fit(X:vecvecdouble, y:vecint64) -> bool,
-    predict(X:vecvecdouble) -> vecint64
-);
+ FUNCTIONS (
+     newtree(height:int, f:int64, sparse:vecint, forget:double, maxf:int64, noclasses:int64, e:int, r:int64, rb:int64) -> bool,
+     fit(X:vecvecdouble, y:vecint64) -> bool,
+     predict(X:vecvecdouble) -> vecint
+ );
 
-create table source(x1 double, x2 double, x3 double, x4 double, x5 int64);
-load data infile "data/benchmark" into table source fields terminated by ",";
+ create table source(x1 double, x2 double, x3 double, x4 double, x5 int64);
+ load data infile "data/benchmark" into table source fields terminated by ",";
 
-create table sparse(x int64);
-insert into sparse values (1);
-insert into sparse values (1);
-insert into sparse values (1);
-insert into sparse values (1);
+ create table sparse(x int);
+ insert into sparse values (1);
+ insert into sparse values (1);
+ insert into sparse values (1);
+ insert into sparse values (1);
 
-select * from source;
+ select newtree(6, 4, sparse.x, 0, 4, 2, 0, 400, 2147483647) from sparse
 
-select newtree(6, 4, sparse.x, 0, 4, 2, 0, 400, 2147483647) from sparse;
+ select fit(pack(x1, x2, x3, x4), x5) from source
 
-select fit(pack(x1, x2, x3, x4), x5) from source limit 100;
-select fit(pack(x1, x2, x3, x4), x5) from source limit 100;
-select fit(pack(x1, x2, x3, x4), x5) from source limit 100;
-select fit(pack(x1, x2, x3, x4), x5) from source limit 100;
-
-select predict(pack(x1, x2, x3, x4)) from source limit 100;
\ No newline at end of file
+-- select pack(x1, x2, x3, x4) from source
+  select predict(pack(x1, x2, x3, x4)) from source
\ No newline at end of file
diff --git a/tests/q1.sql b/tests/q1.sql
index eab8904..b57c2d1 100644
--- a/tests/q1.sql
+++ b/tests/q1.sql
@@ -8,5 +8,3 @@ SELECT sum(c), b, d
 FROM testq1
 group by a,b,d
 order by d DESC, b ASC;
-
--- aaaa
\ No newline at end of file