source: src/CodeTools/DeclStats.cc @ 14a33790

//
// Cforall Version 1.0.0 Copyright (C) 2015 University of Waterloo
//
// The contents of this file are covered under the licence agreement in the
// file "LICENCE" distributed with Cforall.
//
// DeclStats.cc --
//
// Author           : Aaron Moss
// Created On       : Wed Jan 31 16:40:00 2016
// Last Modified By : Aaron Moss
// Last Modified On : Wed Jan 31 16:40:00 2016
// Update Count     : 1
//

#include "DeclStats.h"

#include <iostream>
#include <map>
#include <sstream>
#include <string>
#include <unordered_map>
#include <unordered_set>

#include "Common/VectorMap.h"
#include "GenPoly/GenPoly.h"
#include "Parser/LinkageSpec.h"
#include "SynTree/Declaration.h"
#include "SynTree/Visitor.h"

namespace CodeTools {
        
        class DeclStats : public Visitor {
                template<typename T>
                static void sum(T& a, const T& b) { a += b; }

                static void sum(VectorMap<unsigned>& a, const VectorMap<unsigned>& b) {
                        a.reserve( b.size() );
                        for ( unsigned i = 0; i < b.size(); ++i ) {
                                a[i] += b[i];
                        }
                }

                template<typename K>
                static void sum(std::map<K, unsigned>& a, const std::map<K, unsigned>& b) {
                        for ( const auto& entry : b ) {
                                a[ entry.first ] += entry.second;
                        }
                }

                template<typename K>
                static void sum(std::unordered_map<K, unsigned>& a, const std::unordered_map<K, unsigned>& b) {
                        for ( const auto& entry : b ) {
                                a[ entry.first ] += entry.second;
                        }
                }

                struct ArgPackStats {
                        VectorMap<unsigned> n;                 ///< Count of decls with each number of elements
                        VectorMap<unsigned> n_basic;           ///< Count of decls with each number of basic type elements
                        VectorMap<unsigned> n_poly;            ///< Count of decls with each number of polymorphic elements
                        std::map<unsigned, unsigned> p_basic;  ///< Count of decls with each percentage of basic type elements
                        std::map<unsigned, unsigned> p_poly;   ///< Count of decls with each percentage of polymorphic elements
                        VectorMap<unsigned> n_types;           ///< Count of decls with each number of distinct types in the pack
                        /// Count of decls with each percentage of new types in lists.
                        /// Types used in the parameter list that recur in the return list are not considered to be new.
                        std::map<unsigned, unsigned> p_new;

                        ArgPackStats& operator+= (const ArgPackStats& o) {
                                sum(n, o.n);
                                sum(n_basic, o.n_basic);
                                sum(n_poly, o.n_poly);
                                sum(p_basic, o.p_basic);
                                sum(p_poly, o.p_poly);
                                sum(n_types, o.n_types);
                                sum(p_new, o.p_new);
                                
                                return *this;
                        }
                };
                
                struct Stats {
                        unsigned n_decls;     ///< Total number of declarations
                        /// Count of declarations with each number of assertion parameters
                        VectorMap<unsigned> n_type_params;
                        /// Count of declarations with each name
                        std::unordered_map<std::string, unsigned> by_name;
                        /// Count of uses of each basic type
                        std::unordered_map<std::string, unsigned> basic_type_names;
                        /// Count of uses of each non-basic type
                        std::unordered_map<std::string, unsigned> compound_type_names;
                        /// Count of decls using each basic type
                        std::unordered_map<std::string, unsigned> basic_type_decls;
                        /// Count of decls using each compound type
                        std::unordered_map<std::string, unsigned> compound_type_decls;
                        /// Stats for the parameter list
                        ArgPackStats params;
                        /// Stats for the return list
                        ArgPackStats returns;
                        
                        /// Count of declarations with each number of assertions
                        std::map<unsigned, unsigned> n_assns;
                        /// Stats for the assertions' parameters
                        ArgPackStats assn_params;
                        /// Stats for the assertions' return types
                        ArgPackStats assn_returns;
                        
                        Stats() : n_decls(0), n_type_params(), by_name(), basic_type_names(), compound_type_names(), basic_type_decls(), compound_type_decls(), params(), returns(), n_assns(), assn_params(), assn_returns() {}

                public:
                        Stats& operator+= (const Stats& o) {
                                sum( n_decls, o.n_decls );
                                sum( n_type_params, o.n_type_params );
                                sum( by_name, o.by_name );
                                sum( basic_type_names, o.basic_type_names );
                                sum( compound_type_names, o.compound_type_names );
                                sum( basic_type_decls, o.basic_type_decls );
                                sum( compound_type_decls, o.compound_type_decls );
                                sum( params, o.params );
                                sum( returns, o.returns );
                                sum( n_assns, o.n_assns );
                                sum( assn_params, o.assn_params );
                                sum( assn_returns, o.assn_returns );
                                
                                return *this;
                        }
                };

                Stats for_linkage[LinkageSpec::NoOfSpecs];   ///< Stores separate stats per linkage
                std::unordered_set<std::string> seen_names;  ///< Stores manglenames already seen to avoid double-counting
                Stats total;
                /// Count of expressions with (depth, fanout)
                std::map<std::pair<unsigned, unsigned>, unsigned> exprs_by_fanout_at_depth;

                /// Update arg pack stats based on a declaration list
                void analyze( Stats& stats, std::unordered_set<std::string>& seen, ArgPackStats& pstats, std::list<DeclarationWithType*>& decls ) {
                        std::unordered_set<std::string> types;
                        unsigned n = 0;        ///< number of args/returns
                        unsigned n_basic = 0;  ///< number of basic types
                        unsigned n_poly = 0;   ///< number of polymorphic types
                        unsigned n_new = 0;    ///< number of new types
                        for ( auto decl : decls ) {
                                Type* dt = decl->get_type();

                                n += dt->size();
                                
                                std::stringstream ss;
                                dt->print( ss );
                                types.insert( ss.str() );
                                bool this_new = seen.insert( ss.str() ).second;
                                if ( this_new ) { ++n_new; }

                                if ( dynamic_cast<BasicType*>( dt ) ) {
                                        ++n_basic;
                                        ++stats.basic_type_names[ ss.str() ];
                                        if ( this_new ) {
                                                ++stats.basic_type_decls[ ss.str() ];
                                        }
                                } else if ( GenPoly::hasPolyBase( dt ) ) {
                                        ++n_poly;
                                } else {
                                        ++stats.compound_type_names[ ss.str() ];
                                        if ( this_new ) {
                                                ++stats.compound_type_decls[ ss.str() ];
                                        }
                                }
                        }
                        ++pstats.n.at( n );
                        ++pstats.n_basic.at( n_basic );
                        ++pstats.n_poly.at( n_poly );
                        if ( n > 0 ) {
                                ++pstats.p_basic[ n_basic*100/n ];
                                ++pstats.p_poly[ n_poly*100/n ];
                                ++pstats.p_new[ n_new*100/n ];
                        }
                        ++pstats.n_types.at( types.size() );
                }
                
                void analyzeFunc( FunctionType* fnTy, Stats& stats, ArgPackStats& params, ArgPackStats& returns ) {
                        std::unordered_set<std::string> seen;
                        analyze( stats, seen, params, fnTy->get_parameters() );
                        analyze( stats, seen, returns, fnTy->get_returnVals() );
                }

                void analyzeExpr( UntypedExpr *expr, unsigned depth ) {
                        auto& args = expr->get_args();
                        unsigned fanout = args.size();
                        
                        ++exprs_by_fanout_at_depth[ std::make_pair(depth, fanout) ];
                        for ( Expression* arg : args ) {
                                if ( UntypedExpr *uearg = dynamic_cast<UntypedExpr*>(arg) ) {
                                        analyzeExpr( uearg, depth+1 );
                                }
                        }
                }

        public:
                using Visitor::visit;

                virtual void visit( FunctionDecl *decl ) {
                        // skip if already seen declaration for this function
                        const std::string& mangleName = decl->get_mangleName().empty() ? decl->get_name() : decl->get_mangleName();
                        if ( ! seen_names.insert( mangleName ).second ) {
                                maybeAccept( decl->get_statements(), *this );
                                return;
                        }
                        
                        Stats& stats = for_linkage[ decl->get_linkage() ];

                        ++stats.n_decls;
                        FunctionType* fnTy = decl->get_functionType();
                        const Type::ForallList& forall = fnTy->get_forall();
                        ++stats.n_type_params.at( forall.size() );
                        unsigned n_assns = 0;
                        for ( TypeDecl* fdecl : forall ) {
                                n_assns += fdecl->get_assertions().size();
                                for ( DeclarationWithType* assn : fdecl->get_assertions() ) {
                                        FunctionType *assnTy = 0;
                                        if ( ObjectDecl *assnObj = dynamic_cast<ObjectDecl*>(assn) ) {
                                                if ( PointerType *ptrTy = dynamic_cast<PointerType*>(assnObj->get_type()) ) {
                                                        assnTy = dynamic_cast<FunctionType*>(ptrTy->get_base());
                                                } else assnTy = dynamic_cast<FunctionType*>(assnObj->get_type());
                                        } else if ( FunctionDecl *assnDecl = dynamic_cast<FunctionDecl*>(assn) ) {
                                                assnTy = assnDecl->get_functionType();
                                        }
                                        if ( assnTy ) analyzeFunc( assnTy, stats, stats.assn_params, stats.assn_returns );
                                }
                        }
                        ++stats.n_assns[ n_assns ];

                        ++stats.by_name[ decl->get_name() ];

                        analyzeFunc( fnTy, stats, stats.params, stats.returns );

                        // analyze expressions in decl statements
                        maybeAccept( decl->get_statements(), *this );
                }

                virtual void visit( UntypedExpr *expr ) {
                        analyzeExpr( expr, 0 );
                }

        private:
                template<typename F>
                void printAll( const std::string& name, F extract ) {
                        std::cout << "\"" << name << "\",";
                        for ( const auto& stats : for_linkage ) {
                                std::cout << "," << extract(stats);
                        }
                        std::cout << "," << extract(total) << std::endl;
                }

                template<typename F>
                void printAllMap( const std::string& name, F extract ) {
                        for ( const auto& entry : extract(total) ) {
                                const auto& key = entry.first;
                                std::cout << "\"" << name << "\"," << key;
                                for ( const auto& stats : for_linkage ) {
                                        const auto& map = extract(stats);
                                        auto it = map.find( key );
                                        if ( it == map.end() ) std::cout << ",0";
                                        else std::cout << "," << it->second;
                                }
                                std::cout  << "," << entry.second << std::endl;
                        }
                }

                template<typename F>
                void printAllHisto( const std::string& name, F extract ) {
                        VectorMap<unsigned> histos[LinkageSpec::NoOfSpecs];
                        VectorMap<unsigned> thisto;

                        for ( const auto& entry : extract(total) ) { ++thisto.at( entry.second ); }

                        for ( unsigned i = 0; i < LinkageSpec::NoOfSpecs; ++i ) {
                                // can't be a higher count in one of the sub-histograms than the total
                                histos[i].reserve( thisto.size() );

                                for ( const auto& entry : extract(for_linkage[i]) ) { ++histos[i][entry.second]; }
                        }

                        for ( unsigned i = 0; i < thisto.size(); ++i ) {
                                std::cout << "\"" << name << "\"," << i;
                                for ( const auto& histo : histos ) {
                                        std::cout << "," << histo[i];
                                }
                                std::cout << "," << thisto[i] << std::endl;
                        }
                }

                template<typename F>
                void printAllSparseHisto( const std::string& name, F extract ) {
                        std::map<unsigned, unsigned> histos[LinkageSpec::NoOfSpecs];
                        std::map<unsigned, unsigned> thisto;

                        for ( const auto& entry : extract(total) ) { ++thisto[ entry.second ]; }

                        for ( unsigned i = 0; i < LinkageSpec::NoOfSpecs; ++i ) {
                                for ( const auto& entry : extract(for_linkage[i]) ) { ++histos[i][entry.second]; }
                        }

                        for ( const auto& entry : thisto ) {
                                const auto& key = entry.first;
                                std::cout << "\"" << name << "\"," << key;
                                for ( unsigned i = 0; i < LinkageSpec::NoOfSpecs; ++i ) {
                                        auto it = histos[i].find( key );
                                        if ( it == histos[i].end() ) std::cout << ",0";
                                        else std::cout << "," << it->second;
                                }
                                std::cout << "," << entry.second << std::endl;
                        }
                }

                template<typename F>
                void printAllPack( const std::string& name, F extract ) {
                        printAllMap("n_basic_" + name, [&extract](const Stats& stats) { return extract(stats).n_basic; });
                        printAllMap("n_poly_" + name, [&extract](const Stats& stats) { return extract(stats).n_poly; });
                        printAllMap("n_" + name, [&extract](const Stats& stats) { return extract(stats).n; });
                        printAllMap("%_basic_" + name, [&extract](const Stats& stats) { return extract(stats).p_basic; });
                        printAllMap("%_poly_" + name, [&extract](const Stats& stats) { return extract(stats).p_poly; });
                        printAllMap("n_distinct_types_" + name, [&extract](const Stats& stats) { return extract(stats).n_types; });
                        printAllMap("%_new_types_in_" + name, [&extract](const Stats& stats) { return extract(stats).p_new; });
                }

                void printPairMap( const std::string& name, 
                                   const std::map<std::pair<unsigned, unsigned>, unsigned>& map ) {
                        for ( const auto& entry : map ) {
                                const auto& key = entry.first;
                                std::cout << "\"" << name << "\"," << key.first << "," << key.second << "," 
                                          << entry.second << std::endl;
                        }
                }
                
        public:
                void print() {
                        for ( auto& stats : for_linkage ) {
                                total += stats;
                        }

                        std::cout << ",,\"intrinsic\",\"Cforall\",\"C\",\"autogen\",\"builtin\",\"TOTAL\"" << std::endl;

                        printAllMap("n_type_params", [](const Stats& stats) { return stats.n_type_params; });
                        printAll("n_decls", [](const Stats& stats) { return stats.n_decls; });
                        printAll("unique_names", [](const Stats& stats) { return stats.by_name.size(); });
                        printAllSparseHisto("overloads", [](const Stats& stats) { return stats.by_name; });
                        printAll("basic_type_names", [](const Stats& stats) { return stats.basic_type_names.size(); });
                        printAllSparseHisto("basic_type_uses", [](const Stats& stats) { return stats.basic_type_names; });
                        printAllSparseHisto("decls_using_basic_type", [](const Stats& stats) { return stats.basic_type_decls; });
                        printAll("compound_type_names", [](const Stats& stats) { return stats.compound_type_names.size(); });
                        printAllSparseHisto("compound_type_uses", [](const Stats& stats) { return stats.compound_type_names; });
                        printAllSparseHisto("decls_using_compound_type", [](const Stats& stats) { return stats.compound_type_decls; });
                        printAllPack("params", [](const Stats& stats) { return stats.params; });
                        printAllPack("returns", [](const Stats& stats) { return stats.returns; });
                        printAllMap("n_assns", [](const Stats& stats) { return stats.n_assns; });
                        printAllPack("assn_params", [](const Stats& stats) { return stats.assn_params; });
                        printAllPack("assn_returns", [](const Stats& stats) { return stats.assn_returns; });
                        std::cout << std::endl;

                        printPairMap("exprs_by_depth+fanout", exprs_by_fanout_at_depth);
                }
        };

        void printDeclStats( std::list< Declaration * > &translationUnit ) {
                DeclStats stats;
                acceptAll( translationUnit, stats );
                stats.print();
        }
        
} // namespace CodeTools

// Local Variables: //
// tab-width: 4 //
// mode: c++ //
// compile-command: "make install" //
// End: //