1------------------------------------------------------------------------------ 2-- -- 3-- GNAT COMPILER COMPONENTS -- 4-- -- 5-- S E M _ D I M -- 6-- -- 7-- S p e c -- 8-- -- 9-- Copyright (C) 2011-2012, Free Software Foundation, Inc. -- 10-- -- 11-- GNAT is free software; you can redistribute it and/or modify it under -- 12-- terms of the GNU General Public License as published by the Free Soft- -- 13-- ware Foundation; either version 3, or (at your option) any later ver- -- 14-- sion. GNAT is distributed in the hope that it will be useful, but WITH- -- 15-- OUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY -- 16-- or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License -- 17-- for more details. You should have received a copy of the GNU General -- 18-- Public License distributed with GNAT; see file COPYING3. If not, go to -- 19-- http://www.gnu.org/licenses for a complete copy of the license. -- 20-- -- 21-- GNAT was originally developed by the GNAT team at New York University. -- 22-- Extensive contributions were provided by Ada Core Technologies Inc. -- 23-- -- 24------------------------------------------------------------------------------ 25 26-- This package provides support for numerical systems with dimensions. A 27-- "dimension" is a compile-time property of a numerical type which represents 28-- a relation between various quantifiers such as length, velocity, etc. 29 30-- Package System.Dim.Mks offers a ready-to-use system of SI base units. In 31-- addition, the implementation of this feature offers the ability to define 32-- an arbitrary system of units through the use of Ada 2012 aspects. 33 34-- Dimensionality checking is part of type analysis performed by the compiler. 35-- It ensures that manipulation of quantified numeric values is sensible with 36-- respect to the system of units. 37 38----------------------------- 39-- Aspect_Dimension_System -- 40----------------------------- 41 42-- In order to enable the user to create their own dimension system, a new 43-- aspect: Aspect_Dimension_System has been created. 44 45-- Note that this aspect applies for type declaration of type derived from any 46-- numeric type. 47 48-- It defines the names of each dimension 49 50---------------------- 51-- Aspect_Dimension -- 52---------------------- 53 54-- This new aspect applies for subtype and object declarations in order to 55-- define new dimensions. 56 57-- Using this aspect, the user is able to create new subtype/object with any 58-- dimension needed. 59 60-- Note that the base type of the subtype/object must be the type that defines 61-- the corresponding dimension system. 62 63-- The expression of this aspect is an aggregate of rational values for each 64-- dimension in the corresponding dimension system. 65 66------------------------------------------- 67-- Dimensionality checking & propagation -- 68------------------------------------------- 69 70-- For each node (when needed), a dimension analysis (Analyze_Dimension) is 71-- performed as part of the Resolution routine or the Analysis routine if no 72-- Resolution. 73 74-- The dimension analysis is divided into two phases: 75 76-- Phase 1: dimension checking 77 78-- Phase 2: propagation of dimensions 79 80-- Depending on the node kind, either none, one phase or two phases are 81-- executed. 82 83-- Phase 2 is called only when the node allows a dimension (see body of 84-- Sem_Dim to get the list of nodes that permit dimensions). 85 86with Types; use Types; 87 88package Sem_Dim is 89 90 procedure Analyze_Aspect_Dimension 91 (N : Node_Id; 92 Id : Entity_Id; 93 Aggr : Node_Id); 94 -- Analyze the contents of aspect Dimension. Associate the provided values 95 -- and quantifiers with the related context N. Id is the corresponding 96 -- Aspect_Id (Aspect_Dimension) Aggr is the corresponding expression for 97 -- the aspect Dimension declared by the declaration of N. 98 99 procedure Analyze_Aspect_Dimension_System 100 (N : Node_Id; 101 Id : Entity_Id; 102 Aggr : Node_Id); 103 -- Analyze the contents of aspect Dimension_System. Extract the numerical 104 -- type, unit name and corresponding symbol from each indivitual dimension. 105 -- Id is the corresponding Aspect_Id (Aspect_Dimension_System) 106 -- Aggr is the corresponding expression for the aspect Dimension_System 107 -- declared by the declaration of N. 108 109 procedure Analyze_Dimension (N : Node_Id); 110 -- N may denote any of the following contexts: 111 -- * aggregate 112 -- * assignment statement 113 -- * attribute reference 114 -- * binary operator 115 -- * call 116 -- * compontent declaration 117 -- * extended return statement 118 -- * expanded name 119 -- * identifier 120 -- * indexed component 121 -- * object declaration 122 -- * object renaming declaration 123 -- * procedure call statement 124 -- * qualified expression 125 -- * selected component 126 -- * simple return statement 127 -- * slice 128 -- * subtype declaration 129 -- * type conversion 130 -- * unary operator 131 -- * unchecked type conversion 132 -- Depending on the context, ensure that all expressions and entities 133 -- involved do not violate the rules of a system. 134 135 procedure Analyze_Dimension_Array_Aggregate 136 (N : Node_Id; 137 Comp_Typ : Entity_Id); 138 -- Check, for each component of the array aggregate denoted by N, the 139 -- dimensions of the component expression match the dimensions of the 140 -- component type Comp_Typ. 141 142 procedure Analyze_Dimension_Call (N : Node_Id; Nam : Entity_Id); 143 -- This routine is split in two steps. Note the second step applies only to 144 -- function calls. 145 -- Step 1. Dimension checking: 146 -- * General case: check the dimensions of each actual parameter match 147 -- the dimensions of the corresponding formal parameter. 148 -- * Elementary function case: check each actual is dimensionless except 149 -- for Sqrt call. 150 -- Step 2. Dimension propagation (only for functions): 151 -- * General case: propagate the dimensions from the returned type to the 152 -- function call. 153 -- * Sqrt case: the resulting dimensions equal to half the dimensions of 154 -- the actual 155 156 procedure Analyze_Dimension_Extension_Or_Record_Aggregate (N : Node_Id); 157 -- Check, for each component of the extension or record aggregate denoted 158 -- by N, the dimensions of the component expression match the dimensions of 159 -- the component type. 160 161 procedure Analyze_Dimension_Formals (N : Node_Id; Formals : List_Id); 162 -- For sub spec N, issue a warning for each dimensioned formal with a 163 -- literal default value in the list of formals Formals. 164 165 procedure Copy_Dimensions (From, To : Node_Id); 166 -- Copy dimension vector of node From to node To. Note that To must be a 167 -- node that is allowed to contain a dimension (see OK_For_Dimension in 168 -- body of Sem_Dim). 169 170 procedure Eval_Op_Expon_For_Dimensioned_Type 171 (N : Node_Id; 172 Btyp : Entity_Id); 173 -- Evaluate the Expon operator for dimensioned type with rational exponent. 174 -- Indeed the regular Eval_Op_Expon routine (see package Sem_Eval) is 175 -- restricted to Integer exponent. This routine deals only with rational 176 -- exponent which is not an integer if Btyp is a dimensioned type. 177 178 procedure Expand_Put_Call_With_Symbol (N : Node_Id); 179 -- Determine whether N denotes a subprogram call to one of the routines 180 -- defined in System.Dim.Float_IO or System.Dim.Integer_IO and add an 181 -- extra actual to the call to represent the symbolic representation of 182 -- a dimension. 183 184 function Has_Dimension_System (Typ : Entity_Id) return Boolean; 185 -- Return True if type Typ has aspect Dimension_System applied to it 186 187 function Is_Dim_IO_Package_Instantiation (N : Node_Id) return Boolean; 188 -- Return True if N is a package instantiation of System.Dim.Integer_IO or 189 -- of System.Dim.Float_IO. 190 191 procedure Remove_Dimension_In_Statement (Stmt : Node_Id); 192 -- Remove the dimensions associated with Stmt 193 194end Sem_Dim; 195