Merge branch 'master' into cmake

Conflicts:
	CMakeLists.txt

Monicelli.lpp

@@ -22,7 +22,6 @@
 #include "Parser.hpp"
 
 #include <string>
-#include <cstdlib>
 
 static void meta(const char *);
 
@@ -30,6 +29,12 @@ using namespace monicelli;
 typedef Parser::token token;
 
 #define YY_USER_ACTION location->begin.columns(yyleng);
+
+static inline
+bool in(const char *sub, const std::string &str) {
+    return str.find(sub) != std::string::npos;
+}
+
 %}
 
 %option ecs stack warn c++
@@ -88,7 +93,7 @@ CHAR  [a-zA-Z_]
 ("il"|"lo"|"la"|"l'"|"i"|"gli"|"le"|"un"|"un'"|"una"|"dei"|"delle") {
     return token::ARTICLE;
 }
-"più" {
+"pi"("ù"|"u`") {
     return token::OP_PLUS;
 }
 "meno" {
@@ -142,7 +147,7 @@ CHAR  [a-zA-Z_]
 "e "("b"|"p")"rematura anche, se" {
     return token::LOOP_CONDITION;
 }
-"che cos'è" {
+"che cos'"("è"|"e`") {
     return token::BRANCH_CONDITION;
 }
 "?" {
@@ -151,14 +156,14 @@ CHAR  [a-zA-Z_]
 "o tarapia tapioco" {
     return token::BRANCH_ELSE;
 }
-"e velocità di esecuzione" {
+"e velocit"("à"|"a`")" di esecuzione" {
     return token::BRANCH_END;
 }
 ":" {
     return token::COLON;
 }
-"blinda la supercazzola" {
+"blinda la supercazzo"("r"|"l")"a" {
-    return token::FUNDECL;
+    return token::FUN_DECL;
 }
 "con" {
     return token::PARAMS;
@@ -166,8 +171,8 @@ CHAR  [a-zA-Z_]
 "," {
     return token::COMMA;
 }
-("b"|"p")"rematurata la supercazzola" {
+("b"|"p")"rematurata la supercazzo"("r"|"l")"a" {
-    return token::FUNCALL;
+    return token::FUN_CALL;
 }
 "o scherziamo"("?")? {
     return token::FUN_END;
@@ -191,13 +196,20 @@ CHAR  [a-zA-Z_]
     return token::ID;
 } 
 
-{DIGIT}+ {
+[-+]?(({DIGIT}*".")?{DIGIT}+|{DIGIT}+".")([eE][-+]?{DIGIT}+)? {
-    lval->intval = strtol(yytext, NULL, 10);
+    std::string value(yytext);
-    return token::NUMBER;
+
+    if (in(".", value) || in("e", value) || in("E", value)) {
+        lval->floatval = std::stod(value);
+        return token::FLOAT;
+    } else {
+        lval->intval = std::stol(value);
+        return token::NUMBER;
+    }
 }
 
 <INITIAL,shift>. {
-    return -1;
+    return token::ERROR;
 }
 
 %%

Monicelli.ypp

@@ -1,6 +1,6 @@
 /*
  * Monicelli: an esoteric language compiler
- * 
+ *
  * Copyright (C) 2014 Stefano Sanfilippo
  *
  * This program is free software: you can redistribute it and/or modify
@@ -50,7 +50,7 @@
     static int yylex(Parser::semantic_type*, Parser::location_type*, Scanner&);
 }
 
-%token MAIN
+%token MAIN ERROR
 %token RETURN
 %token ARTICLE TYPENAME STAR
 %token VARDECL ASSIGN
@@ -59,7 +59,7 @@
 %token LOOP_BEGIN LOOP_CONDITION
 %token BRANCH_CONDITION BRANCH_BEGIN BRANCH_ELSE BRANCH_END CASE_END
 %token COLON COMMA
-%token FUNDECL PARAMS FUNCALL FUN_END
+%token FUN_DECL PARAMS FUN_CALL FUN_END
 %token ABORT
 %token ID NUMBER FLOAT
 
@@ -71,15 +71,6 @@
 %nonassoc LOWER_THAN_ELSE
 %nonassoc BRANCH_ELSE
 
-%code {
-#include <stack>
-
-static std::stack<StatementList*> stmtStack;
-static std::stack<ExpressionList*> argsStack;
-static std::stack<IdList*> paramsStack;
-static std::stack<BranchCaseList*> branchCaseStack;
-}
-
 %union {
     int intval;
     double floatval;
@@ -94,39 +85,48 @@ static std::stack<BranchCaseList*> branchCaseStack;
     Input* inputval;
     Abort* abortval;
     Branch* branchval;
+    Branch::Body* branchbodyval;
     VarDeclaration* declval;
     Assignment* assignval;
     Loop* loopval;
     BranchCase *caseval;
+    BranchCaseList *caselistval;
     Return* returnval;
     Expression* expressionval;
+    ExpressionList* exprlistval;
     SemiExpression *semiexpval;
     Id* idval;
     Number* numericval;
     Function* funval;
+    FunArg *argval;
+    FunArgList *arglistval;
     Main* mainval;
 }
 
 %type<intval> NUMBER
 %type<floatval> FLOAT
 %type<strval> ID
-%type<typeval> TYPENAME
+%type<typeval> TYPENAME fun_return
-
 %type<statementval> statement
-%type<statlistval> branch_body
+%type<statlistval> statements
+%type<branchbodyval> branch_body
 %type<assertval> assert_stmt
 %type<callval> fun_call
+%type<argval> arg_decl
+%type<arglistval> args_decl args
 %type<funval> fun_decl
 %type<printval> print_stmt
 %type<inputval> input_stmt
 %type<abortval> abort_stmt
 %type<branchval> branch_stmt
 %type<caseval> case_stmt
+%type<caselistval> cases
 %type<declval> var_decl
 %type<assignval> assign_stmt
 %type<loopval> loop_stmt
 %type<returnval> return_stmt
 %type<expressionval> expression maybe_expression simple_expression var_init
+%type<exprlistval> call_arglist call_args
 %type<semiexpval> semi_expression
 %type<idval> variable
 %type<numericval> numeric
@@ -151,47 +151,45 @@ fun_decls:
     fun_decls
 ;
 fun_decl:
-    FUNDECL ID {
+    FUN_DECL fun_return ID args FUN_END statements {
-        paramsStack.push(new IdList());
+        $$ = new Function(new Id($3), $2, $4, $6);
-    } args FUN_END {
-        stmtStack.push(new StatementList());
-    }
-    statements {
-        $$ = new Function(new Id($2), paramsStack.top(), stmtStack.top());
-        paramsStack.pop();
-        stmtStack.pop();
     }
 ;
+fun_return:
+    /* epsilon */ { $$ = Type::VOID; } | TYPENAME { $$ = $1; }
+;
 args:
-    /* epsilon */ | PARAMS arglist
+    /* epsilon */ { $$ = new FunArgList(); }| PARAMS args_decl { $$ = $2; }
 ;
-arglist:
+args_decl:
-    variable {
+    arg_decl {
-        paramsStack.top()->push_back($1); 
+        $$ = new FunArgList();
+        $$->push_back($1);
     }
-    | variable {
+    | args_decl COMMA arg_decl {
-        paramsStack.top()->push_back($1); 
+        $1->push_back($3);
+    }
+;
+arg_decl:
+    variable pointer TYPENAME {
+        $$ = new FunArg($1, $3, $2);
     }
-    arglist
 ;
 main:
-    MAIN {
+    MAIN statements {
-        stmtStack.push(new StatementList());
+        $$ = new Main($2);
-    }
-    statements {
-        $$ = new Main(stmtStack.top());
-        stmtStack.pop();
     }
 ;
 statements:
     /* epsilon */ {
+        $$ = new StatementList();
     }
-    | statement {
+    | statements statement  {
-        if ($1 != nullptr) {
+        if ($2 != nullptr) {
-            stmtStack.top()->push_back($1);
+            $1->push_back($2);
         }
+        $$ = $1;
     }
-    statements
 ;
 statement:
     assert_stmt   { $$ = $1; }
@@ -252,61 +250,59 @@ maybe_expression:
     expression { $$ = $1; } | /* epsilon */ { $$ = nullptr; }
 ;
 loop_stmt:
-    LOOP_BEGIN {
+    LOOP_BEGIN statements LOOP_CONDITION expression {
-        stmtStack.push(new StatementList());
+        $$ = new Loop($2, $4);
-    }
-    statements LOOP_CONDITION expression {
-        $$ = new Loop(stmtStack.top(), $5);
-        stmtStack.pop();
     }
 ;
 branch_stmt:
-    BRANCH_CONDITION variable BRANCH_BEGIN {
+    BRANCH_CONDITION variable BRANCH_BEGIN branch_body BRANCH_END {
-        branchCaseStack.push(new BranchCaseList());
+        $$ = new Branch($2, $4);
-    }
-    branch_body BRANCH_END {
-        $$ = new Branch($2, branchCaseStack.top(), $5);
-        branchCaseStack.pop();
     }
 ;
 branch_body:
     cases %prec LOWER_THAN_ELSE {
-        $$ = nullptr;
+        $$ = new Branch::Body($1);
     }
-    | cases BRANCH_ELSE COLON {
+    | cases BRANCH_ELSE COLON statements {
-        stmtStack.push(new StatementList());
+        $$ = new Branch::Body($1, $4);
-    }
-    statements {
-        $$ = stmtStack.top();
-        stmtStack.pop();
     }
 ;
 cases:
-    case_stmt | case_stmt CASE_END cases
+    case_stmt {
+        $$ = new BranchCaseList();
+        $$->push_back($1);
+    }
+    | cases CASE_END case_stmt {
+        $1->push_back($3);
+        $$ = $1;
+    }
 ;
 case_stmt:
-    semi_expression COLON {
+    semi_expression COLON statements {
-        stmtStack.push(new StatementList());
+        $$ = new BranchCase($1, $3);
-    } statements {
-        branchCaseStack.top()->push_back(new BranchCase($1, stmtStack.top()));
-        stmtStack.pop();
     }
 ;
 fun_call:
-    FUNCALL {
+    FUN_CALL ID call_args FUN_END {
-        argsStack.push(new ExpressionList());
+        $$ = new FunctionCall(new Id($2), $3);
-    }
-    ID call_args FUN_END {
-        $$ = new FunctionCall(new Id($3), argsStack.top());
-        argsStack.pop();
     }
 ;
 call_args:
-    /* epsilon */ | PARAMS call_arglist
+    /* epsilon */ {
+        $$ = new ExpressionList();
+    }
+    | PARAMS call_arglist {
+        $$ = $2;
+    }
 ;
 call_arglist:
-    expression { argsStack.top()->push_back($1); } 
+    expression {
-    | expression { argsStack.top()->push_back($1); } COMMA call_arglist
+        $$ = new ExpressionList();
+        $$->push_back($1);
+    }
+    | call_arglist COMMA expression {
+        $$->push_back($3);
+    }
 ;
 abort_stmt:
     ABORT {
@@ -400,6 +396,7 @@ simple_expression:
 void Parser::error(const location_type& loc, const std::string &message) {
     std::cerr << "line " << loc.begin.line << ", col " << loc.begin.column;
     std::cerr << ": " << message << std::endl;
+    std::exit(1);
 }
 
 int yylex(Parser::semantic_type *lval, Parser::location_type *loc, Scanner &scanner) {

Nodes.cpp

@@ -40,6 +40,9 @@ std::ostream& monicelli::operator<<(std::ostream &stream, const Type &type) {
         case Type::DOUBLE:
             stream << "double";
             break;
+        case Type::VOID:
+            stream << "void";
+            break;
     }
 
     return stream;
@@ -119,9 +122,9 @@ void Branch::emit(std::ostream &stream, int indent) {
     stream << "if (";
     var->emit(stream);
 
-    if (cases->size() > 0) {
+    if (body->cases->size() > 0) {
-        BranchCase *last = cases->back();
+        BranchCase *last = body->cases->back();
-        for (BranchCase *c: *cases) {
+        for (BranchCase *c: *body->cases) {
             c->emit(stream, indent + 1);
             if (c != last) {
                 stream << " else if (";
@@ -130,12 +133,12 @@ void Branch::emit(std::ostream &stream, int indent) {
         }
     }
 
-    if (els == nullptr) {
+    if (body->els == nullptr) {
         return;
     }
 
     stream << " else {\n";
-    els->emit(stream, indent + 1);
+    body->els->emit(stream, indent + 1);
     emitIndent(stream, indent);
     stream << "}";
 }
@@ -158,12 +161,16 @@ void Assignment::emit(std::ostream &stream, int indent) {
 }
 
 void Print::emit(std::ostream &stream, int indent) {
-    stream << "std::cout << ";
+    stream << "std::cout << (";
     expression->emit(stream);
-    stream << " << std::endl";
+    stream << ") << std::endl";
 }
 
 void Input::emit(std::ostream &stream, int indent) {
+    stream << "std::cout << \"";
+    variable->emit(stream);
+    stream << "? \";\n";
+    emitIndent(stream, indent);
     stream << "std::cin >> ";
     variable->emit(stream);
 }
@@ -185,16 +192,26 @@ void FunctionCall::emit(std::ostream &stream, int indent) {
     stream << ")";
 }
 
+void FunArg::emit(std::ostream &stream, int indent) {
+    stream << type << (pointer? "* ": " ");
+    name->emit(stream);
+}
+
 void Function::emit(std::ostream &stream, int indent) {
+    emitSignature(stream, indent);
+    stream << " {\n";
+    body->emit(stream, indent + 1);
+    stream << "}\n\n";
+}
+
+void Function::emitSignature(std::ostream &stream, int indent) {
     emitIndent(stream, indent);
 
-    stream << "void ";
+    stream << type << ' ';
     name->emit(stream);
     stream << "(";
     args->emit(stream);
-    stream << ") {\n";
+    stream << ")";
-    body->emit(stream, indent + 1);
-    stream << "}\n\n";
 }
 
 void Main::emit(std::ostream &stream, int indent) {
@@ -209,6 +226,15 @@ void Program::emit(std::ostream &stream, int indent) {
     stream << "#include <iostream>\n";
     stream << "#include <cstdlib>\n\n";
 
+    for (Function *f: functions) {
+        f->emitSignature(stream);
+        stream << ";\n";
+    }
+
+    if (!functions.empty()) {
+        stream << "\n";
+    }
+
     for (Function *f: functions) {
         f->emit(stream);
     }

Nodes.hpp

@@ -31,14 +31,18 @@ enum class Type {
     CHAR,
     FLOAT,
     BOOL,
-    DOUBLE
+    DOUBLE,
+    VOID
 };
 
 std::ostream& operator<<(std::ostream &stream, const Type &type);
 
 
 template <class T>
-using Pointer = std::unique_ptr<T>;
+class Pointer: public std::unique_ptr<T> {
+public:
+    Pointer(T *p = nullptr): std::unique_ptr<T>(p) {}
+};
 
 
 class Emittable {
@@ -264,14 +268,23 @@ typedef PointerList<BranchCase> BranchCaseList;
 
 class Branch: public Statement {
 public:
-    Branch(Id *v, BranchCaseList *c, StatementList *e):
+    struct Body {
-        var(v), cases(c), els(e) {}
+    public:
+        Body(BranchCaseList *c, StatementList *e = nullptr): cases(c), els(e) {}
+
+    private:
+        Pointer<BranchCaseList> cases;
+        Pointer<StatementList> els;
+
+        friend class Branch;
+    };
+
+    Branch(Id *v, Branch::Body *b): var(v), body(b) {}
     virtual void emit(std::ostream &stream, int indent = 0);
 
 private:
     Pointer<Id> var;
-    Pointer<BranchCaseList> cases;
+    Pointer<Branch::Body> body;
-    Pointer<StatementList> els;
 };
 
 
@@ -285,20 +298,40 @@ private:
 };
 
 
-class Function: public Emittable {
+class FunArg: public Emittable {
 public:
-    Function(Id *n, IdList *a, StatementList *b):
+    FunArg(Id *n, Type t, bool p): name(n), type(t), pointer(p) {}
-        name(n), args(a), body(b) {}
+    virtual ~FunArg() {}
-    virtual ~Function() {}
 
     virtual void emit(std::ostream &stream, int indent = 0);
 
 private:
     Pointer<Id> name;
-    Pointer<IdList> args;
+    Type type;
+    bool pointer;
+};
+
+
+typedef ListEmittable<FunArg> FunArgList;
+
+
+class Function: public Emittable {
+public:
+    Function(Id *n, Type r, FunArgList *a, StatementList *b):
+        name(n), type(r), args(a), body(b) {}
+    virtual ~Function() {}
+
+    virtual void emit(std::ostream &stream, int indent = 0);
+    void emitSignature(std::ostream &stream, int indent = 0);
+
+private:
+    Pointer<Id> name;
+    Type type;
+    Pointer<FunArgList> args;
     Pointer<StatementList> body;
 };
 
+
 class Program: public Emittable {
 public:
     virtual void emit(std::ostream &stream, int indent = 0);

README.md

@@ -34,7 +34,7 @@ Usage
 subset of C++. For those of you who want to get to the code ASAP, the `examples/`
 folder contains a set of programs covering most of the features of the language.
 
-A good wat to learn on the field is comparing the resulting C++ with the
+A good way to learn on the field is comparing the resulting C++ with the
 input. Well, mostly with the beautified version of the input, `*.beauty.mc`.
 
 The compiler reads from standard input and print result to standard output.
@@ -46,10 +46,14 @@ The compiler reads from standard input and print result to standard output.
 Language overview
 =================
 
-The original specification can be found in `Specification.txt`, which
+The original specification can be found in `Specification.txt`, and was
-unfortunately is not complete. This project is an ongoing effort to implement
+initially conceived by my colleagues and dear friends Alessandro Barenghi,
-it, which means filling gaps and ambiguities. This file only documents
+Michele Tartara and Nicola Vitucci, to whom goes my gratitude.
-usable features of the language.
+
+Unfortunately, their proposal was meant to be a joke and is not complete.
+This project is an ongoing effort to produce a rigorous specification for the
+language and implement a compiler, which implies filling gaps and ambiguities
+with sensible choices.
 
 Statements have no terminator, i.e. no semicolon `;` or the like. A single
 statement can be split across multiple lines and multiple statements can be
@@ -58,6 +62,10 @@ words **cannot** be split on multiple lines.
 
 A comma might be inserted after each statement, if it fits the sentence ;)
 
+Accented letters can be replaced by the non-accented letter followed by a
+backtick `` ` ``, although the use of the correct Italian spelling is strongly
+encouraged for maximizing the antani effect.
+
 Main
 ----
 
@@ -194,7 +202,7 @@ defined as follows:
 For example:
 
     voglio antani, Necchi come se fosse 10
-    stuzzica:
+    stuzzica
         antani come fosse antani meno 1
     e brematura anche, se antani maggiore di 0
 
@@ -285,10 +293,62 @@ where the `o tarapia tapioco` part is like the `default` block.
 Functions
 ---------
 
-_Coming in a few releases..._. 
+**Note**: the alternate spelling `supercazzora` might be used in place
+ of `supercazzola` wherever the latter appears.
 
-We can already parse and emit them, although there are some ambiguities 
+###Declaration
-yet to be solved in the specification.
+
+A function is declared with the `blinda la supercazzola` statement:
+
+    blinda la supercazzola [<type>] <name> [con <param> <type>[, <param> <type>...]] o scherziamo?
+        <statements>
+
+Where `<type>` can be omitted for a void function. For instance:
+
+    blinda la supercazzola Necchi antanizzata con alfio Mascetti o scherziamo?
+        vaffanzum alfio meno 2!
+
+is a function of type `Necchi`, taking one argument of type `Mascetti`.
+Multiple arguments must be comma-separed, like in:
+
+    blinda la supercazzola Necchi antanizzata con alfio Mascetti, barilotto Necchi o scherziamo?
+        vaffanzum alfio meno 2!
+
+which is a function of type `Necchi`, taking two arguments of type `Mascetti`
+and `Necchi`. It maps to:
+
+    int antanizzata(char alfio, int barilotto) {
+        return alfio - 2;
+    }
+
+Finally, this:
+
+    blinda la supercazzola antanizzata o scherziamo?
+        vaffanzum!
+
+is a `void` function taking no arguments and becomes:
+
+    void antanizzata() {
+        return;
+    }
+
+Functions cannot be nested and can be declared before or after the main in any
+order. `mcc` will not check that a return statement is always reachable inside
+ a non-void function. Failing to return a value leads to undefined behaviour.
+
+###Invocation
+
+A function is called with the `brematurata la supercazzola` statement:
+
+    brematurata la supercazzola <name> [con <expression>[, <expression>...] o scherziamo?
+
+Functions might be called inside expressions. For instance, this:
+
+    antani come se fosse brematurata la supercazzola alfio con barilotto diviso 3 o scherziamo? per 2
+
+maps to:
+
+    antani = alfio(barilotto / 3) * 2;
 
 Exceptions
 ----------
@@ -311,9 +371,11 @@ Comments
 --------
 
 Any character after `bituma` is ignored until a line break is encountered. For
-instance:
+instance, in:
 
-    antani come se fossee 4 bituma lorem ipsum
+    antani come se fosse 4 bituma, scusi, noi siamo in quattro
+
+`, scusi, noi siamo in quattro` is ignored.
 
 Comments are useful to fill the "supercazzola" and make it more readable, since
 any word (including reserved words) can be inserted into it.

Specification.txt

@@ -1,3 +1,14 @@
+-------------------------------------------------------------------------
+
+NOTE: this document was left here only for historical purposes,
+      refer to REAMDE.md for the current spec.
+
+NOTA: questo documento è obsoleto ed è qui solo per memoria storica, in 
+      quanto rappresenta la prima specifica proposta per il linguaggio.
+      README.md contiene l'attuale versione.
+
+-------------------------------------------------------------------------
+
 Proposte per la definizione di un linguaggio di programmazione esoterico "Monicelli"
 
 Il "main" del programma inizia in corrispondenza di:

examples/float.mc

@@ -0,0 +1,12 @@
+# Test all possible numerical forms (integer and float)
+
+Lei ha clacsonato
+
+2.0 a posterdati
+-2.0 a posterdati
+2. a posterdati
+.232 a posterdati
+-2 a posterdati
+-.2233 a posterdati
++2 a posterdati
++2.233e-23 a posterdati

examples/hello-world.mc

@@ -0,0 +1,16 @@
+# Author: Alessandro Pellegrini <alessandro@pellegrini.tk>
+# Released under GPLv3
+
+Lei ha clacsonato
+voglio una bucaiola, Necchi come se fosse 0 voglio prematurata, Mascetti come se fosse 72
+prematurata a posterdati voglio antifurto, Mascetti come se fosse 87.
+voglio una cofandina, Mascetti come se fosse prematurata con scappellamento a sinistra per 1.
+cofandina come fosse cofandina meno 33 brematurata la supercazzola antanizzata con antifurto,
+cofandina o scherziamo? vaffanzum bucaiola! bituma scusi, noi siamo in quattro.
+blinda la supercazzola antanizzata con Alfio Mascetti, tarapia Mascetti o scherziamo?
+voglio vicesindaco, Mascetti come se fosse 101 vicesindaco a posterdati voglio pastene,
+Mascetti come se fosse vicesindaco più 7  pastene a posterdati bituma in un certo senso.
+pastene a posterdati tarapia a posterdati Alfio a posterdati tarapia a posterdati 
+voglio scappellamento, Mascetti come se fosse 114 scappellamento a posterdati
+pastene a posterdati voglio Antani, Mascetti come se fosse pastene meno 8 Antani a posterdati
+vaffanzum!

examples/syntax.beauty.mc

@@ -40,7 +40,7 @@ Lei ha clacsonato
     vicesindaco come se fosse brematurata la supercazzola avanti con il vicesindaco o scherziamo,
     vaffanzum 0!
 
-blinda la supercazzola antanizzata con alfio o scherziamo?
+blinda la supercazzola Necchi antanizzata con alfio Mascetti, barilotto Necchi o scherziamo?
     vaffanzum alfio meno 2!
 
 bituma al finale?

examples/syntax.mc

@@ -10,5 +10,5 @@ o tarapia tapioco: mi porga il cappello e velocità di esecuzione,
 vicesindaco a posterdati, mi porga il vicesindaco, brematurata la supercazzola 
 tombale con alfio, serio o scherziamo? avvertite don ulrico, ho visto la signora! 
 vicesindaco come se fosse brematurata la supercazzola avanti con il vicesindaco
-o scherziamo, vaffanzum 0! blinda la supercazzola antanizzata con alfio o scherziamo?
+o scherziamo, vaffanzum 0! blinda la supercazzola Necchi antanizzata con alfio 
-vaffanzum alfio meno 2! bituma al finale?
+Mascetti o scherziamo? vaffanzum alfio meno 2! bituma al finale?