Difference between revisions of "SourcePawn Transitional Syntax"

From AlliedModders Wiki
Jump to: navigation, search
m
Line 416: Line 416:
 
};
 
};
  
void DoStuff(Rectangle& r) {
+
void DoStuff(Rectangle r) {
 
   PrintToServer("%d, %d, %d, %d", r.x, r.y, r.width, r.height);
 
   PrintToServer("%d, %d, %d, %d", r.x, r.y, r.width, r.height);
 
}
 
}
 
</pawn>
 
</pawn>
  
Enum structs are syntactic sugar and are internally represented as arrays. They should never be used as arrays, however, unless interacting with opaque data structures like <tt>ArrayList</tt>. For example, this is considered valid:
+
Enum structs are syntactic sugar and are internally represented as arrays. This means they pass by-reference to function arguments, and the "&" token is not required (nor is it allowed).
 +
 
 +
Note that even though enum structs are actually arrays, for the most part they cannot be used as arrays. The exception is when interacting with opaque data structures like <tt>ArrayList</tt>. For example, this is considered valid:
  
 
<pawn>
 
<pawn>
void SaveRectangle(ArrayList list, Rectangle r) {
+
void SaveRectangle(ArrayList list, const Rectangle r) {
 
   list.PushArray(r, sizeof(r));
 
   list.PushArray(r, sizeof(r));
 
}
 
}

Revision as of 01:43, 17 December 2018

We would like to give our users a more modern language. Pawn is showing its age; manual memory management, buffers, tags, and lack of object-oriented API are very frustrating. We can't solve everything all at once, but we can begin to take steps in the right direction.

SourceMod 1.7 introduces a Transitional API. It is built on a new Transitional Syntax in SourcePawn, which is a set of language tools to make Pawn feel more modern. In particular, it allows developers to use older APIs in an object-oriented manner, without breaking compatibility. Someday, if and when SourcePawn can become a full-fledged modern language, the transitional API will making porting efforts very minimal.

The transitional API has the following major features and changes:

  • New Declarators - A cleaner way of declaring variables, similar to Java and C#.
  • Methodmaps - Object-oriented wrappers around older APIs.
  • Real Types - SourcePawn now has "int", "float", "bool", "void", and "char" as real types.
  • null - A new, general keyword to replace INVALID_HANDLE.

New Declarators

Developers familiar with pawn will recognize Pawn's original declaration style:

new Float:x = 5.0;
new y = 7;

In the transitional syntax, this can be reworded as:

float x = 5.0;
int y = 7;

The following builtin tags now have types:

  • Float: is float.
  • bool: is bool.
  • _: (or no tag) is int.
  • String: is char.
  • void can now be used as a return type for functions.

Rationale

In the old style, tagged variables are not real types. Float:x does not indicate a float-typed variable, it indicates a 32-bit "cell" tagged as a float. It is possible to remove the tag or change the tag, which while flexible, is dangerous and confusing. The syntax itself is also problematic. The parser does not have the ability to identify characters in between the tag name and the colon. Internally, the compiler cannot represent values that are not exactly 32-bit.

The takeaway message is: there is no sensible way to represent a concept like "int64" or "X is a type that represents an array of floats". The tagging grammar makes it too awkward, and the compiler itself is incapable of attaching such information to a tag. We can't fix that right away, but we can begin to deprecate the tag system via a more normal declaration syntax.

An easy example to see why this is necessary, is the weirdness in expressing something like this with tags:

native float[3] GetEntOrigin();

Note on the String tag: The introduction of char might initially seem confusing. The reason for this is that in the future, we would like to introduce a true "string" type that acts as an object, like strings in most other languages. The existing behavior in Pawn is an array of characters, which is much lower-level. The renaming clarifies what the type really is, and leaves the door open for better types in the future.

Arrays

The new style of declaration disambiguates between two kinds of arrays. Pawn has indeterminate arrays, where the size is not known, and determinate arrays, where the size is known. We refer to these as "dynamic" and "fixed-length" arrays, respectively.

A fixed-length array is declared by placing brackets after a variable name. For example:

int CachedStuff[1000];
int PlayerData[MAXPLAYERS + 1] = { 0, ... };
int Weapons[] = { WEAPON_AK47, WEAPON_GLOCK, WEAPON_KNIFE };

In these examples, the array size is fixed. The size is known ahead of time and cannot change. When using brackets in this position, the array size must be specified, either via an explicit size or inferred from an initial value.

A dynamic-length array has the brackets before the variable name, that is, after the type. The most common case is when specifying functions that take an array as input:

native void SetPlayerName(int player, const char[] name);

Here, we are specifying that the length of name is not always known - it could be anything.

Dynamic arrays can also be created in local scopes. For example,

void FindPlayers()
{
  int[] players = new int[MaxClients + 1];
}

This allocates a new array of the given size and places a reference in players. The memory is automatically freed when no longer in use.

It is illegal to initialize a fixed-length array with an indeterminate array, and it is illegal to initialize a dynamic array with a fixed-array. It is also illegal to specify a fixed size on a dynamic length array.

For the most part, this does not change existing Pawn semantics. It is simply new syntax intended to clarify the way arrays work.

Rationale

In the original syntax, there was a subtle difference in array declaration: new array1[MAXPLAYERS + 1]; new array2[MaxClients + 1];

Here, array1 and array2 are very different types: the first is an int[65] and the latter is an int[]. However, there is no syntactic difference: the compiler has to deduce whether the size expression is constant to determine the type. The new syntax clearly and explicitly disambiguates these cases, so in the future when we introduce fully dynamic and flexible arrays, we're less likely to break existing code.

This may result in some confusion. Hopefully won't matter long term, as once we have true dynamic arrays, fixed arrays will become much less useful and more obscure.

The rationale for restricting initializers is similar. Once we have true dynamic arrays, the restrictions will be lifted. In the meantime, we need to make sure we're limited to semantics that won't have subtle differences in the future.

Examples

float x = 5.0;    // Replacement for "new Float:x = 5.0;"
int y = 4;        // Replacement for "new y = 4;"
char name[32];    // Replacement for "new String:name[32];" and "decl String:name[32];"
 
void DoStuff(float x, int y, char[] name, int length) { //Replacement for "DoStuff(Float:x, y, String:name[], length)"
  Format(name, length, "%f %d", x, y); //No replacement here
}

View As

A new operator is available for reinterpreting the bits in a value as another type. This operator is called view_as. It is not a safe cast, in that, it can transform one type to another even if the actual value does not conform to either.

In pre-transitional syntax, this was called "retagging". Retagging does not support new-style types, which is why this operator has been introduced. Example of before and after code:

// Before:
float x = Float:array.Get(i);
 
// After:
float y = view_as<float>(array.Get(i));

It is worth reiterating that this is not a cast. If the value in the array is not a float, then when "viewed as" a float it will probably look very odd.

Grammar

The new and old declaration grammar is below.

return-type ::= return-old | return-new
return-new ::= type-expr new-dims?        // Note, dims not yet supported.
return-old ::= old-dims? label?

argdecl ::= arg-old | arg-new
arg-new ::= "const"? type-expr '&'? symbol old-dims? ('=' arg-init)?
arg-old ::= "const"? tags? '&'? symbol old-dims? ('=' arg-init)?

vardecl ::= var-old | var-new
var-new ::= var-new-prefix type-expr symbol old-dims?
var-new-prefix ::= "static" | "const"
var-old ::= var-old-prefix tag? symbol old-dims?
var-old-prefix ::= "new" | "decl" | "static" | "const"

global ::= global-old | global-new
global-new ::= storage-class* type-expr symbol old-dims?
global-old ::= storage-class* tag? symbol old-dims?

storage-class ::= "public" | "static" | "const" | "stock"

type-expr ::= (builtin-type | symbol) new-dims?
builtin-type ::= "void"
               | "int"
               | "float"
               | "char"
               | "bool"

tags ::= tag-vector | tag
tag-vector ::= '{' symbol (',' symbol)* '}' ':'
tag ::= label

new-dims ::= ('[' ']')*
old-dims ::= ('[' expr? ']')+

label ::= symbol ':'
symbol ::= [A-Za-z_]([A-Za-z0-9_]*)

Also note, there is no equivalent of decl in the new declarator syntax. decl is considered to be dangerous and unnecessary. If an array's zero initialization is too costly, consider making it static or global.

Methodmaps

Introduction

Methodmaps are simple: they attach methods onto an enum. For example, here is our legacy API for Handles:

native CloneHandle(Handle handle);
native CloseHandle(Handle handle);

This is a good example of our legacy API. Using it generally looks something like:

Handle array = CreateAdtArray();
PushArrayCell(array, 4);
CloseHandle(array);

Gross! A Methodmap can clean it up by attaching functions to the Handle tag, like this:

methodmap Handle {
    public Clone() = CloneHandle;
    public Close() = CloseHandle;
};

Now, our earlier array code can start to look object-oriented:

Handle array = CreateAdtArray();
PushArrayCell(array, 4);
array.Close();

With a full methodmap for Arrays, for example,

methodmap ArrayList < Handle
{
  public native ArrayList(); // constructor
  public native void Push(any value);
};

We can write even more object-like code:

ArrayList array = new ArrayList();
array.Push(4);
delete array;

(Note: the official API does not expose methods on raw Handles.)

Inheritance

The Handle system has a "weak" hierarchy. All handles can be passed to CloseHandle, but only AdtArray handles can be passed to functions like PushArrayCell. This hierarchy is not enforced via tags (unfortunately), but instead by run-time checks. Methodmaps allow us to make individual handle types object-oriented, while also moving type-checks into the compiler.

For example, here is a transitional API for arrays:

native AdtArray CreateAdtArray();
 
methodmap AdtArray < Handle {
    public PushCell() = PushArrayCell;
};

Note that CreateAdtArray now returns AdtArray instead of Handle. Normally that would break older code, but since AdtArray inherits from Handle, there is a special rule in the type system that allows coercing an AdtArray to a Handle (but not vice versa).

Now, the API looks much more object-oriented:

AdtArray array = CreateAdtArray();
array.PushCell(4);
array.Close();

Inline Methods

Methodmaps can declare inline methods and accessors. Inline methods can be either natives or Pawn functions. For example:

methodmap AdtArray {
    public native void PushCell(value);
};

This example requires that an "AdtArray.PushCell" native exists somewhere in SourceMod. It has a magic initial parameter called "this", so the signature will look something like:

native void AdtArray.PushCell(AdtArray this, value);

(Of course, this exact signature will not appear in an include file - it's the signature that the C++ implementation should expect, however.)

It's also possible to define new functions without a native:

methodmap AdtArray {
    public native void PushCell(value);
 
    public void PushCells(list[], count) {
        for (int i = 0; i < count; i++) {
            this.PushCell(i);
        }
    }
};

Lastly, we can also define accessors. or example,

methodmap AdtArray {
    property int Size {
        public get() = GetArraySize;
    }
    property bool Empty {
        public get() {
            return this.Size == 0;
        }
    }
    property int Capacity {
        public native get();
    }
};

The first accessor simply assigns an existing function as an accessor for "Size". The second accessor is an inline method with an implicit "this" parameter. The third accessor will bind to a native with the following name and signature:

native int AdtArray.Capacity.get(AdtArray this);

Setters are also supported. For example:

methodmap Player {
    property int Health {
        public native get();
        public native set(int health);
    }
}

Custom Tags

Methodmaps don't have to be used with Handles. It is possible to define custom methodmaps on new or existing tags. For example:

methodmap AdminId {
    public int Rights() {
        return GetAdminFlags(this);
    }
};

Now, for example, it is possible to do:

GetPlayerAdmin(id).Rights()

Constructors and Destructors

Methodmaps can also define constructors and destructors, which is useful if they are intended to behave like actual objects. For example,

methodmap AdtArray {
    public AdtArray(blocksize = 1);
    public ~AdtArray();
    public void PushCell(value);
};

Now AdtArrays can be used in a fully object-oriented style:

AdtArray array = new AdtArray();
array.PushCell(10);
array.PushCell(20);
array.PushCell(30);
delete array;

Caveats

There are a few caveats to methodmaps:

  1. CloseHandle() is not yet gone. It is required to call delete on any object that previously would have required CloseHandle().
  2. There can be only one methodmap for a tag.
  3. When using existing natives, the first parameter of the native must coerce to the tag of the methodmap. Tag mismatches of the "this" parameter will result in an error. Not a warning!
  4. Methodmaps can only be defined on tags. Pawn has some ways of creating actual types (like via struct or class). Methodmaps cannot be created on those types.
  5. Methodmaps do not have strong typing. For example, it is still possible to perform "illegal" casts like Float:CreateAdtArray(). This is necessary for backwards compatibility, so methodmap values can flow into natives like PrintToServer or CreateTimer.
  6. It is not possible to inherit from anything other than another previously declared methodmap.
  7. Methodmaps can only be defined over scalars - that is, the "this" parameter can never be an array. This means they cannot be used for enum-structs.
  8. Destructors can only be native. When we are able to achieve garbage collection, destructors will be removed.
  9. The signatures of methodmaps must use the new declaration syntax.
  10. Methodmaps must be declared before they are used.

Grammar

The grammar for methodmaps is:

visibility ::= "public"
method-args ::= arg-new* "..."?

methodmap ::= "methodmap" symbol? { methodmap-item* } term
methodmap-item ::=
           visibility "~"? symbol "(" ")" "=" symbol term
         | visibility "native" type-expr "~"? symbol methodmap-symbol "(" method-args ")" term
         | visibility type-expr symbol "(" method-args ")" func-body term
         | "property" type-expr symbol { property-decl } term
property-func ::= "get" | "set"
property-decl ::= visibility property-impl
property-impl ::=
           "native" property-func "(" ")" term
         | property-func "(" ")" func-body term
         | property-func "(" ")" "=" symbol

Typedefs

Function tags and function enums have been deprecated in favor of a more modern syntax. Currently, they can still only create tag names for functions. Future versions will support arbitrary types.

Upgrading both functags and funcenums is simple. Below are two examples:

functag public Action:SrvCmd(args);
 
funcenum Timer {
  Action:public(Handle:Timer, Handle:hndl),
  Action:public(Handle:timer),
};

Now, this becomes:

typedef SrvCmd = function Action (int args);
 
typeset Timer {
  function Action (Handle timer, Handle hndl);
  function Action (Handle timer);
};

The grammar for the new syntax is:

typedef ::= "typedef" symbol "=" full-type-expr term
full-type-expr ::= "(" type-expr ")"
                 | type-expr
type-expr ::= "function" type-name "(" typedef-args? ")"
typedef-args ::= "..."
               | typedef-arg (", " "...")?

Note that typedefs only support new-style types.

Enum Structs

Enum structs were a previously unsupported mechanism for emulating structs through arrays. As of SourceMod 1.10, this mechanism is now fully supported through Transitional Syntax. Here is an example of enum struct syntax:

enum struct Rectangle {
  int x;
  int y;
  int width;
  int height;
 
  int Area() {
    return this.width * this.height;
  }
};
 
void DoStuff(Rectangle r) {
  PrintToServer("%d, %d, %d, %d", r.x, r.y, r.width, r.height);
}

Enum structs are syntactic sugar and are internally represented as arrays. This means they pass by-reference to function arguments, and the "&" token is not required (nor is it allowed).

Note that even though enum structs are actually arrays, for the most part they cannot be used as arrays. The exception is when interacting with opaque data structures like ArrayList. For example, this is considered valid:

void SaveRectangle(ArrayList list, const Rectangle r) {
  list.PushArray(r, sizeof(r));
}
 
void PopArray(ArrayList list, Rectangle r) {
  list.GetArray(list.Length - 1, r, sizeof(r));
  list.Erase(list.Length - 1);
}

But this is not allowed:

Rectangle r;
PrintToServer("%d", r[0]);

The grammar for enum structs is as follows:

enum-struct ::= "enum" "struct" symbol "{" newline enum-struct-entry enum-struct-entry* "}" term
enum-struct-entry ::= enum-struct-field
                    | enum-struct-method
enum-struct-field ::= type-expr symbol old-dims? term
enum-struct-method ::= type-expr symbol "(" method-args ")" func-body term

Enforcing new syntax

You can enforce the new syntax in 1.7 by using

#pragma newdecls required

ontop of your code, after the includes (or else current 1.7 includes which contain old syntax will be read with new-syntax rules).