Nim/doc/nimrodc.txt

===================================
   Nimrod Compiler User Guide
===================================

:Author: Andreas Rumpf
:Version: |nimrodversion|

.. contents::

Introduction
============

This document describes the usage of the *Nimrod compiler*
on the different supported platforms. It is not a definition of the Nimrod
programming system (therefore is the Nimrod manual).

Nimrod is free software; it is licensed under the
`GNU General Public License <gpl.html>`_.


Compiler Usage
==============

Command line switches
---------------------
Basis command line switches are:

.. include:: ../data/basicopt.txt

Advanced command line switches are:

.. include:: ../data/advopt.txt


Configuration file
------------------
The default configuration file is ``nimrod.cfg``. The ``nimrod`` executable
looks for it in the following directories (in this order):

1. ``/home/$user/.config/nimrod.cfg`` (UNIX) or ``$APPDATA/nimrod.cfg`` (Windows)
2. ``$nimrod/config/nimrod.cfg`` (UNIX, Windows)
3. ``/etc/nimrod.cfg`` (UNIX)

The search stops as soon as a configuration file has been found. The reading 
of ``nimrod.cfg`` can be suppressed by the ``--skip_cfg`` command line option.
Configuration settings can be overwritten in a project specific
configuration file that is read automatically. This specific file has to
be in the same directory as the project and be of the same name, except
that its extension should be ``.cfg``.

Command line settings have priority over configuration file settings.


Nimrod's directory structure
----------------------------
The generated files that Nimrod produces all go into a subdirectory called
``nimcache`` in your project directory. This makes it easy to delete all 
generated files.

However, the generated C code is not platform independant. C code generated for
Linux does not compile on Windows, for instance. The comment on top of the
C file lists the OS, CPU and CC the file has been compiled for.

The library lies in ``lib``. Directly in the library directory are essential
Nimrod modules like the ``system`` and ``os`` modules. Under ``lib/base``
are additional specialized libraries or interfaces to foreign libraries which
are included in the standard distribution. The ``lib/extra`` directory is
initially empty. Third party libraries should go there. In the default
configuration the compiler always searches for libraries in ``lib``,
``lib/base`` and ``lib/extra``.


Additional Features
===================

This section describes Nimrod's additional features that are not listed in the
Nimrod manual.

New Pragmas and Options
-----------------------

Because Nimrod generates C code it needs some "red tape" to work properly.
Thus lots of options and pragmas for tweaking the generated C code are
available.

Importc Pragma
~~~~~~~~~~~~~~
The `importc`:idx: pragma provides a means to import a type, a variable, or a
procedure from C. The optional argument is a string containing the C 
identifier. If the argument is missing, the C name is the Nimrod 
identifier *exactly as spelled*:

.. code-block::
  proc printf(formatstr: cstring) {.importc: "printf", varargs.}


Exportc Pragma
~~~~~~~~~~~~~~
The `exportc`:idx: pragma provides a means to export a type, a variable, or a
procedure to C. The optional argument is a string containing the C 
identifier. If the argument is missing, the C name is the Nimrod 
identifier *exactly as spelled*:

.. code-block:: Nimrod
  proc callme(formatstr: cstring) {.exportc: "callMe", varargs.}


Dynlib Pragma
~~~~~~~~~~~~~
With the `dynlib`:idx: pragma a procedure or a variable can be imported from
a dynamic library (``.dll`` files for Windows, ``lib*.so`` files for UNIX). The
non-optional argument has to be the name of the dynamic library: 

.. code-block:: Nimrod
  proc gtk_image_new(): PGtkWidget {.cdecl, dynlib: "libgtk-x11-2.0.so", importc.}

In general, importing a dynamic library does not require any special linker
options or linking with import libraries. This also
implies that no *devel* packages need to be installed.


No_decl Pragma
~~~~~~~~~~~~~~
The `no_decl`:idx: pragma can be applied to almost any symbol (variable, proc,
type, etc.) and is sometimes useful for interoperability with C:
It tells Nimrod that it should not generate a declaration for the symbol in
the C code. Thus it makes the following possible, for example:

.. code-block:: Nimrod
  var
    EOF {.importc: "EOF", no_decl.}: cint # pretend EOF was a variable, as
                                          # Nimrod does not know its value

However, the ``header`` pragma is often the better alternative.


Header Pragma
~~~~~~~~~~~~~
The `header`:idx: pragma is very similar to the ``no_decl`` pragma: It can be
applied to almost any symbol and specifies that not only it should not be
declared but also that it leads to the inclusion of a given header file:

.. code-block:: Nimrod
  type
    PFile {.importc: "FILE*", header: "<stdio.h>".} = pointer
      # import C's FILE* type; Nimrod will treat it as a new pointer type

The ``header`` pragma expects always a string constant. The string contant
contains the header file: As usual for C, a system header file is enclosed
in angle brackets: ``<>``. If no angle brackets are given, Nimrod
encloses the header file in ``""`` in the generated C code.


Varargs Pragma
~~~~~~~~~~~~~~
The `varargs`:idx: pragma can be applied to procedures only. It tells Nimrod
that the proc can take a variable number of parameters after the last
specified parameter. Nimrod string values will be converted to C
strings automatically:

.. code-block:: Nimrod
  proc printf(formatstr: cstring) {.nodecl, varargs.}

  printf("hallo %s", "world") # "world" will be passed as C string


No_static Pragma
~~~~~~~~~~~~~~~~
The `no_static`:idx: pragma can be applied to almost any symbol and specifies
that it shall not be declared ``static`` in the generated C code. Note that
symbols in the interface part of a module never get declared ``static``, so
only in very special cases this pragma is necessary.


Line_dir Option
~~~~~~~~~~~~~~~
The `line_dir`:idx: option can be turned on or off. If on the generated C code
contains ``#line`` directives.


Stack_trace Option
~~~~~~~~~~~~~~~~~~
If the `stack_trace`:idx: option is turned on, the generated C contains code to
ensure that proper stack traces are given if the program crashes or an
uncaught exception is raised.


Line_trace Option
~~~~~~~~~~~~~~~~~
The `line_trace`:idx: option implies the ``stack_trace`` option. If turned on,
the generated C contains code to ensure that proper stack traces with line
number information are given if the program crashes or an uncaught exception
is raised.

Debugger Option
~~~~~~~~~~~~~~~
The `debugger`:idx: option enables or disables the *Embedded Nimrod Debugger*.
See the documentation of endb_ for further information.


Breakpoint Pragma
~~~~~~~~~~~~~~~~~
The *breakpoint* pragma was specially added for the sake of debugging with
ENDB. See the documentation of `endb <endb.html>`_ for further information.


Volatile Pragma
~~~~~~~~~~~~~~~
The `volatile`:idx: pragma is for variables only. It declares the variable as
``volatile``, whatever that means in C/C++.

Register Pragma
~~~~~~~~~~~~~~~
The `register`:idx: pragma is for variables only. It declares the variable as
``register``, giving the compiler a hint that the variable should be placed
in a hardware register for faster access. C compilers usually ignore this
though and for good reason: Often they do a better job without it anyway.

In highly specific cases (a dispatch loop of an bytecode interpreter for 
example) it may provide benefits, though.


Acyclic Pragma
~~~~~~~~~~~~~~
The `acyclic`:idx: pragma can be used for object types to mark them as acyclic
even though they seem to be cyclic. This is an **optimization** for the garbage 
collector to not consider objects of this type as part of a cycle::

  type
    PNode = ref TNode
    TNode {.acyclic, final.} = object
      left, right: PNode
      data: string
      
In the example a tree structure is declared with the ``TNode`` type. Note that
the type definition is recursive thus the GC has to assume that objects of 
this type may form a cyclic graph. The ``acyclic`` pragma passes the 
information that this cannot happen to the GC. If the programmer uses the
``acyclic`` pragma for data types that are in reality cyclic, the GC may leak
memory, but nothing worse happens.



Disabling certain messages
--------------------------
Nimrod generates some warnings and hints ("line too long") that may annoy the
user. Thus a mechanism for disabling certain messages is provided: Each hint
and warning message contains a symbol in brackets. This is the message's
identifier that can be used to enable or disable it:

.. code-block:: Nimrod
  {.warning[LineTooLong]: off.} # turn off warning about too long lines

This is often better than disabling all warnings at once.


Debugging with Nimrod
=====================

Nimrod comes with its own *Embedded Nimrod Debugger*. See
the documentation of endb_ for further information.


Optimizing for Nimrod
=====================

Nimrod has no separate optimizer, but the C code that is produced is very
efficient. Most C compilers have excellent optimizers, so usually it is
not needed to optimize one's code. Nimrod has been designed to encourage
efficient code: The most readable code in Nimrod is often the most efficient
too.

However, sometimes one has to optimize. Do it in the following order:

1. switch off the embedded debugger (it is **slow**!)
2. turn on the optimizer and turn off runtime checks
3. profile your code to find where the bottlenecks are
4. try to find a better algorithm
5. do low-level optimizations

This section can only help you with the last item. Note that rewriting parts
of your program in C is *never* necessary to speed up your program, because
everything that can be done in C can be done in Nimrod. Rewriting parts in
assembler *might*.

Optimizing string handling
--------------------------

String assignments are sometimes expensive in Nimrod: They are required to
copy the whole string. However, the compiler is often smart enough to not copy
strings. Due to the argument passing semantics, strings are never copied when
passed to subroutines. The compiler does not copy strings that are returned by
a routine, because a routine returns a new string anyway. Thus it is efficient
to do:

.. code-block:: Nimrod
  var s = procA() # assignment will not copy the string; procA allocates a new
                  # string anyway

However it is not efficient to do:

.. code-block:: Nimrod
  var s = varA    # assignment has to copy the whole string into a new buffer!

String case statements are optimized too. A hashing scheme is used for them
if several different string constants are used. This is likely to be more
efficient than any hand-coded scheme.


The ECMAScript code generator
=============================

Note: As of version 0.7.0 the ECMAScript code generator is not maintained any
longer. Help if you are interested.

Note: I use the term `ECMAScript`:idx: here instead of `JavaScript`:idx:, since
it is the proper term.

The ECMAScript code generator is experimental!

Nimrod targets ECMAScript 1.5 which is supported by any widely used browser.
Since ECMAScript does not have a portable means to include another module,
Nimrod just generates a long ``.js`` file.

Features or modules that the ECMAScript platform does not support are not
available. This includes:

* manual memory management (``alloc``, etc.)
* casting and other unsafe operations (``cast`` operator, ``zeroMem``, etc.)
* file management (``openfile``, etc.)
* most modules of the Standard library
* proper 64 bit integer arithmetic
* proper unsigned integer arithmetic

However, the modules `strutils`:idx:, `math`:idx:, and `times`:idx: are
available! To access the DOM, use the `dom`:idx: module that is only available
for the ECMAScript platform.