• Stars
    star
    113
  • Rank 310,115 (Top 7 %)
  • Language
    Fortran
  • License
    GNU General Publi...
  • Created almost 8 years ago
  • Updated 4 months ago

Reviews

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Repository Details

The Fortran Template Library

The Fortran Template Library

The Fortran Template Library (FTL) is a general purpose library for Fortran 2003. Its intention is to bring all these nice things we take for granted in modern languages like Python and C++ to the Fortran world: Generic containers, versatile algorithms, easy string manipulation, and more. It is heavily inspired by C++'s standard library, especially the part that is commonly referred to as the Standard Template Library (STL).

Introduction

Fortran has come pretty far with the 2003 and 2008 standards, yet one thing that is still missing are generic programming facilities, meaning a way to write code that works with any type. In C++ this is done with templates. Let's have a look at how one would define a generic max() function.

template <typename T>
T max(T x, T y) {
   T value;
   if (x < y)
      value = y;
   else
      value = x;
   return value;
}

Once this function template is defined one can just call it like any other function.

bigger_integer = max(1,6)
bigger_double  = max(1.2,6.4)

The compiler will determine that the type T is int for the first call and double for the second and will generate two versions of the max function template: One for int and double. This is called template instantiation. This does not only work for the plain old types like numbers, but for any type, provided that it overloads the < operator. If it doesn't, the compiler can not instantiate the template and will produce an error.

In Python we can do pretty much the same thing, only that we would get a run-time error in the max() function if it is called with a type for which the < operator is not overloaded.

Now Fortran obviously does not have something like C++'s templates or Python's duck typing. But really what we need to do is not that complicated: We only need to replace T with our desired type!

T function max(x, y)
   T :: x, y
   if (x < y) then
      max = y
   else
      max = x
   endif
end function

If we put this into a separate file, say max.F90_template, we can easily instantiate our templates manually using the C preprocessor.

#define T integer
#include "max.F90_template"

#define T real
#include "max.F90_template"

Sure, it's not as convenient as in C++ where the compiler instantiates the templates automatically, but it's better than nothing. Now for something as simple as a max() function this is probably not worth it. But if your template is a dictionary class with two template types (key and value) you really don't want to duplicate its implementation for all combinations of key and value types!

In practice we would need to put our template function into a module and and wrap in an interface so that the different instantiations don't collide with each other, but these are technical details that are all hidden inside the template file. From the outside we don't mind; all we have to do is instantiate our template once.

Components

ftlDynArray
A resizeable array container. It can slowly grow in size as elements are added at its end. Note that insertion at the end is an amortized constant time operation. Basically, ftlDynArray is exactly the same as C++'s std::vector. We just changed the name because calling a resizeable array a vector makes no sense from a mathematical point of view.
ftlList
A linked list container that allows constant time insert and erase operations anywhere within the list. Exactly the same as C++'s std::list.
ftlHashMap
An associative containers that stores elements formed by the combination of a key value and a mapped value, and which allows for fast retrieval of individual elements based on their keys. It's basically a dictionary that internally uses a hash table to allow constant time retrieval of elements. ftlHashMap is very similar to C++'s std::unordered_map (though its interface is a bit less awkward).
ftlHashSet
A container representing a set of unique elements. It's pretty much like an ftlHashMap where the key is at the same time also the value.
ftlHash
A small utility library that provides hash functions for the Fortran intrinsic types. This allows them to be used as key types in ftlHashMap and as elements in ftlHashSet. Furthermore these basic has functions can be used to implement hash functions for other derived types, so that these can also be used as keys in ftlHashMap. This file is not a template.
ftlString
A variable length string type that integrates seamlessly with plain Fortran strings. The provided ftlString type is not a template. It is quite similar to C++'s std::string in the sense that it has the interface of a container of single characters. However, since the std::string interface is in practice a bit basic, it also offers Python's string manipulation methods.
ftlRegex
A convenient Fortran wrapper around the POSIX regular expression functionality in the C standard library (aka regex.h) or alternatively the PCRE (Perl Compatible Regular Expressions) library. It's nicely integrated with the ftlString type. ftlRegex is not a template.
ftlAlgorithms
A library of generic algorithms that work on all FTL containers. Exactly the same as C++'s std::algorithm header.
ftlArray
A little module that provides the FTL style container iterators for plain one-dimensional Fortran arrays. This allows the ftlAlgorithms to work on normal Fortran arrays.
ftlSharedPtr
Provides a reference counted ftlSharedPtr in the spirit of C++'s std::shared_ptr.

Implementation progress

ftlDynArray, ftlList, ftlHashMap, ftlHashSet and the plain Fortran array wrapper ftlArray are pretty much finished.

ftlAlgorithms is incomplete. Ultimately we would like all of the algorithms in C++'s std::algorithm header to be implemented, but so far we only did maybe 30% of them. It's quite a lot of work as there are many algorithms to implement. We would absolutely appreciate some help here.

ftlString is incomplete. The basics are there, but we would like to have all Python string manipulation methods. Only a handful are implemented at the moment. Again, there are just many of them. Help is much appreciated.

Definitely on the TODO list are:

  • An equivalent of std::deque, a double-ended queue. A container with random access iterators but constant time insertion at both ends. It should be reasonably local in memory.

These things might be nice:

  • Random number generators and distributions like in std::random.
  • File system access like std::filesystem.
  • Parsing and evaluating equations from strings.

Documentation

The documentation of the Fortran Template Library is hosted on the Wiki associated with this GitHub repository.

License

The Fortran Template Library is an Open Source project supported by Software for Chemistry and Materials BV (SCM). The terms of the LGPL-3.0 license apply. As an exception to the LGPL-3.0 license, you agree to grant SCM a BSD 3-clause license to the contributions you commit to this Github repository or provide to SCM in another manner.