RES - Automatically Resizing Contiguous Memory for OCaml
What is RES?
This OCaml-library consists of a set of modules which implement automatically resizing (= reallocating) datastructures that consume a contiguous part of memory. This allows appending and removing of elements to/from arrays (both boxed and unboxed), strings (buffers), bit strings and weak arrays while still maintaining fast constant-time access to elements.
There are also functors that allow the generation of similar modules which use different reallocation strategies.
Fast constant-time access to indexed elements (e.g. in arrays and strings) is often a prerequisite for short execution times of programs.
Still, operations like adding and/or removing elements to/from the end of such datastructures are often needed. Unfortunately, having both properties at the same time sometimes requires reallocating this contiguous part of memory.
This module does not eliminate this problem, but hides the process of reallocation from the user, i.e. it happens automatically.
Thus, the user is liberated from this bug-attracting (e.g. index errors) task.
This library allows the user to parameterize allocation strategies at runtime. This is an important feature, because it is impossible for any allocation algorithm to perform optimally without having knowledge about the user program.
For example, the programmer might know that a consecutive series of operations will alternately add and remove large batches of elements. In such a case it would be wise to keep a high reserve of available slots in the datastructure, because otherwise it will resize very often during this procedure which requires a significant amount of time.
By raising a corresponding threshold in appropriate places at runtime, programmers can fine-tune the behaviour of e.g. their buffers for optimal performance and set this parameter back later to save memory.
Because optimal reallocation strategies may be quite complex, it was also a design goal to have users supply their own ones (if required).
By using functors users can parameterize these datastructures with their own reallocation strategies, giving them even more control over how and when reallocations are triggered.
Users may want to add support for additional low-level implementations that require reallocations. In this case, too, it is fairly easy to create new modules by using functors.
The library implements a large interface of functions, all of which are completely independent of the reallocation strategy and the low-level implementation.
All the interfaces of the corresponding low-level implementations of datastructures (e.g. array, string) are fully supported and have been extended with further functionality. There is even a new buffer module which can be used in every context of the standard one.
OCaml makes a distinction between unboxed and boxed arrays. If the type of an array is
float, the representation will be unboxed in cases in which the array is not used in a polymorphic context (native code only).
To benefit from these much faster representations there are specialized versions of automatically resizing arrays in the distribution.
The API is fully documented and can be built as HTML using
It is also available online.
The preparameterized modules (default strategy) and the functors for mapping
strategy-implementations to this kind of modules are contained and documented
For examples of how to use the functors to implement new strategies and/or
low-level representations, take a look at the implementation in
Their function interface, however, is documented in files
(for parameterized "low-level" types like e.g. normal arrays) and in
lib/nopres_intf.ml (for non-parameterized "low-level" types like e.g. float
arrays, strings (buffers), etc.).
It should be noted that it is possible to use the standard notation for
accessing elements (e.g.
ar.(42)) with resizable arrays (and even with
Bits, etc...). This requires a short explanation of how OCaml
treats such syntactic sugar:
All that OCaml does is that it replaces such syntax with an appropriate
Array.set. This may be any module that happens to be
bound to this name in the current scope. The same principle is true for the
String-module and the
Thus, the following works:
module Array = Res.Bits module String = Res.Buffer let () = let ar = Array.empty () in Array.add_one ar true; print_endline (string_of_bool ar.(0)); let str = String.empty () in String.add_one str 'x'; print_char str.; print_newline ()
Do not forget that it is even possible to bind modules locally. Example:
let () = let module Array = Res.Array in Printf.printf "%d\n" (Array.init 10 (fun x -> x * x)).(7)
If you want to change one of your files to make use of resizable arrays instead of standard ones without much trouble, please read the following:
You may want to "save" the standard
Array-module and its type for later
module StdArray = Array type 'a std_array = 'a array
Make the resizable implementation (includes the index operators!) available:
Or more explicitly:
module Array = Res.Array
Or if you want to use a specific
module Array = Res.Bits
Then set the type:
type 'a array = 'a Array.t
If you create standard arrays with the builtin syntax, change lines like:
let ar = [| 1; 2; 3; 4 |] in
let ar = Array.of_array [| 1; 2; 3; 4 |] in
This should allow all of your sources to compile out-of-the-box with the additional functionality. In places where you still need the standard implementation you should have no problems to use the rebound module and type to do so.
This trick works similarly for the old and the new Buffer-module. You might
also want to replace the
String-module in this fashion. The latter one,
however, supports a number of functions like e.g.
escape, which are not
Contact Information and Contributing
In the case of bugs, feature requests, contributions and similar, you can contact me here: email@example.com
Up-to-date information should be available at: https://bitbucket.org/mmottl/res
Markus Mottl in Rutherford, NJ on July 10, 2012