Related
I've been dabbling in Haskell - so still very much a beginner.
I'm been thinking about the counting the frequency of items in a list. In languages with mutable data structures, this is typically solved using a hash table - a dict in Python or a HashMap in Java for example. The complexity of such a solution is O(n) - assuming the hash table can fit entirely in memory.
In Haskell, there seem to be two (mainstream) choices - to sort the data then group and count it or use a Data.Map. If a sort is used, it dominates the run-time of the solution, so the complexity is O(n log n). Likewise, Data.Map uses a balanced tree, so inserting n elements into it will also have complexity O(n log n).
If my analysis is correct, then I assume that this particular problem is most efficiently solved by resorting to a mutable data structure. Are there other types of problems where this is also true? How in general do people using Haskell approach something like this?
The question whether we can implement any algorithm with optimal complexity in a pure language is currently unknown. Nicholas Pippenger has proven that there is a problem that must necessarily have a log(n) penalty in a pure strict language compared to the optimal algorithm. However, there is a followup paper which shows that this problem have an optimal solution in a lazy language. So at the end of the day we really don't know. Though it seems that most people think that there is an inherent log(n) penalty for some problems, even for lazy languages.
A while ago, I ran across an article on FingerTrees (See Also an accompanying Stack Overflow Question) and filed the idea away. I have finally found a reason to make use of them.
My problem is that the Data.FingerTree package seems to have a little bit rot around the edges. Moreover, Data.Sequence in the Containers package which makes use of the data structure re-implements a (possibly better) version, but doesn't export it.
As theoretically useful as this structure seems to be, it doesn't seem to get a lot of actual use or attention. Have people found that FingerTrees are not useful as a practical matter, or is this a case not enough attention?
further explanation:
I'm interested in building a data structure holding text that has good concatenation properties. Think about building an HTML document from assorted fragments. Most pre-built solutions use bytestrings, but I really want something that deals with Unicode text properly. My plan at the moment is to layer Data.Text fragments into a FingerTree.
I would also like to borrow the trick from Data.Vector of taking slices without copying using (offset,length) manipulation. Data.Text.Text has this built in to the data type, but only uses it for efficient uncons and unsnoc opperations. In FingerTree this information could very easily becomes the v or annotation of the tree.
To answer your question about finger trees in particular, I think the problem is that they have relatively high constant costs compared to arrays, and are more complex than other ways of achieving efficient concatenation. A Builder has a more efficient interface for just appending chunks, and they're usually readily available (see the links in #informatikr's answer). Suppose that Data.Text.Lazy is implemented with a linked list of chunks, and you're creating a Data.Text.Lazy from a builder. Unless you have a lot of chunks (probably more than 50), or are accessing data near the end of the list repeatedly, the high constant cost of a finger tree probably isn't worth it.
The Data.Sequence implementation is specialized for performance reasons, and isn't as general as the full interface provided by the fingertree package. That's why it isn't exported; it's not really possible to use it for anything other than a Sequence.
I also suspect that many programmers are at a loss as to how to actually use the monoidal annotation, as it's behind a fairly significant abstraction barrier. So many people wouldn't use it because they don't see how it can be useful compared to other data types.
I didn't really get it until I read Chung-chieh Shan's blog series on word numbers (part2, part3, part4). That's proof that the idea can definitely be used in practical code.
In your case, if you need to both inspect partial results and have efficient appends, using a fingertree may be better than a builder. Depending on the builder's implementation, you may end up doing a lot of repeated work as you convert to Text, add more stuff to the builder, convert to Text again, etc. It would depend on your usage pattern though.
You might be interested in my splaytree package, which provides splay trees with monoidal annotations, and several different structures build upon them. Other than the splay tree itself, the Set and RangeSet modules have more-or-less complete API's, the Sequence module is mostly a skeleton I used for testing. It's not a "batteries included" solution to what you're looking for (again, #informatikr's answer provides those), but if you want to experiment with monoidal annotations it may be more useful than Data.FingerTree. Be aware that a splay tree can get unbalanced if you traverse all the elements in sequence (or continually snoc onto the end, or similar), but if appends and lookups are interleaved performance can be excellent.
In addition to John Lato's answer, I'll add some specific details about the performance of finger trees, since I spent some time looking at that in the past.
The broad summary is:
Data.Sequence has great constant factors and asymptotics: it is almost as fast as [] when accessing the front of the list (where both data structures have O(1) asymptotics), and much faster elsewhere in the list (where Data.Sequence's logarithmic asymptotics trounce []'s linear asymptotics).
Data.FingerTree has the same asymptotics as Data.Sequence, but is about an order of magnitude slower.
Just like lists, finger trees have high per-element memory overheads, so they should be combined with chunking for better memory and cache use. Indeed, a few packages do this (yi, trifecta, rope). If Data.FingerTree could be brought close to Data.Sequence in performance, I would hope to see a Data.Text.Sequence type, which implemented a finger tree of Data.Text values. Such a type would lose the streaming behaviour of Data.Text.Lazy, but benefit from improved random access and concatenation performance. (Similarly, I would want to see Data.ByteString.Sequence and Data.Vector.Sequence.)
The obstacle to implementing these now is that no efficient and generic implementation of finger trees exists (see below where I discuss this further). To produce efficient implementations of Data.Text.Sequence one would have to completely reimplement finger trees, specialised to Text - just as Data.Text.Lazy completely reimplements lists, specialised to Text. Unfortunately, finger trees are much more complex than lists (especially concatenation!), so this is a considerable amount of work.
So as I see it the answer is:
specialised finger trees are great, but a lot of work to implement
chunked finger trees (e.g. Data.Text.Sequence) would be great, but at present the poor performance of Data.FingerTree means they are not a viable alternative to chunked lists in the common case
builders and chunked lists achieve many of the benefits of chunked finger trees, and so they suffice for the common case
in the uncommon case where builders and chunked lists don't suffice, we grit our teeth and put up with the poor constant factors of chunked finger trees (e.g. in yi and trifecta).
Obstacles to an efficient and generic finger tree
Much of the performance gap between Data.Sequence and Data.FingerTree is due to two optimisations in Data.Sequence:
The measure type is specialised to Int, so measure manipulations will compile down to efficient integer arithmetic rather
The measure type is unpacked into the Deep constructor, which saves pointer dereferences in the inner loops of the tree operations.
It is possible to apply these optimisations in the general case of Data.FingerTree by using data families for generic unpacking and by exploiting GHC's inliner and specialiser - see my fingertree-unboxed package, which brings generic finger tree performance almost up to that of Data.Sequence. Unfortunately, these techniques have some significant problems:
data families for generic unpacking is unpleasant for the user, because they have to define lots of instances. There is no clear solution to this problem.
finger trees use polymorphic recursion, which GHC's specialiser doesn't handle well (1, 2). This means that, to get sufficient specialisation on the measure type, we need lots of INLINE pragmas, which causes GHC to generate huge amounts of code.
Due to these problems, I never released the package on Hackage.
Ignoring your Finger Tree question and only responding to your further explanation: did you look into Data.Text.Lazy.Builder or, specifically for building HTML, blaze-html?
Both allow fast concatenation. For slicing, if that is important for solving your problem, they might not have ideal performance.
It seems that Template Haskell is often viewed by the Haskell community as an unfortunate convenience. It's hard to put into words exactly what I have observed in this regard, but consider these few examples
Template Haskell listed under "The Ugly (but necessary)" in response to the question Which Haskell (GHC) extensions should users use/avoid?
Template Haskell considered a temporary/inferior solution in Unboxed Vectors of newtype'd values thread (libraries mailing list)
Yesod is often criticized for relying too much on Template Haskell (see the blog post in response to this sentiment)
I've seen various blog posts where people do pretty neat stuff with Template Haskell, enabling prettier syntax that simply wouldn't be possible in regular Haskell, as well as tremendous boilerplate reduction. So why is it that Template Haskell is looked down upon in this way? What makes it undesirable? Under what circumstances should Template Haskell be avoided, and why?
One reason for avoiding Template Haskell is that it as a whole isn't type-safe, at all, thus going against much of "the spirit of Haskell." Here are some examples of this:
You have no control over what kind of Haskell AST a piece of TH code will generate, beyond where it will appear; you can have a value of type Exp, but you don't know if it is an expression that represents a [Char] or a (a -> (forall b . b -> c)) or whatever. TH would be more reliable if one could express that a function may only generate expressions of a certain type, or only function declarations, or only data-constructor-matching patterns, etc.
You can generate expressions that don't compile. You generated an expression that references a free variable foo that doesn't exist? Tough luck, you'll only see that when actually using your code generator, and only under the circumstances that trigger the generation of that particular code. It is very difficult to unit test, too.
TH is also outright dangerous:
Code that runs at compile-time can do arbitrary IO, including launching missiles or stealing your credit card. You don't want to have to look through every cabal package you ever download in search for TH exploits.
TH can access "module-private" functions and definitions, completely breaking encapsulation in some cases.
Then there are some problems that make TH functions less fun to use as a library developer:
TH code isn't always composable. Let's say someone makes a generator for lenses, and more often than not, that generator will be structured in such a way that it can only be called directly by the "end-user," and not by other TH code, by for example taking a list of type constructors to generate lenses for as the parameter. It is tricky to generate that list in code, while the user only has to write generateLenses [''Foo, ''Bar].
Developers don't even know that TH code can be composed. Did you know that you can write forM_ [''Foo, ''Bar] generateLens? Q is just a monad, so you can use all of the usual functions on it. Some people don't know this, and because of that, they create multiple overloaded versions of essentially the same functions with the same functionality, and these functions lead to a certain bloat effect. Also, most people write their generators in the Q monad even when they don't have to, which is like writing bla :: IO Int; bla = return 3; you are giving a function more "environment" than it needs, and clients of the function are required to provide that environment as an effect of that.
Finally, there are some things that make TH functions less fun to use as an end-user:
Opacity. When a TH function has type Q Dec, it can generate absolutely anything at the top-level of a module, and you have absolutely no control over what will be generated.
Monolithism. You can't control how much a TH function generates unless the developer allows it; if you find a function that generates a database interface and a JSON serialization interface, you can't say "No, I only want the database interface, thanks; I'll roll my own JSON interface"
Run time. TH code takes a relatively long time to run. The code is interpreted anew every time a file is compiled, and often, a ton of packages are required by the running TH code, that have to be loaded. This slows down compile time considerably.
This is solely my own opinion.
It's ugly to use. $(fooBar ''Asdf) just does not look nice. Superficial, sure, but it contributes.
It's even uglier to write. Quoting works sometimes, but a lot of the time you have to do manual AST grafting and plumbing. The API is big and unwieldy, there's always a lot of cases you don't care about but still need to dispatch, and the cases you do care about tend to be present in multiple similar but not identical forms (data vs. newtype, record-style vs. normal constructors, and so on). It's boring and repetitive to write and complicated enough to not be mechanical. The reform proposal addresses some of this (making quotes more widely applicable).
The stage restriction is hell. Not being able to splice functions defined in the same module is the smaller part of it: the other consequence is that if you have a top-level splice, everything after it in the module will be out of scope to anything before it. Other languages with this property (C, C++) make it workable by allowing you to forward declare things, but Haskell doesn't. If you need cyclic references between spliced declarations or their dependencies and dependents, you're usually just screwed.
It's undisciplined. What I mean by this is that most of the time when you express an abstraction, there is some kind of principle or concept behind that abstraction. For many abstractions, the principle behind them can be expressed in their types. For type classes, you can often formulate laws which instances should obey and clients can assume. If you use GHC's new generics feature to abstract the form of an instance declaration over any datatype (within bounds), you get to say "for sum types, it works like this, for product types, it works like that". Template Haskell, on the other hand, is just macros. It's not abstraction at the level of ideas, but abstraction at the level of ASTs, which is better, but only modestly, than abstraction at the level of plain text.*
It ties you to GHC. In theory another compiler could implement it, but in practice I doubt this will ever happen. (This is in contrast to various type system extensions which, though they might only be implemented by GHC at the moment, I could easily imagine being adopted by other compilers down the road and eventually standardized.)
The API isn't stable. When new language features are added to GHC and the template-haskell package is updated to support them, this often involves backwards-incompatible changes to the TH datatypes. If you want your TH code to be compatible with more than just one version of GHC you need to be very careful and possibly use CPP.
There's a general principle that you should use the right tool for the job and the smallest one that will suffice, and in that analogy Template Haskell is something like this. If there's a way to do it that's not Template Haskell, it's generally preferable.
The advantage of Template Haskell is that you can do things with it that you couldn't do any other way, and it's a big one. Most of the time the things TH is used for could otherwise only be done if they were implemented directly as compiler features. TH is extremely beneficial to have both because it lets you do these things, and because it lets you prototype potential compiler extensions in a much more lightweight and reusable way (see the various lens packages, for example).
To summarize why I think there are negative feelings towards Template Haskell: It solves a lot of problems, but for any given problem that it solves, it feels like there should be a better, more elegant, disciplined solution better suited to solving that problem, one which doesn't solve the problem by automatically generating the boilerplate, but by removing the need to have the boilerplate.
* Though I often feel that CPP has a better power-to-weight ratio for those problems that it can solve.
EDIT 23-04-14: What I was frequently trying to get at in the above, and have only recently gotten at exactly, is that there's an important distinction between abstraction and deduplication. Proper abstraction often results in deduplication as a side effect, and duplication is often a telltale sign of inadequate abstraction, but that's not why it's valuable. Proper abstraction is what makes code correct, comprehensible, and maintainable. Deduplication only makes it shorter. Template Haskell, like macros in general, is a tool for deduplication.
I'd like to address a few of the points dflemstr brings up.
I don't find the fact that you can't typecheck TH to be that worrying. Why? Because even if there is an error, it will still be compile time. I'm not sure if this strengthens my argument, but this is similar in spirit to the errors that you receive when using templates in C++. I think these errors are more understandable than C++'s errors though, as you'll get a pretty printed version of the generated code.
If a TH expression / quasi-quoter does something that's so advanced that tricky corners can hide, then perhaps it's ill-advised?
I break this rule quite a bit with quasi-quoters I've been working on lately (using haskell-src-exts / meta) - https://github.com/mgsloan/quasi-extras/tree/master/examples . I know this introduces some bugs such as not being able to splice in the generalized list comprehensions. However, I think that there's a good chance that some of the ideas in http://hackage.haskell.org/trac/ghc/blog/Template%20Haskell%20Proposal will end up in the compiler. Until then, the libraries for parsing Haskell to TH trees are a nearly perfect approximation.
Regarding compilation speed / dependencies, we can use the "zeroth" package to inline the generated code. This is at least nice for the users of a given library, but we can't do much better for the case of editing the library. Can TH dependencies bloat generated binaries? I thought it left out everything that's not referenced by the compiled code.
The staging restriction / splitting of compilation steps of the Haskell module does suck.
RE Opacity: This is the same for any library function you call. You have no control over what Data.List.groupBy will do. You just have a reasonable "guarantee" / convention that the version numbers tell you something about the compatibility. It is somewhat of a different matter of change when.
This is where using zeroth pays off - you're already versioning the generated files - so you'll always know when the form of the generated code has changed. Looking at the diffs might be a bit gnarly, though, for large amounts of generated code, so that's one place where a better developer interface would be handy.
RE Monolithism: You can certainly post-process the results of a TH expression, using your own compile-time code. It wouldn't be very much code to filter on top-level declaration type / name. Heck, you could imagine writing a function that does this generically. For modifying / de-monolithisizing quasiquoters, you can pattern match on "QuasiQuoter" and extract out the transformations used, or make a new one in terms of the old.
This answer is in response to the issues brought up by illissius, point by point:
It's ugly to use. $(fooBar ''Asdf) just does not look nice. Superficial, sure, but it contributes.
I agree. I feel like $( ) was chosen to look like it was part of the language - using the familiar symbol pallet of Haskell. However, that's exactly what you /don't/ want in the symbols used for your macro splicing. They definitely blend in too much, and this cosmetic aspect is quite important. I like the look of {{ }} for splices, because they are quite visually distinct.
It's even uglier to write. Quoting works sometimes, but a lot of the time you have to do manual AST grafting and plumbing. The [API][1] is big and unwieldy, there's always a lot of cases you don't care about but still need to dispatch, and the cases you do care about tend to be present in multiple similar but not identical forms (data vs. newtype, record-style vs. normal constructors, and so on). It's boring and repetitive to write and complicated enough to not be mechanical. The [reform proposal][2] addresses some of this (making quotes more widely applicable).
I also agree with this, however, as some of the comments in "New Directions for TH" observe, the lack of good out-of-the-box AST quoting is not a critical flaw. In this WIP package, I seek to address these problems in library form: https://github.com/mgsloan/quasi-extras . So far I allow splicing in a few more places than usual and can pattern match on ASTs.
The stage restriction is hell. Not being able to splice functions defined in the same module is the smaller part of it: the other consequence is that if you have a top-level splice, everything after it in the module will be out of scope to anything before it. Other languages with this property (C, C++) make it workable by allowing you to forward declare things, but Haskell doesn't. If you need cyclic references between spliced declarations or their dependencies and dependents, you're usually just screwed.
I've run into the issue of cyclic TH definitions being impossible before... It's quite annoying. There is a solution, but it's ugly - wrap the things involved in the cyclic dependency in a TH expression that combines all of the generated declarations. One of these declarations generators could just be a quasi-quoter that accepts Haskell code.
It's unprincipled. What I mean by this is that most of the time when you express an abstraction, there is some kind of principle or concept behind that abstraction. For many abstractions, the principle behind them can be expressed in their types. When you define a type class, you can often formulate laws which instances should obey and clients can assume. If you use GHC's [new generics feature][3] to abstract the form of an instance declaration over any datatype (within bounds), you get to say "for sum types, it works like this, for product types, it works like that". But Template Haskell is just dumb macros. It's not abstraction at the level of ideas, but abstraction at the level of ASTs, which is better, but only modestly, than abstraction at the level of plain text.
It's only unprincipled if you do unprincipled things with it. The only difference is that with the compiler implemented mechanisms for abstraction, you have more confidence that the abstraction isn't leaky. Perhaps democratizing language design does sound a bit scary! Creators of TH libraries need to document well and clearly define the meaning and results of the tools they provide. A good example of principled TH is the derive package: http://hackage.haskell.org/package/derive - it uses a DSL such that the example of many of the derivations /specifies/ the actual derivation.
It ties you to GHC. In theory another compiler could implement it, but in practice I doubt this will ever happen. (This is in contrast to various type system extensions which, though they might only be implemented by GHC at the moment, I could easily imagine being adopted by other compilers down the road and eventually standardized.)
That's a pretty good point - the TH API is pretty big and clunky. Re-implementing it seems like it could be tough. However, there are only really only a few ways to slice the problem of representing Haskell ASTs. I imagine that copying the TH ADTs, and writing a converter to the internal AST representation would get you a good deal of the way there. This would be equivalent to the (not insignificant) effort of creating haskell-src-meta. It could also be simply re-implemented by pretty printing the TH AST and using the compiler's internal parser.
While I could be wrong, I don't see TH as being that complicated of a compiler extension, from an implementation perspective. This is actually one of the benefits of "keeping it simple" and not having the fundamental layer be some theoretically appealing, statically verifiable templating system.
The API isn't stable. When new language features are added to GHC and the template-haskell package is updated to support them, this often involves backwards-incompatible changes to the TH datatypes. If you want your TH code to be compatible with more than just one version of GHC you need to be very careful and possibly use CPP.
This is also a good point, but somewhat dramaticized. While there have been API additions lately, they haven't been extensively breakage inducing. Also, I think that with the superior AST quoting I mentioned earlier, the API that actually needs to be used can be very substantially reduced. If no construction / matching needs distinct functions, and are instead expressed as literals, then most of the API disappears. Moreover, the code you write would port more easily to AST representations for languages similar to Haskell.
In summary, I think that TH is a powerful, semi-neglected tool. Less hate could lead to a more lively eco-system of libraries, encouraging the implementation of more language feature prototypes. It's been observed that TH is an overpowered tool, that can let you /do/ almost anything. Anarchy! Well, it's my opinion that this power can allow you to overcome most of its limitations, and construct systems capable of quite principled meta-programming approaches. It's worth the usage of ugly hacks to simulate the "proper" implementation, as this way the design of the "proper" implementation will gradually become clear.
In my personal ideal version of nirvana, much of the language would actually move out of the compiler, into libraries of these variety. The fact that the features are implemented as libraries does not heavily influence their ability to faithfully abstract.
What's the typical Haskell answer to boilerplate code? Abstraction. What're our favorite abstractions? Functions and typeclasses!
Typeclasses let us define a set of methods, that can then be used in all manner of functions generic on that class. However, other than this, the only way classes help avoid boilerplate is by offering "default definitions". Now here is an example of an unprincipled feature!
Minimal binding sets are not declarable / compiler checkable. This could lead to inadvertent definitions that yield bottom due to mutual recursion.
Despite the great convenience and power this would yield, you cannot specify superclass defaults, due to orphan instances http://lukepalmer.wordpress.com/2009/01/25/a-world-without-orphans/ These would let us fix the numeric hierarchy gracefully!
Going after TH-like capabilities for method defaults led to http://www.haskell.org/haskellwiki/GHC.Generics . While this is cool stuff, my only experience debugging code using these generics was nigh-impossible, due to the size of the type induced for and ADT as complicated as an AST. https://github.com/mgsloan/th-extra/commit/d7784d95d396eb3abdb409a24360beb03731c88c
In other words, this went after the features provided by TH, but it had to lift an entire domain of the language, the construction language, into a type system representation. While I can see it working well for your common problem, for complex ones, it seems prone to yielding a pile of symbols far more terrifying than TH hackery.
TH gives you value-level compile-time computation of the output code, whereas generics forces you to lift the pattern matching / recursion part of the code into the type system. While this does restrict the user in a few fairly useful ways, I don't think the complexity is worth it.
I think that the rejection of TH and lisp-like metaprogramming led to the preference towards things like method-defaults instead of more flexible, macro-expansion like declarations of instances. The discipline of avoiding things that could lead to unforseen results is wise, however, we should not ignore that Haskell's capable type system allows for more reliable metaprogramming than in many other environments (by checking the generated code).
One rather pragmatic problem with Template Haskell is that it only works when GHC's bytecode interpreter is available, which is not the case on all architectures. So if your program uses Template Haskell or relies on libraries that use it, it will not run on machines with an ARM, MIPS, S390 or PowerPC CPU.
This is relevant in practice: git-annex is a tool written in Haskell that makes sense to run on machines worrying about storage, such machines often have non-i386-CPUs. Personally, I run git-annex on a NSLU 2 (32 MB of RAM, 266MHz CPU; did you know Haskell works fine on such hardware?) If it would use Template Haskell, this is not possible.
(The situation about GHC on ARM is improving these days a lot and I think 7.4.2 even works, but the point still stands).
Why is TH bad? For me, it comes down to this:
If you need to produce so much repetitive code that you find yourself trying to use TH to auto-generate it, you're doing it wrong!
Think about it. Half the appeal of Haskell is that its high-level design allows you to avoid huge amounts of useless boilerplate code that you have to write in other languages. If you need compile-time code generation, you're basically saying that either your language or your application design has failed you. And we programmers don't like to fail.
Sometimes, of course, it's necessary. But sometimes you can avoid needing TH by just being a bit more clever with your designs.
(The other thing is that TH is quite low-level. There's no grand high-level design; a lot of GHC's internal implementation details are exposed. And that makes the API prone to change...)
I know that memoization seems to be a perennial topic here on the haskell tag on stack overflow, but I think this question has not been asked before.
I'm aware of several different 'off the shelf' memoization libraries for Haskell:
The memo-combinators and memotrie packages, which make use of a beautiful trick involving lazy infinite data structures to achieve memoization in a purely functional way. (As I understand it, the former is slightly more flexible, while the latter is easier to use in simple cases: see this SO answer for discussion.)
The uglymemo package, which uses unsafePerformIO internally but still presents a referentially transparent interface. The use of unsafePerformIO internally results in better performance than the previous two packages. (Off the shelf, its implementation uses comparison-based search data structures, rather than perhaps-slightly-more-efficient hash functions; but I think that if you find and replace Cmp for Hashable and Data.Map for Data.HashMap and add the appropraite imports, you get a hash based version.)
However, I'm not aware of any library that looks answers up based on object identity rather than object value. This can be important, because sometimes the kinds of object which are being used as keys to your memo table (that is, as input to the function being memoized) can be large---so large that fully examining the object to determine whether you've seen it before is itself a slow operation. Slow, and also unnecessary, if you will be applying the memoized function again and again to an object which is stored at a given 'location in memory' 1. (This might happen, for example, if we're memoizing a function which is being called recursively over some large data structure with a lot of structural sharing.) If we've already computed our memoized function on that exact object before, we can already know the answer, even without looking at the object itself!
Implementing such a memoization library involves several subtle issues and doing it properly requires several special pieces of support from the language. Luckily, GHC provides all the special features that we need, and there is a paper by Peyton-Jones, Marlow and Elliott which basically worries about most of these issues for you, explaining how to build a solid implementation. They don't provide all details, but they get close.
The one detail which I can see which one probably ought to worry about, but which they don't worry about, is thread safety---their code is apparently not threadsafe at all.
My question is: does anyone know of a packaged library which does the kind of memoization discussed in the Peyton-Jones, Marlow and Elliott paper, filling in all the details (and preferably filling in proper thread-safety as well)?
Failing that, I guess I will have to code it up myself: does anyone have any ideas of other subtleties (beyond thread safety and the ones discussed in the paper) which the implementer of such a library would do well to bear in mind?
UPDATE
Following #luqui's suggestion below, here's a little more data on the exact problem I face. Let's suppose there's a type:
data Node = Node [Node] [Annotation]
This type can be used to represent a simple kind of rooted DAG in memory, where Nodes are DAG Nodes, the root is just a distinguished Node, and each node is annotated with some Annotations whose internal structure, I think, need not concern us (but if it matters, just ask and I'll be more specific.) If used in this way, note that there may well be significant structural sharing between Nodes in memory---there may be exponentially more paths which lead from the root to a node than there are nodes themselves. I am given a data structure of this form, from an external library with which I must interface; I cannot change the data type.
I have a function
myTransform : Node -> Node
the details of which need not concern us (or at least I think so; but again I can be more specific if needed). It maps nodes to nodes, examining the annotations of the node it is given, and the annotations its immediate children, to come up with a new Node with the same children but possibly different annotations. I wish to write a function
recursiveTransform : Node -> Node
whose output 'looks the same' as the data structure as you would get by doing:
recursiveTransform Node originalChildren annotations =
myTransform Node recursivelyTransformedChildren annotations
where
recursivelyTransformedChildren = map recursiveTransform originalChildren
except that it uses structural sharing in the obvious way so that it doesn't return an exponential data structure, but rather one on the order of the same size as its input.
I appreciate that this would all be easier if say, the Nodes were numbered before I got them, or I could otherwise change the definition of a Node. I can't (easily) do either of these things.
I am also interested in the general question of the existence of a library implementing the functionality I mention quite independently of the particular concrete problem I face right now: I feel like I've had to work around this kind of issue on a few occasions, and it would be nice to slay the dragon once and for all. The fact that SPJ et al felt that it was worth adding not one but three features to GHC to support the existence of libraries of this form suggests that the feature is genuinely useful and can't be worked around in all cases. (BUT I'd still also be very interested in hearing about workarounds which will help in this particular case too: the long term problem is not as urgent as the problem I face right now :-) )
1 Technically, I don't quite mean location in memory, since the garbage collector sometimes moves objects around a bit---what I really mean is 'object identity'. But we can think of this as being roughly the same as our intuitive idea of location in memory.
If you only want to memoize based on object identity, and not equality, you can just use the existing laziness mechanisms built into the language.
For example, if you have a data structure like this
data Foo = Foo { ... }
expensive :: Foo -> Bar
then you can just add the value to be memoized as an extra field and let the laziness take care of the rest for you.
data Foo = Foo { ..., memo :: Bar }
To make it easier to use, add a smart constructor to tie the knot.
makeFoo ... = let foo = Foo { ..., memo = expensive foo } in foo
Though this is somewhat less elegant than using a library, and requires modification of the data type to really be useful, it's a very simple technique and all thread-safety issues are already taken care of for you.
It seems that stable-memo would be just what you needed (although I'm not sure if it can handle multiple threads):
Whereas most memo combinators memoize based on equality, stable-memo does it based on whether the exact same argument has been passed to the function before (that is, is the same argument in memory).
stable-memo only evaluates keys to WHNF.
This can be more suitable for recursive functions over graphs with cycles.
stable-memo doesn't retain the keys it has seen so far, which allows them to be garbage collected if they will no longer be used. Finalizers are put in place to remove the corresponding entries from the memo table if this happens.
Data.StableMemo.Weak provides an alternative set of combinators that also avoid retaining the results of the function, only reusing results if they have not yet been garbage collected.
There is no type class constraint on the function's argument.
stable-memo will not work for arguments which happen to have the same value but are not the same heap object. This rules out many candidates for memoization, such as the most common example, the naive Fibonacci implementation whose domain is machine Ints; it can still be made to work for some domains, though, such as the lazy naturals.
Ekmett just uploaded a library that handles this and more (produced at HacPhi): http://hackage.haskell.org/package/intern. He assures me that it is thread safe.
Edit: Actually, strictly speaking I realize this does something rather different. But I think you can use it for your purposes. It's really more of a stringtable-atom type interning library that works over arbitrary data structures (including recursive ones). It uses WeakPtrs internally to maintain the table. However, it uses Ints to index the values to avoid structural equality checks, which means packing them into the data type, when what you want are apparently actually StableNames. So I realize this answers a related question, but requires modifying your data type, which you want to avoid...
Closed. This question needs to be more focused. It is not currently accepting answers.
Closed 7 years ago.
Locked. This question and its answers are locked because the question is off-topic but has historical significance. It is not currently accepting new answers or interactions.
I keep hearing this term tossed around in several different contexts. What is it?
Declarative programming is when you write your code in such a way that it describes what you want to do, and not how you want to do it. It is left up to the compiler to figure out the how.
Examples of declarative programming languages are SQL and Prolog.
The other answers already do a fantastic job explaining what declarative programming is, so I'm just going to provide some examples of why that might be useful.
Context Independence
Declarative Programs are context-independent. Because they only declare what the ultimate goal is, but not the intermediary steps to reach that goal, the same program can be used in different contexts. This is hard to do with imperative programs, because they often depend on the context (e.g. hidden state).
Take yacc as an example. It's a parser generator aka. compiler compiler, an external declarative DSL for describing the grammar of a language, so that a parser for that language can automatically be generated from the description. Because of its context independence, you can do many different things with such a grammar:
Generate a C parser for that grammar (the original use case for yacc)
Generate a C++ parser for that grammar
Generate a Java parser for that grammar (using Jay)
Generate a C# parser for that grammar (using GPPG)
Generate a Ruby parser for that grammar (using Racc)
Generate a tree visualization for that grammar (using GraphViz)
simply do some pretty-printing, fancy-formatting and syntax highlighting of the yacc source file itself and include it in your Reference Manual as a syntactic specification of your language
And many more …
Optimization
Because you don't prescribe the computer which steps to take and in what order, it can rearrange your program much more freely, maybe even execute some tasks in parallel. A good example is a query planner and query optimizer for a SQL database. Most SQL databases allow you to display the query that they are actually executing vs. the query that you asked them to execute. Often, those queries look nothing like each other. The query planner takes things into account that you wouldn't even have dreamed of: rotational latency of the disk platter, for example or the fact that some completely different application for a completely different user just executed a similar query and the table that you are joining with and that you worked so hard to avoid loading is already in memory anyway.
There is an interesting trade-off here: the machine has to work harder to figure out how to do something than it would in an imperative language, but when it does figure it out, it has much more freedom and much more information for the optimization stage.
Loosely:
Declarative programming tends towards:-
Sets of declarations, or declarative statements, each of which has meaning (often in the problem domain) and may be understood independently and in isolation.
Imperative programming tends towards:-
Sequences of commands, each of which perform some action; but which may or may not have meaning in the problem domain.
As a result, an imperative style helps the reader to understand the mechanics of what the system is actually doing, but may give little insight into the problem that it is intended to solve. On the other hand, a declarative style helps the reader to understand the problem domain and the approach that the system takes towards the solution of the problem, but is less informative on the matter of mechanics.
Real programs (even ones written in languages that favor the ends of the spectrum, such as ProLog or C) tend to have both styles present to various degrees at various points, to satisfy the varying complexities and communication needs of the piece. One style is not superior to the other; they just serve different purposes, and, as with many things in life, moderation is key.
Here's an example.
In CSS (used to style HTML pages), if you want an image element to be 100 pixels high and 100 pixels wide, you simply "declare" that that's what you want as follows:
#myImageId {
height: 100px;
width: 100px;
}
You can consider CSS a declarative "style sheet" language.
The browser engine that reads and interprets this CSS is free to make the image appear this tall and this wide however it wants. Different browser engines (e.g., the engine for IE, the engine for Chrome) will implement this task differently.
Their unique implementations are, of course, NOT written in a declarative language but in a procedural one like Assembly, C, C++, Java, JavaScript, or Python. That code is a bunch of steps to be carried out step by step (and might include function calls). It might do things like interpolate pixel values, and render on the screen.
I am sorry, but I must disagree with many of the other answers. I would like to stop this muddled misunderstanding of the definition of declarative programming.
Definition
Referential transparency (RT) of the sub-expressions is the only required attribute of a declarative programming expression, because it is the only attribute which is not shared with imperative programming.
Other cited attributes of declarative programming, derive from this RT. Please click the hyperlink above for the detailed explanation.
Spreadsheet example
Two answers mentioned spreadsheet programming. In the cases where the spreadsheet programming (a.k.a. formulas) does not access mutable global state, then it is declarative programming. This is because the mutable cell values are the monolithic input and output of the main() (the entire program). The new values are not written to the cells after each formula is executed, thus they are not mutable for the life of the declarative program (execution of all the formulas in the spreadsheet). Thus relative to each other, the formulas view these mutable cells as immutable. An RT function is allowed to access immutable global state (and also mutable local state).
Thus the ability to mutate the values in the cells when the program terminates (as an output from main()), does not make them mutable stored values in the context of the rules. The key distinction is the cell values are not updated after each spreadsheet formula is performed, thus the order of performing the formulas does not matter. The cell values are updated after all the declarative formulas have been performed.
Declarative programming is the picture, where imperative programming is instructions for painting that picture.
You're writing in a declarative style if you're "Telling it what it is", rather than describing the steps the computer should take to get to where you want it.
When you use XML to mark-up data, you're using declarative programming because you're saying "This is a person, that is a birthday, and over there is a street address".
Some examples of where declarative and imperative programming get combined for greater effect:
Windows Presentation Foundation uses declarative XML syntax to describe what a user interface looks like, and what the relationships (bindings) are between controls and underlying data structures.
Structured configuration files use declarative syntax (as simple as "key=value" pairs) to identify what a string or value of data means.
HTML marks up text with tags that describe what role each piece of text has in relation to the whole document.
Declarative Programming is programming with declarations, i.e. declarative sentences. Declarative sentences have a number of properties that distinguish them from imperative sentences. In particular, declarations are:
commutative (can be reordered)
associative (can be regrouped)
idempotent (can repeat without change in meaning)
monotonic (declarations don't subtract information)
A relevant point is that these are all structural properties and are orthogonal to subject matter. Declarative is not about "What vs. How". We can declare (represent and constrain) a "how" just as easily as we declare a "what". Declarative is about structure, not content. Declarative programming has a significant impact on how we abstract and refactor our code, and how we modularize it into subprograms, but not so much on the domain model.
Often, we can convert from imperative to declarative by adding context. E.g. from "Turn left. (... wait for it ...) Turn Right." to "Bob will turn left at intersection of Foo and Bar at 11:01. Bob will turn right at the intersection of Bar and Baz at 11:06." Note that in the latter case the sentences are idempotent and commutative, whereas in the former case rearranging or repeating the sentences would severely change the meaning of the program.
Regarding monotonic, declarations can add constraints which subtract possibilities. But constraints still add information (more precisely, constraints are information). If we need time-varying declarations, it is typical to model this with explicit temporal semantics - e.g. from "the ball is flat" to "the ball is flat at time T". If we have two contradictory declarations, we have an inconsistent declarative system, though this might be resolved by introducing soft constraints (priorities, probabilities, etc.) or leveraging a paraconsistent logic.
Describing to a computer what you want, not how to do something.
imagine an excel page. With columns populated with formulas to calculate you tax return.
All the logic is done declared in the cells, the order of the calculation is by determine by formula itself rather than procedurally.
That is sort of what declarative programming is all about. You declare the problem space and the solution rather than the flow of the program.
Prolog is the only declarative language I've use. It requires a different kind of thinking but it's good to learn if just to expose you to something other than the typical procedural programming language.
I have refined my understanding of declarative programming, since Dec 2011 when I provided an answer to this question. Here follows my current understanding.
The long version of my understanding (research) is detailed at this link, which you should read to gain a deep understanding of the summary I will provide below.
Imperative programming is where mutable state is stored and read, thus the ordering and/or duplication of program instructions can alter the behavior (semantics) of the program (and even cause a bug, i.e. unintended behavior).
In the most naive and extreme sense (which I asserted in my prior answer), declarative programming (DP) is avoiding all stored mutable state, thus the ordering and/or duplication of program instructions can NOT alter the behavior (semantics) of the program.
However, such an extreme definition would not be very useful in the real world, since nearly every program involves stored mutable state. The spreadsheet example conforms to this extreme definition of DP, because the entire program code is run to completion with one static copy of the input state, before the new states are stored. Then if any state is changed, this is repeated. But most real world programs can't be limited to such a monolithic model of state changes.
A more useful definition of DP is that the ordering and/or duplication of programming instructions do not alter any opaque semantics. In other words, there are not hidden random changes in semantics occurring-- any changes in program instruction order and/or duplication cause only intended and transparent changes to the program's behavior.
The next step would be to talk about which programming models or paradigms aid in DP, but that is not the question here.
It's a method of programming based around describing what something should do or be instead of describing how it should work.
In other words, you don't write algorithms made of expressions, you just layout how you want things to be. Two good examples are HTML and WPF.
This Wikipedia article is a good overview: http://en.wikipedia.org/wiki/Declarative_programming
Since I wrote my prior answer, I have formulated a new definition of the declarative property which is quoted below. I have also defined imperative programming as the dual property.
This definition is superior to the one I provided in my prior answer, because it is succinct and it is more general. But it may be more difficult to grok, because the implication of the incompleteness theorems applicable to programming and life in general are difficult for humans to wrap their mind around.
The quoted explanation of the definition discusses the role pure functional programming plays in declarative programming.
Declarative vs. Imperative
The declarative property is weird, obtuse, and difficult to capture in a technically precise definition that remains general and not ambiguous, because it is a naive notion that we can declare the meaning (a.k.a semantics) of the program without incurring unintended side effects. There is an inherent tension between expression of meaning and avoidance of unintended effects, and this tension actually derives from the incompleteness theorems of programming and our universe.
It is oversimplification, technically imprecise, and often ambiguous to define declarative as “what to do” and imperative as “how to do”. An ambiguous case is the “what” is the “how” in a program that outputs a program— a compiler.
Evidently the unbounded recursion that makes a language Turing complete, is also analogously in the semantics— not only in the syntactical structure of evaluation (a.k.a. operational semantics). This is logically an example analogous to Gödel's theorem— “any complete system of axioms is also inconsistent”. Ponder the contradictory weirdness of that quote! It is also an example that demonstrates how the expression of semantics does not have a provable bound, thus we can't prove2 that a program (and analogously its semantics) halt a.k.a. the Halting theorem.
The incompleteness theorems derive from the fundamental nature of our universe, which as stated in the Second Law of Thermodynamics is “the entropy (a.k.a. the # of independent possibilities) is trending to maximum forever”. The coding and design of a program is never finished— it's alive!— because it attempts to address a real world need, and the semantics of the real world are always changing and trending to more possibilities. Humans never stop discovering new things (including errors in programs ;-).
To precisely and technically capture this aforementioned desired notion within this weird universe that has no edge (ponder that! there is no “outside” of our universe), requires a terse but deceptively-not-simple definition which will sound incorrect until it is explained deeply.
Definition:
The declarative property is where there can exist only one possible set of statements that can express each specific modular semantic.
The imperative property3 is the dual, where semantics are inconsistent under composition and/or can be expressed with variations of sets of statements.
This definition of declarative is distinctively local in semantic scope, meaning that it requires that a modular semantic maintain its consistent meaning regardless where and how it's instantiated and employed in global scope. Thus each declarative modular semantic should be intrinsically orthogonal to all possible others— and not an impossible (due to incompleteness theorems) global algorithm or model for witnessing consistency, which is also the point of “More Is Not Always Better” by Robert Harper, Professor of Computer Science at Carnegie Mellon University, one of the designers of Standard ML.
Examples of these modular declarative semantics include category theory functors e.g. the Applicative, nominal typing, namespaces, named fields, and w.r.t. to operational level of semantics then pure functional programming.
Thus well designed declarative languages can more clearly express meaning, albeit with some loss of generality in what can be expressed, yet a gain in what can be expressed with intrinsic consistency.
An example of the aforementioned definition is the set of formulas in the cells of a spreadsheet program— which are not expected to give the same meaning when moved to different column and row cells, i.e. cell identifiers changed. The cell identifiers are part of and not superfluous to the intended meaning. So each spreadsheet result is unique w.r.t. to the cell identifiers in a set of formulas. The consistent modular semantic in this case is use of cell identifiers as the input and output of pure functions for cells formulas (see below).
Hyper Text Markup Language a.k.a. HTML— the language for static web pages— is an example of a highly (but not perfectly3) declarative language that (at least before HTML 5) had no capability to express dynamic behavior. HTML is perhaps the easiest language to learn. For dynamic behavior, an imperative scripting language such as JavaScript was usually combined with HTML. HTML without JavaScript fits the declarative definition because each nominal type (i.e. the tags) maintains its consistent meaning under composition within the rules of the syntax.
A competing definition for declarative is the commutative and idempotent properties of the semantic statements, i.e. that statements can be reordered and duplicated without changing the meaning. For example, statements assigning values to named fields can be reordered and duplicated without changed the meaning of the program, if those names are modular w.r.t. to any implied order. Names sometimes imply an order, e.g. cell identifiers include their column and row position— moving a total on spreadsheet changes its meaning. Otherwise, these properties implicitly require global consistency of semantics. It is generally impossible to design the semantics of statements so they remain consistent if randomly ordered or duplicated, because order and duplication are intrinsic to semantics. For example, the statements “Foo exists” (or construction) and “Foo does not exist” (and destruction). If one considers random inconsistency endemical of the intended semantics, then one accepts this definition as general enough for the declarative property. In essence this definition is vacuous as a generalized definition because it attempts to make consistency orthogonal to semantics, i.e. to defy the fact that the universe of semantics is dynamically unbounded and can't be captured in a global coherence paradigm.
Requiring the commutative and idempotent properties for the (structural evaluation order of the) lower-level operational semantics converts operational semantics to a declarative localized modular semantic, e.g. pure functional programming (including recursion instead of imperative loops). Then the operational order of the implementation details do not impact (i.e. spread globally into) the consistency of the higher-level semantics. For example, the order of evaluation of (and theoretically also the duplication of) the spreadsheet formulas doesn't matter because the outputs are not copied to the inputs until after all outputs have been computed, i.e. analogous to pure functions.
C, Java, C++, C#, PHP, and JavaScript aren't particularly declarative.
Copute's syntax and Python's syntax are more declaratively coupled to
intended results, i.e. consistent syntactical semantics that eliminate the extraneous so one can readily
comprehend code after they've forgotten it. Copute and Haskell enforce
determinism of the operational semantics and encourage “don't repeat
yourself” (DRY), because they only allow the pure functional paradigm.
2 Even where we can prove the semantics of a program, e.g. with the language Coq, this is limited to the semantics that are expressed in the typing, and typing can never capture all of the semantics of a program— not even for languages that are not Turing complete, e.g. with HTML+CSS it is possible to express inconsistent combinations which thus have undefined semantics.
3 Many explanations incorrectly claim that only imperative programming has syntactically ordered statements. I clarified this confusion between imperative and functional programming. For example, the order of HTML statements does not reduce the consistency of their meaning.
Edit: I posted the following comment to Robert Harper's blog:
in functional programming ... the range of variation of a variable is a type
Depending on how one distinguishes functional from imperative
programming, your ‘assignable’ in an imperative program also may have
a type placing a bound on its variability.
The only non-muddled definition I currently appreciate for functional
programming is a) functions as first-class objects and types, b)
preference for recursion over loops, and/or c) pure functions— i.e.
those functions which do not impact the desired semantics of the
program when memoized (thus perfectly pure functional
programming doesn't exist in a general purpose denotational semantics
due to impacts of operational semantics, e.g. memory
allocation).
The idempotent property of a pure function means the function call on
its variables can be substituted by its value, which is not generally
the case for the arguments of an imperative procedure. Pure functions
seem to be declarative w.r.t. to the uncomposed state transitions
between the input and result types.
But the composition of pure functions does not maintain any such
consistency, because it is possible to model a side-effect (global
state) imperative process in a pure functional programming language,
e.g. Haskell's IOMonad and moreover it is entirely impossible to
prevent doing such in any Turing complete pure functional programming
language.
As I wrote in 2012 which seems to the similar consensus of
comments in your recent blog, that declarative programming is an
attempt to capture the notion that the intended semantics are never
opaque. Examples of opaque semantics are dependence on order,
dependence on erasure of higher-level semantics at the operational
semantics layer (e.g. casts are not conversions and reified generics
limit higher-level semantics), and dependence on variable values
which can not be checked (proved correct) by the programming language.
Thus I have concluded that only non-Turing complete languages can be
declarative.
Thus one unambiguous and distinct attribute of a declarative language
could be that its output can be proven to obey some enumerable set of
generative rules. For example, for any specific HTML program (ignoring
differences in the ways interpreters diverge) that is not scripted
(i.e. is not Turing complete) then its output variability can be
enumerable. Or more succinctly an HTML program is a pure function of
its variability. Ditto a spreadsheet program is a pure function of its
input variables.
So it seems to me that declarative languages are the antithesis of
unbounded recursion, i.e. per Gödel's second incompleteness
theorem self-referential theorems can't be proven.
Lesie Lamport wrote a fairytale about how Euclid might have
worked around Gödel's incompleteness theorems applied to math proofs
in the programming language context by to congruence between types and
logic (Curry-Howard correspondence, etc).
Declarative programming is "the act of programming in languages that conform to the mental model of the developer rather than the operational model of the machine".
The difference between declarative and imperative programming is well
illustrated by the problem of parsing structured data.
An imperative program would use mutually recursive functions to consume input
and generate data. A declarative program would express a grammar that defines
the structure of the data so that it can then be parsed.
The difference between these two approaches is that the declarative program
creates a new language that is more closely mapped to the mental model of the
problem than is its host language.
It may sound odd, but I'd add Excel (or any spreadsheet really) to the list of declarative systems. A good example of this is given here.
I'd explain it as DP is a way to express
A goal expression, the conditions for - what we are searching for. Is there one, maybe or many?
Some known facts
Rules that extend the know facts
...and where there is a deduct engine usually working with a unification algorithm to find the goals.
As far as I can tell, it started being used to describe programming systems like Prolog, because prolog is (supposedly) about declaring things in an abstract way.
It increasingly means very little, as it has the definition given by the users above. It should be clear that there is a gulf between the declarative programming of Haskell, as against the declarative programming of HTML.
A couple other examples of declarative programming:
ASP.Net markup for databinding. It just says "fill this grid with this source", for example, and leaves it to the system for how that happens.
Linq expressions
Declarative programming is nice because it can help simplify your mental model* of code, and because it might eventually be more scalable.
For example, let's say you have a function that does something to each element in an array or list. Traditional code would look like this:
foreach (object item in MyList)
{
DoSomething(item);
}
No big deal there. But what if you use the more-declarative syntax and instead define DoSomething() as an Action? Then you can say it this way:
MyList.ForEach(DoSometing);
This is, of course, more concise. But I'm sure you have more concerns than just saving two lines of code here and there. Performance, for example. The old way, processing had to be done in sequence. What if the .ForEach() method had a way for you to signal that it could handle the processing in parallel, automatically? Now all of a sudden you've made your code multi-threaded in a very safe way and only changed one line of code. And, in fact, there's a an extension for .Net that lets you do just that.
If you follow that link, it takes you to a blog post by a friend of mine. The whole post is a little long, but you can scroll down to the heading titled "The Problem" _and pick it up there no problem.*
It depends on how you submit the answer to the text. Overall you can look at the programme at a certain view but it depends what angle you look at the problem. I will get you started with the programme:
Dim Bus, Car, Time, Height As Integr
Again it depends on what the problem is an overall. You might have to shorten it due to the programme. Hope this helps and need the feedback if it does not.
Thank You.