GitHub/gf-core
1
0
Fork 1
mirror of https://github.com/GrammaticalFramework/gf-core.git synced 2026-04-14 07:19:31 -06:00
Code Issues Packages Projects Releases Wiki Activity
Files
351f8aada4cb750947af892d72ee935703de1cfc
gf-core/doc/gf-modules.html
aarne 351f8aada4 more tutorial to the module system
2005-04-10 10:39:34 +00:00

682 lines
23 KiB
HTML
Raw Blame History

<html>
<body bgcolor="#FFFFFF" text="#000000">
<center>
<img src="gf-logo.gif">
<h1>The Module System of GF</h1>
<p>
8/4/2005
<p>
<a href="http://www.cs.chalmers.se/~aarne">Aarne Ranta</a>
</center>
A GF grammar consists of a set of <b>modules</b>, which can be
combined in different ways to build different grammars.
There are several different <b>types of modules</b>:
<ul>
<li> <tt>abstract</tt>
<li> <tt>concrete</tt>
<li> <tt>resource</tt>
<li> <tt>interface</tt>
<li> <tt>instance</tt>
<li> <tt>incomplete concrete</tt>
<li> <tt>transfer</tt>
</ul>
We will go through the module types in this order, which is also
their order of "importance" from the most frequently used to
the more esoteric/advanced ones.
<p>
This document is meant as an appendix to the GF tutorial, and
presupposes knowledge of GF judgements and expressions. It aims
just to tell what module system adds to the old functionality;
some information is repeated to give understanding on how the
module system related to the already familiar uses of GF grammars.
<h2>The principal module types</h2>
<h3>Abstract syntax</h3>
Any GF grammar that is used in an application
will probably contain at least one module
of the <tt>abstract</tt> module type. Here is an example of
such a module, defining a fragment of propositional logic.
<pre>
abstract Logic = {
cat Prop ;
fun Conj : Prop -> Prop -> Prop ;
fun Disj : Prop -> Prop -> Prop ;
fun Impl : Prop -> Prop -> Prop ;
fun Falsum : Prop ;
}
</pre>
The <b>name</b> of this module is <tt>Logic</tt>.
<p>
An <tt>abstract</tt> module defines an <b>abstract syntax</b>, which
is a language-independent representation of a fragment of language.
It consists of two kinds of <b>judgements</b>:
<ul>
<li> <tt>cat</tt> judgements telling what <b>categories</b> there are
(types of abstract syntax trees)
<li> <tt>fun</tt> judgements telling what <b>functions</b> there are
(to build abstract syntax trees)
</ul>
There can also be <tt>def</tt> and <tt>data</tt> judgements in an
abstract syntax.
<h4>Compilation of abstract syntax</h4>
The GF grammar compiler expects to find the module <tt>Logic</tt> in a file named
<tt>Logic.gf</tt>. When the compiler is run, it produces
another file, named <tt>Logic.gfc</tt>. This file is in the
format called <b>canonical GF</b>, which is the "machine language"
of GF. Next time that the module <tt>Logic</tt> is needed in
compiling a grammar, it can be read from the compiled (<tt>gfc</tt>)
file instead of the source (<tt>gf</tt>) file, unless the source
has been changed after the compilation.
<h3>Concrete syntax</h3>
In order for a GF grammar to describe a concrete language, the abstract
syntax must be completed with a <b>concrete syntax</b> of it.
For this purpose, we use modules of type <tt>concrete</tt>: for instance,
<pre>
concrete LogicEng of Logic = {
lincat Prop = {s : Str} ;
lin Conj a b = {s = a.s ++ "and" ++ b.s} ;
lin Disj a b = {s = a.s ++ "or" ++ b.s} ;
lin Impl a b = {s = "if" ++ a.s ++ "then" ++ b.s} ;
lin Falsum = {s = ["we have a contradiction"]} ;
}
</pre>
The module <tt>LogicEng</tt> is a concrete syntax <tt>of</tt> the
abstract syntax <tt>Logic</tt>. The GF grammar compiler checks that
the concrete is valid with respect to the abstract syntax <tt>of</tt>
which it is claimed to be. The validity requires that there has to be
<ul>
<li> a <tt>lincat</tt> judgement for each <tt>cat</tt> judgement, telling what the
<b>linearization types</b> of categories are
<li> a <tt>lin</tt> judgement for each <tt>fun</tt> judgement, telling what the
<b>linearization functions</b> corresponding to functions are
</ul>
Validity also requires that the linearization functions defined by
<tt>lin</tt> judgements are type-correct with respect to the
linearization types of the arguments and value of the function.
<p>
There can also be <tt>lindef</tt> and <tt>printname</tt> judgements in a
concrete syntax.
<h3>Top-level grammar</h3>
When a <tt>concrete</tt> module is successfully compiled, a <tt>gfc</tt>
file is produced in the same way as for <tt>abstract</tt> modules. The
pair of an <tt>abstract</tt> and a corresponding <tt>concrete</tt> module
is a <b>top-level grammar</b>, which can be used in the GF system to
perform various tasks. The most fundamental tasks are
<ul>
<li> <b>linearization</b>: take an abstract syntax tree and find the corresponding string
<li> <b>parsing</b>: take a string and find the corresponding abstract syntax
trees (which can be zero, one, or many)
</ul>
In the current grammar, infinitely many trees and strings are recognized, although
no very interesting ones. For example, the tree
<pre>
Impl (Disj Falsum Falsum) Falsum
</pre>
has the linearization
<pre>
if we have a contradiction or we have a contradiction then we have a contradiction
</pre>
which in turn can be parsed uniquely as that tree.
<h4>Compiling top-level grammars</h4>
When GF compiles the module <tt>LogicEng</tt> it also has to compile
all modules that it <b>depends</b> on (in this case, just <tt>Logic</tt>).
The compilation process starts with dependency analysis to find
all these modules, recursively, starting from the explicitly imported one.
The compiler then reads either <tt>gf</tt> or <tt>gfc</tt> files, in
a dependency order. The decision on which files to read depends on
time stamps and dependencies in a natural way, so that all and only
those modules that have to be compiled are compiled. (This behaviour can
be changed with flags, see below.)
<h4>Using top-level grammars</h4>
To use a top-level grammar in the GF system, one uses the <tt>import</tt>
command (short name <tt>i</tt>). For instance,
<pre>
i LogicEng.gf
</pre>
It is also possible to specify the imported grammar(s) on the command
line when invoking GF:
<pre>
gf LogicEng.gf
</pre>
Various <b>compilation flags</b> can be added to both ways of compiling a module:
<ul>
<li> <tt>-src</tt> forces compilation form source files
<li> <tt>-v</tt> gives more verbose information on compilation
<li> <tt>-s</tt> makes compilation silent (except if it fails with an error message)
</ul>
Importing a grammar makes it visible in GF's <b>internal state</b>. To see
what modules are available, use the command <tt>print_options</tt> (<tt>po</tt>).
You can empty the state with the command <tt>empty</tt> (<tt>e</tt>); this is
needed if you want to read in grammars with a different abstract syntax
than the current one without exiting GF.
<p>
Grammar modules can reside in different directories. They can then be found
by means of a <b>search path</b>, which is a flag such as
<pre>
-path=.:../prelude
</pre>
given to the <tt>import</tt> command or the shell command invoking GF.
(It can also be defined in the grammar file; see below.) The compiler
writes every <tt>gfc</tt> file in the same directory as the corresponding
<tt>gf</tt> file.
<p>
Parsing and linearization can be performed with the <tt>parse</tt>
(<tt>p</tt>) and <tt>linearize</tt> (<tt>l</tt>) commands, respectively.
For instance,
<pre>
> l Impl (Disj Falsum Falsum) Falsum
if we have a contradiction or we have a contradiction then we have a contradiction
> p -cat=Prop "we have a contradiction"
Falsum
</pre>
Notice that the <tt>parse</tt> command needs the parsing category
as a flag. This necessary since a grammar can have several
possible parsing categories ("entry points").
<h3>Multilingual grammar</h3>
One <tt>abstract</tt> syntax can have several <tt>concrete</tt> syntaxes.
Here are two new ones for <tt>Logic</tt>:
<pre>
concrete LogicFre of Logic = {
lincat Prop = {s : Str} ;
lin Conj a b = {s = a.s ++ "et" ++ b.s} ;
lin Disj a b = {s = a.s ++ "ou" ++ b.s} ;
lin Impl a b = {s = "si" ++ a.s ++ "alors" ++ b.s} ;
lin Falsum = {s = ["nous avons une contradiction"]} ;
}
concrete LogicSymb of Logic = {
lincat Prop = {s : Str} ;
lin Conj a b = {s = "(" ++ a.s ++ "&" ++ b.s ++ ")"} ;
lin Disj a b = {s = "(" ++ a.s ++ "v" ++ b.s ++ ")"} ;
lin Impl a b = {s = "(" ++ a.s ++ "->" ++ b.s ++ ")"} ;
lin Falsum = {s = "_|_"} ;
}
</pre>
The four modules <tt>Logic</tt>, <tt>LogicEng</tt>, <tt>LogicFre</tt>, and
<tt>LogicSymb</tt> together form a <b>multilingual grammar</b>, in which
it is possible to perform parsing and linearization with respect to any
of the concrete syntaxes. As a combination of parsing and linearization,
one can also perform <b>translation</b> from one language to another.
(By <b>language</b> we mean the set of expressions generated by one
concrete syntax.)
<h4>Using multilingual grammars</h4>
Any combination of abstract syntax and corresponding concrete syntaxes
is thus a multilingual grammar. With many languages and other enrichments
(as described below), a multilingual grammar easily grows to the size of
tens of modules. The grammar developer, having finished her job, can
package the result in a <b>multilingual canonical grammar</b>, a file
with the suffix <tt>.gfcm</tt>. For instance, to compile the set of grammars
described by now, the following sequence of GF commands can be used:
<pre>
i LogicEng.gf
i LogicFre.gf
i LogicSymb.gf
pm | wf logic.gfcm
</pre>
The "end user" of the grammar only needs the file <tt>logic.gfcm</tt> to
access all the functionality of the multilingual grammar. It can be
imported in the GF system in the same way as <tt>.gf</tt> files. But
it can also be used in the <b>Embedded Java Interpreter for GF</b> to
build Java programs of which the multilingual grammar functionalities
(linearization, parsing, translation) form a part.
<p>
In a multilingual grammar, the concrete syntax module names work as
names of languages that can be selected for linearization and parsing:
<pre>
> l -lang=LogicFre Impl Falsum Falsum
si nous avons une contradiction alors nous avons une contradiction
> l -lang=LogicSymb Impl Falsum Falsum
( _|_ -> _|_ )
> p -cat=Prop -lang=LogicSymb "( _|_ & _|_ )"
Conj Falsum Falsum
</pre>
The option <tt>-multi</tt> gives linearization to all languages:
<pre>
> l -multi Impl Falsum Falsum
if we have a contradiction then we have a contradiction
si nous avons une contradiction alors nous avons une contradiction
( _|_ -> _|_ )
</pre>
Translation can be obtained by using a <b>pipe</b> from a parser
to a linearizer:
<pre>
> p -cat=Prop -lang=LogicSymb "( _|_ & _|_ )" | l -lang=LogicEng
if we have a contradiction then we have a contradiction
</pre>
<h4>Exercise</h4>
Write yet another concrete syntax of <tt>Logic</tt>, for
a language or symbolic notation of your choice.
<h3>Resource modules</h3>
The <tt>concrete</tt> modules shown above would look much nicer if
we used the main idea of functional programming: avoid repetitive
code by using <b>functions</b> that capture repeated patterns of
expressions. A collection of such functions can be a valuable
<b>resource</b> for a programmer, reusable in many different
top-level grammars. Thus we introduce the <tt>resource</tt>
module type, with the first example
<pre>
resource Util = {
oper SS : Type = {s : Str} ;
oper ss : Str -> SS = \s -> {s = s} ;
oper paren : Str -> Str = \s -> "(" ++ s ++ ")" ;
oper infix : Str -> SS -> SS -> SS = \h,x,y ->
ss (x.s ++ h ++ y.s) ;
oper infixp : Str -> SS -> SS -> SS = \h,x,y ->
ss (paren (infix h x y)) ;
}
</pre>
Modules of <tt>resource</tt> type have two forms of judgement:
<ul>
<li> <tt>oper</tt> defining auxiliary operations
<li> <tt>param</tt> defining parameter types
</ul>
A <tt>resource</tt> can be used in a <tt>concrete</tt> (or another
<tt>resource</tt>) by <tt>open</tt>ing it. This means that
all operations (and parameter types) defined in the resource
module become usable in module that opens it. For instance,
we can rewrite the module <tt>LogicSymb</tt> much more concisely:
<pre>
concrete LogicSymb of Logic = open Util in {
lincat Prop = SS ;
lin Conj = infixp "&" ;
lin Disj = infixp "v" ;
lin Impl = infixp "->" ;
lin Falsum = ss "_|_" ;
}
</pre>
What happens when this variant of <tt>LogicSymb</tt> is
compiled is that the <tt>oper</tt>-defined constants
of <tt>Util</tt> are <b>inlined</b> in the
right-hand-sides of the judgements of <tt>LogicSymb</tt>,
and these expressions are <b>partially evaluated</b>, i.e.
computed as far as possible. The generated <tt>gfc</tt> file
will look just like the file generated for the first version
of <tt>LogicSymb</tt> - at least, it will do the same job.
<p>
Several <tt>resource</tt> modules can be <tt>open</tt>ed
at the same time. If the modules contain same names, the
conflict can be resolved by <b>qualified</b> opening and
reference. For instance,
<pre>
concrete LogicSymb of Logic = open Util, Prelude in { ...
} ;
</pre>
(where <tt>Prelude</tt> is a standard library of GF) brings
into scope two definitions of the constant <tt>SS</tt>.
To specify which one is used, you can write
<tt>Util.SS</tt> or <tt>Prelude.SS</tt> instead of just <tt>SS</tt>.
You can also introduce abbreviations to avoid long qualifiers, e.g.
<pre>
concrete LogicSymb of Logic = open (U=Util), (P=Prelude) in { ...
} ;
</pre>
which means that you can write <tt>U.SS</tt> and <tt>P.SS</tt>.
<h4>Compiling resource modules</h4>
The compilation of a <tt>resource</tt> module differs
from the compilation of <tt>abstract</tt> and
<tt>concrete</tt> modules because <tt>oper</tt> operations
do not in general have values in <tt>gfc</tt>. A <tt>gfc</tt>
file <i>is</i> generated, but it contains only
<tt>param</tt> judgements (also recall that <tt>oper</tt>s
are inlined in their top-level use sites, so it is not
necessary to save them in the compiled grammar).
However, since computing the operations over and over
again can be time comsuming, and since type checking
<tt>resource</tt> modules also takes time, a third kind
of file is generated for resource modules: a <tt>.gfr</tt>
file. This file is written in the GF source code notation,
but it is type checked and type annotated, and <tt>oper</tt>s
are computed as far as possible.
<p>
If you look at any <tt>gfc</tt> or <tt>gfr</tt> file generated
by the GF compiler, you see that all names have been replaced by
their qualified variants. This is an important first step (after parsing)
the compiler does. As for the commands in the GF shell, some output
qualified names and some not. The difference does not always result
from firm principles.
<h4>Using resource modules</h4>
The typical use is through <tt>open</tt> in a
<tt>concrete</tt> module, which means that
<tt>resource</tt> modules are not imported on their own.
However, when developing and testing phase of grammars, it
can be useful to evaluate <tt>oper</tt>s with different
arguments. To prevent them from being inlined away, the
<tt>-retain</tt> option can be used:
<pre>
> i -retain Util.gf
</pre>
The command <tt>compute_concrete</tt> (<tt>cc</tt>)
can now be used for evaluating expressions that may concain
operations defined in <tt>Util</tt>:
<pre>
> cc ss (paren "foo")
{s = "(" ++ "foo" ++ ")"}
</pre>
To find out, what <tt>oper</tt>s are available for a given type,
the command <tt>show_operations</tt> (<tt>so</tt>) can be used:
<pre>
> so SS
Util.ss : Str -> SS ;
Util.infix : Str -> SS -> SS -> SS ;
Util.infixp : Str -> SS -> SS -> SS ;
</pre>
<h4>Exercise</h4>
Rewrite the modules <tt>LogicEng</tt> and <tt>LogicFre</tt>
by making use of the resource.
<h3>Inheritance</h3>
The most characteristic modularity of GF lies in the division of
grammars into <tt>abstract</tt>, <tt>concrete</tt>, and
<tt>resource</tt> modules. This permits writing multilingual
grammar and sharing the maximum of code between different
languages.
<p>
In addition to its special modularity, GF provides <b>inheritance</b>,
which is familiar from other programming languages (in particular,
object-oriented ones). Inheritance means that a module inherits all
judgements from another module; we also say that it <b>extends</b>
the other module. Inheritance is useful to divide big grammars into
smaller units, and also to reuse the same units in different bigger
grammars.
<p>
The first example of inheritance is for abstract syntax. Let us
extend the module <tt>Logic</tt> to <tt>Arithmetic</tt>:
<pre>
abstract Arithmetic = Logic ** {
cat Nat ;
fun Even : Nat -> Prop ;
fun Odd : Nat -> Prop ;
fun Zero : Nat ;
fun Succ : Nat -> Nat ;
}
</pre>
In parallel with the extension of the abstract syntax
<tt>Logic</tt> to <tt>Arithmetic</tt>, we can extend
the concrete syntax <tt>LogicEng</tt> to <tt>ArithmeticEng</tt>:
<pre>
concrete ArithmeticEng of Arithmetic = LogicEng ** open Util in {
lincat Nat = SS ;
lin Even x = ss (x.s ++ "is" ++ "even") ;
lin Odd x = ss (x.s ++ "is" ++ "odd") ;
lin Zero = ss "zero" ;
lin Succ x = ss ("the" ++ "successor" ++ "of" ++ x.s) ;
}
</pre>
Another extension of <tt>Logic</tt> is <tt>Geometry</tt>,
<pre>
abstract Geometry = Logic ** {
cat Point ;
cat Line ;
fun Incident : Point -> Line -> Prop ;
}
</pre>
The corresponding concrete syntax is left as exercise.
<p>
Inheritance can be <b>multiple</b>, which means that a module
may extend many modules at the same time. Suppose, for instance,
that we want to build a module for mathematics covering both
arithmetic and geometry, and the underlying logic. We then write
<pre>
abstract Mathematics = Arithmetic, Geometry ** {
} ;
</pre>
We could of course add some new judgements in this module, but
it is not necessary to do so.
<p>
The module <tt>Mathematics</tt> also shows that it is possibe
to extend a module already built by extension. The correctness
criterion for extensions is that the same name
(<tt>cat</tt>, <tt>fun</tt>, <tt>oper</tt>, or <tt>param</tt>)
may not be defined twice in the resulting union of names.
That the names defined in <tt>Logic</tt> are "inherited twice"
by <tt>Mathematics</tt> (via both <tt>Arithmetic</tt> and
<tt>Geometry</tt>) is no violation of this rule; the usual
problems of multiple inheritance do not arise, since
the definitions of inherited constants cannot be changed.
<h4>Compiling inheritance</h4>
Inherited judgements are not copied into the inheriting modules.
Instead, an <b>indirection</b> is created for each inherited name,
as can be seen by looking into the generated <tt>gfc</tt> (and
<tt>gfr</tt>) files. Thus for instance the names
<pre>
Mathematics.Prop Arithmetic.Prop Logic.Prop
Geometry.Prop
</pre>
all refer to the same category, declared in the module
<tt>Logic</tt>.
<h4>Inspecting grammar hierarchies</h4>
The command <tt>visualize_graph</tt> (<tt>vg</tt>) shows the
dependency graph in the current GF shell state. The graph can
also be saved in a file and used e.g. in documentation, by the
command <tt>print_multi -graph</tt> (<tt>pm -graph</tt>).
<h3>Reuse of top-level grammars as resources</h3>
Top-level grammars have a straightforward translation to
<tt>resource</tt> modules. The translation concerns
pairs of abstract-concrete judgements:
<pre>
cat C ; ===> oper C : Type = T ;
lincat C = T ;
fun f : A ; ===> oper f : A = t ;
lin f = t ;
</pre>
Due to this translation, a <tt>concrete</tt> module
can be <tt>open</tt>ed in the same way as a
<tt>resource</tt> module; the translation is done
on the fly (it is computationally very cheap).
<p>
Modular grammar engineering often means that some grammarians
focus on the semantics of the domain whereas others take care
of linguistic details. Thus a typical reuse opens a
linguistically oriented <b>resource grammar</b>,
<pre>
abstract Resource = {
cat S ; NP ; A ;
fun PredA : NP -> A -> S ;
}
concrete ResourceEng of Resource = {
lincat S = ... ;
lin PredA = ... ;
}
</pre>
The <b>application grammar</b>, instead of giving linearizations
explicitly, just reduces them to categories and functions in the
resource grammar:
<pre>
concrete ArithmeticEng of Arithmetic = LogicEng ** open ResourceEng in {
lincat Nat = NP ;
lin Even x = PredA x (regA "even") ;
}
</pre>
If the resource grammar is only capable of generating grammatically
correct expressions, then the grammaticality of the application
grammar is also guaranteed: the type checker of GF is used as
grammar checker.
To guarantee distinctions between categories that have
the same linearization type, the actual translation used
in GF adds to every linearization type and linearization
a <b>lock field</b>,
<pre>
cat C ; ===> oper C : Type = T ** {lock_C : {}} ;
lincat C = T ;
fun f : ... -> C ; ===> oper f : A = t ** {lock_C = &lt;>};
lin f = t ;
</pre>
(Notice that the latter translation is type-correct because of
record subtyping, which means that <tt>t</tt> can ignore the
lock fields of its arguments.) An application grammarian who
only uses resource grammar categories and functions never
needs to write these lock fields herself. Having to do so
serves as a warning that the grammaticality guarantee given
by the resource grammar no longer holds.
<h2>Additional module types</h2>
<h3>Interfaces, instances, and incomplete grammars</h3>
One difference between top-level grammars and <tt>resource</tt>
modules is that the former systematically separete the
declarations of categories and functions from their definitions.
In the reuse translation creating and <tt>oper</tt> judgement,
the declaration coming from the <tt>abstract</tt> module is put
together with the definition coming from the <tt>concrete</tt>
module.
<p>
However, the separation of declarations and definitions is so
useful a notion that GF also has specific modules types that
<tt>resource</tt> modules into two parts. In this splitting,
an <tt>interface</tt> module corresponds to an abstract syntax,
in giving the declarations of operations (and parameter types).
For instance, a generic markup interface would look as follows:
<pre>
interface Markup = open Util in {
oper Boldface : Str -> Str ;
oper Heading : Str -> Str ;
oper markupSS : (Str -> Str) -> SS -> SS = \f,r ->
ss (f r.s) ;
}
</pre>
The definitions of the constants declared in an <tt>interface</tt>
are given in an <tt>instance</tt> module (which is always <tt>of</tt>
an interface, in the same way as a <tt>concrete</tt> is always
<tt>of</tt> an abstract). The following <tt>instance</tt>s
define markup in HTML and latex.
<pre>
instance MarkupHTML of Markup = open Util in {
oper Boldface s = "&lt;b>" ++ s ++ "&lt;/b>" ;
oper Heading s = "&lt;h2>" ++ s ++ "&lt;/h2>" ;
}
instance MarkupLatex of Markup = open Util in {
oper Boldface s = "\\textbf{" ++ s ++ "}" ;
oper Heading s = "\\section{" ++ s ++ "}" ;
}
</pre>
Notice that both <tt>interface</tt>s and <tt>instance</tt>s may
<tt>open</tt> <tt>resource</tt>s (and also reused top-level grammars).
An <tt>interface</tt> may moreover define some of the operations it
declares; these definitions are inherited by all instances and cannot
be changed in them. Inheritance by module extension
is possible, as always, between modules of the same type.
<h4>Using an interface</h4>
<h4>Instantiating an interface</h4>
<h4>Compiling interfaces, instances, and parametrized modules</h4>
<h3>Transfer modules</h3>
<h2>Summary of module syntax and semantics</h2>
</body>
</html>
Reference in New Issue View Git Blame Copy Permalink