GitHub/gf-core
1
0
Fork 1
mirror of https://github.com/GrammaticalFramework/gf-core.git synced 2026-04-09 04:59:31 -06:00
Code Issues Packages Projects Releases Wiki Activity

improved final-resource

Browse Source
This commit is contained in:
aarne
2006-06-13 14:40:12 +00:00
parent 161924b0e6
commit bd6929e4a7
14 changed files with 351 additions and 216 deletions
Download Patch File Download Diff File Expand all files Collapse all files

250
doc/final-resource.tex
View File

@@ -6,7 +6,7 @@
\setlength{\parskip}{8pt}\parindent=0pt % no paragraph indentation
\newcommand{\commOut}[1]{}
\newcommand{\subsubsubsection}[1]{\textit{#1}}
\newcommand{\subsubsubsection}[1]{\textbf{#1}.}
\title{The GF Resource Grammar Library}
\author{Author: Aarne Ranta}
@@ -25,7 +25,7 @@ module system, knowledge that can be acquired e.g. from the GF
tutorial. We start with an introduction to the library, and proceed to
details with the aim of covering all that one needs to know
in order to use the library.
How to write one's own resource grammar (i.e. implement the API for
How to write one's own resource grammar (i.e. to implement the API for
a new language), is covered by a separate Resource-HOWTO document.
\section{Motivation}
@@ -34,15 +34,30 @@ The GF Resource Grammar Library contains grammar rules for
is to make these rules available for application programmers,
who can thereby concentrate on the semantic and stylistic
aspects of their grammars, without having to think about
grammaticality. The level of a typical application grammarian
is skilled programmer, without knowledge linguistics, but with
grammaticality. The targeted level of application grammarians
is skilled programmer without knowledge linguistics, but with
a good knowledge of the target languages. Such a combination of
skilles is typical of a programmer who wants to localize a piece
of software to a new language.
skills is typical of programmers who want to localize
software to new languages.
To give an example, an application dealing with
music players may have a semantical category \texttt{Kind}, examples
of Kinds being Song and Artist. In German, for instance, Song
The current resource languages are
-\texttt{Dan}ish
-\texttt{Eng}lish
-\texttt{Fin}nish
-\texttt{Fre}nch
-\texttt{Ger}man
-\texttt{Ita}lian
-\texttt{Nor}wegian
-\texttt{Rus}sian
-\texttt{Spa}nish
-\texttt{Swe}dish
The first three letters (\texttt{Dan} etc) are used in grammar module names.
To give an example application, consider
music playing devices. In the application,
we may have a semantical category \texttt{Kind}, examples
of \texttt{Kind}s being \texttt{Song} and \texttt{Artist}. In German, for instance, \texttt{Song}
is linearized into the noun "Lied", but knowing this is not
enough to make the application work, because the noun must be
produced in both singular and plural, and in four different
@@ -54,13 +69,13 @@ write
\end{verbatim}
and the eight forms are correctly generated. The resource grammar
library contains a complete set of inflectional paradigms (such as
regN2 here), enabling the definition of any lexical items.
\texttt{regN2} here), enabling the definition of any lexical items.
The resource grammar library is not only about inflectional paradigms - it
also has syntax rules. The music player application
might also want to modify songs with properties, such as "American",
"old", "good". The German grammar for adjectival modifications is
particularly complex, because the adjectives have to agree in gender,
particularly complex, because adjectives have to agree in gender,
number, and case, and also depend on what determiner is used
("ein Amerikanisches Lied" vs. "das Amerikanische Lied"). All this
variation is taken care of by the resource grammar function
@@ -81,21 +96,21 @@ given that
lincat Kind = CN
\end{verbatim}
The resource library API is devided into language-specific and language-independet
parts. To put is roughly,
parts. To put it roughly,
\begin{itemize}
\item lexicon is language-specific
\item syntax is language-independent
\item the lexicon API is language-specific
\item the syntax API is language-independent
\end{itemize}
Thus, to render the above example in French instead of German, we need to
pick a different linearization of Song,
pick a different linearization of \texttt{Song},
\begin{verbatim}
lin Song = regGenN "chanson" feminine
\end{verbatim}
But to linearize PropKind, we can use the very same rule as in German.
The resource function AdjCN has different implementations in the two
But to linearize \texttt{PropKind}, we can use the very same rule as in German.
The resource function \texttt{AdjCN} has different implementations in the two
languages, but the application programmer need not care about the difference.
\subsection{A complete example}
@@ -115,7 +130,7 @@ The abstract syntax defines a "domain ontology":
}
\end{verbatim}
The concrete syntax is defined independently of language, by opening
two interfaces: the resource Grammar and an application lexicon.
two interfaces: the resource \texttt{Grammar} and an application lexicon.
\begin{verbatim}
incomplete concrete MusicI of Music = open Grammar, MusicLex in {
@@ -128,8 +143,8 @@ two interfaces: the resource Grammar and an application lexicon.
American = PositA american_A ;
}
\end{verbatim}
The application lexicon MusicLex has an abstract syntax, that extends
the resource category system Cat.
The application lexicon \texttt{MusicLex} has an abstract syntax that extends
the resource category system \texttt{Cat}.
\begin{verbatim}
abstract MusicLex = Cat ** {
@@ -151,11 +166,11 @@ module for that language:
concrete MusicLexFre of MusicLex = CatFre ** open ParadigmsFre in {
lin
song_N = regGenN "chanson" feminine ;
american_A = regA "américain" ;
american_A = regA "américain" ;
}
\end{verbatim}
The top-level Music grammars are obtained by instantiating the two interfaces
of MusicI:
The top-level \texttt{Music} grammars are obtained by instantiating the two interfaces
of \texttt{MusicI}:
\begin{verbatim}
concrete MusicGer of Music = MusicI with
@@ -166,13 +181,21 @@ of MusicI:
(Grammar = GrammarFre),
(MusicLex = MusicLexFre) ;
\end{verbatim}
To localize the system to a new language, all that is needed is two modules,
one implementing MusicLex and the other instantiating Music. The latter is
Both of these files can use the same \texttt{path}, defined as
\begin{verbatim}
--# -path=.:present:prelude
\end{verbatim}
The \texttt{present} category contains the compiled resources, restricted to
present tense; \texttt{alltenses} has the full resources.
To localize the music player system to a new language, all that is needed is two modules,
one implementing \texttt{MusicLex} and the other instantiating \texttt{Music}. The latter is
completely trivial, whereas the former one involves the choice of correct
vocabulary and inflectional paradigms. For instance, Finnish is added as follows:
\begin{verbatim}
concrete MusicLexFin of MusicLex = CatFre ** open ParadigmsFin in {
concrete MusicLexFin of MusicLex = CatFin ** open ParadigmsFin in {
lin
song_N = regN "kappale" ;
american_A = regA "amerikkalainen" ;
@@ -191,7 +214,7 @@ English, but that a relative clause would be preferrable. One can then start as
before,
\begin{verbatim}
concrete MusicLexEng of MusicLex = CatFre ** open ParadigmsEng in {
concrete MusicLexEng of MusicLex = CatEng ** open ParadigmsEng in {
lin
song_N = regN "song" ;
american_A = regA "American" ;
@@ -201,9 +224,10 @@ before,
(Grammar = GrammarEng),
(MusicLex = MusicLexEng) ;
\end{verbatim}
The module MusicEng0 would not be used on the top level, however, but
The module \texttt{MusicEng0} would not be used on the top level, however, but
another module would be built on top of it, with a restricted import from
MusicEng0. MusicEng inherits everything from MusicEng0 except PropKind, and
\texttt{MusicEng0}. \texttt{MusicEng} inherits everything from \texttt{MusicEng0}
except \texttt{PropKind}, and
gives its own definition of this function:
\begin{verbatim}
@@ -238,18 +262,18 @@ All of these problems should be solved in application grammars.
The task of resource grammars is just to take care of low-level linguistic
details such as inflection, agreement, and word order.
For the same reasons, resource grammars are not adequate for parsing.
It is for the same reasons that resource grammars are not adequate for translation.
That the syntax API is implemented for different languages of course makes
it possible to translate via it - but there is no guarantee of translation
equivalence. Of course, the use of parametrized implementations such as MusicI
equivalence. Of course, the use of parametrized implementations such as \texttt{MusicI}
above only extends to those cases where the syntax API does give translation
equivalence - but this must be seen as a limiting case, and real applications
will often use only restricted inheritance of MusicI.
will often use only restricted inheritance of \texttt{MusicI}.
\section{To find rules in the resource grammar library}
\subsection{Inflection paradigms}
Inflection paradigms are defined separately for each language L
in the module ParadigmsL. To test them, the command cc (= compute\_concrete)
Inflection paradigms are defined separately for each language \textit{L}
in the module \texttt{Paradigms}\textit{L}. To test them, the command \texttt{cc} (= \texttt{compute\_concrete})
can be used:
\begin{verbatim}
@@ -285,22 +309,22 @@ For the sake of convenience, every language implements these four paradigms:
\end{verbatim}
It is often possible to initialize a lexicon by just using these functions,
and later revise it by using the more involved paradigms. For instance, in
German we cannot use regN "Lied" for Song, because the result would be a
Masculine noun with the plural form "Liede". The individual Paradigms modules
German we cannot use \texttt{regN "Lied"} for \texttt{Song}, because the result would be a
Masculine noun with the plural form \texttt{"Liede"}. The individual \texttt{Paradigms} modules
tell what cases are covered by the regular heuristics.
As a limiting case, one could even initialize the lexicon for a new language
by copying the English (or some other already existing) lexicon. This will
produce language with correct grammar but content words directly borrowed from
produce language with correct grammar but with content words directly borrowed from
English.
\subsection{Syntax rules}
Syntax rules should be looked for in the abstract modules defining the
API. There are around 10 such modules, each defining constructors for
a group of one or more related categories. For instance, the module
Noun defines how to construct common nouns, noun phrases, and determiners.
\texttt{Noun} defines how to construct common nouns, noun phrases, and determiners.
Thus the proper place to find out how nouns are modified with adjectives
is Noun, because the result of the construction is again a common noun.
is \texttt{Noun}, because the result of the construction is again a common noun.
Browsing the libraries is helped by the gfdoc-generated HTML pages.
However, this is still not easy, and the most efficient way is
@@ -347,13 +371,13 @@ which uses ParadigmsIta.regGenN.
\subsection{Example-based grammar writing}
The technique of parsing with the resource grammar can be used in GF source files,
endowed with the suffix .gfe ("GF examples"). The suffix tells GF to preprocess
endowed with the suffix \texttt{.gfe} ("GF examples"). The suffix tells GF to preprocess
the file by replacing all expressions of the form
\begin{verbatim}
in Module.Cat "example string"
\end{verbatim}
by the syntax trees obtained by parsing "example string" in Cat in Module.
by the syntax trees obtained by parsing "example string" in \texttt{Cat} in \texttt{Module}.
For instance,
\begin{verbatim}
@@ -378,7 +402,7 @@ However, the technique of example-based grammar writing has some limitations:
it may not be the intended one. The other parses are shown in a comment, from
where they must/can be picked manually.
\item Lexicality. The arguments of a function must be atomic identifiers, and are thus
not available for categories that have no lexical items. For instance, the PropKind
not available for categories that have no lexical items. For instance, the \texttt{PropKind}
rule above gives the result
\begin{verbatim}
lin
@@ -391,7 +415,7 @@ all those categories that can be used as arguments, for instance,
cat_CN : CN ;
old_AP : AP ;
\end{verbatim}
and then use this lexicon instead of the standard one included in Lang.
and then use this lexicon instead of the standard one included in \texttt{Lang}.
\end{itemize}
\subsection{Special-purpose APIs}
@@ -407,8 +431,9 @@ develop their own macro packages. The same applies to GF resource grammars:
the application grammarian might not need all the choises that the resource
provides, but would prefer less writing and higher-level programming.
To this end, application grammarians may want to write their own views on the
resource grammar. An example of this is already provided, in mathematical/Predication.
Instead of the NP-VP structure, it permits clause construction directly from
resource grammar. An example of this is already provided, in
\texttt{mathematical/Predication}.
Instead of the \texttt{NP-VP} structure, it permits clause construction directly from
verbs and adjectives and their arguments:
\begin{verbatim}
@@ -419,7 +444,7 @@ verbs and adjectives and their arguments:
predA : A -> NP -> Cl ; -- "x is even"
predA2 : A2 -> NP -> NP -> Cl ; -- "x is divisible by y"
\end{verbatim}
The implementation of this module is the functor PredicationI:
The implementation of this module is the functor \texttt{PredicationI}:
\begin{verbatim}
predV v x = PredVP x (UseV v) ;
@@ -429,27 +454,27 @@ The implementation of this module is the functor PredicationI:
predA a x = PredVP x (UseComp (CompAP (PositA a))) ;
predA2 a x y = PredVP x (UseComp (CompAP (ComplA2 a y))) ;
\end{verbatim}
Of course, Predication can be opened together with Grammar, but using
Of course, \texttt{Predication} can be opened together with \texttt{Grammar}, but using
the resulting grammar for parsing can be frustrating, since having both
ways of building clauses simultaneously available will produce spurious
ambiguities. Using Predication without Verb for parsing is a better idea,
since parsing is also made more efficient without the VP category.
ambiguities. Using \texttt{Predication} without \texttt{Verb} for parsing is a better idea,
since parsing is also made more efficient without rules for the \texttt{VP} category.
The use of special-purpose APIs is to some extent to be seen as an alternative
The use of special-purpose APIs is to some extent just an alternative
to grammar writing by parsing, and its importance may decrease as parsing
with the resource grammars gets more efficient.
with resource grammars gets more efficient.
\section{Overview of syntactic structures}
\subsection{Texts. phrases, and utterances}
The outermost linguistic structure is Text. Texts are composed
from Phrases followed by punctuation marks - either of ".", "?" or
The outermost linguistic structure is \texttt{Text}. \texttt{Text}s are composed
from Phrases (\texttt{Phr}) followed by punctuation marks - either of ".", "?" or
"!" (with their proper variants in Spanish and Arabic). Here is an
example of a Text.
example of a \texttt{Text} string.
\begin{verbatim}
John walks. Why? He doesn't want to sleep!
\end{verbatim}
Phrases are mostly built from Utterances, which in turn are
Phrases are mostly built from Utterances (\texttt{Utt}), which in turn are
declarative sentences, questions, or imperatives - but there
are also "one-word utterances" consisting of noun phrases
or other subsentential phrases. Some Phrases are atomic,
@@ -478,8 +503,8 @@ a Phrase is an Utterance with an optional leading conjunction ("but")
and an optional tailing vocative ("John", "please").
\subsection{Sentences and clauses}
The richest of the categories below Utterance is S, Sentence. A Sentence
is formed from a Clause, by fixing its Tense, Anteriority, and Polarity.
The richest of the categories below Utterance is \texttt{S}, Sentence. A Sentence
is formed from a Clause (\texttt{Cl}), by fixing its Tense, Anteriority, and Polarity.
The difference between Sentence and Clause is thus also rather technical.
For example, each of the following strings has a distinct syntax tree
in the category Sentence:
@@ -549,14 +574,14 @@ many constructors:
The linguistic phenomena mostly discussed in both traditional grammars and modern
syntax belong to the level of Clauses, that is, lines 9-13, and occasionally
to Sentences, lines 5-13. At this level, the major categories are
NP (Noun Phrase) and VP (Verb Phrase). A Clause typically consists of just an
NP and a VP. The internal structure of both NP and VP can be very complex,
and these categories are mutually recursive: not only can a VP contain an NP,
\texttt{NP} (Noun Phrase) and \texttt{VP} (Verb Phrase). A Clause typically consists of just an
\texttt{NP} and a \texttt{VP}. The internal structure of both \texttt{NP} and \texttt{VP} can be very complex,
and these categories are mutually recursive: not only can a \texttt{VP} contain an \texttt{NP},
\begin{verbatim}
[VP loves [NP Mary]]
\end{verbatim}
but an NP can also contain a VP
but also an \texttt{NP} can contain a \texttt{VP}
\begin{verbatim}
[NP every man [RS who [VP walks]]]
@@ -567,32 +592,34 @@ a GF syntax tree, but still a useful device of exposition).
Most of the resource modules thus define functions that are used inside
NPs and VPs. Here is a brief overview:
Noun: How to construct NPs. The main three mechanisms
\textbf{Noun}. How to construct NPs. The main three mechanisms
for constructing NPs are
\begin{itemize}
\item from proper names: John
\item from pronouns: we
\item from common nouns by determiners: this man
\item from proper names: "John"
\item from pronouns: "we"
\item from common nouns by determiners: "this man"
\end{itemize}
The Noun module also defines the construction of common nouns. The most frequent ways are
The \texttt{Noun} module also defines the construction of common nouns.
The most frequent ways are
\begin{itemize}
\item lexical noun items: man
\item adjectival modification: old man
\item relative clause modification: man who sleeps
\item application of relational nouns: successor of the number
\item lexical noun items: "man"
\item adjectival modification: "old man"
\item relative clause modification: "man who sleeps"
\item application of relational nouns: "successor of the number"
\end{itemize}
Verb: How to construct VPs. The main mechanism is verbs with their arguments, for instance,
\textbf{Verb}.
How to construct VPs. The main mechanism is verbs with their arguments, for instance,
\begin{itemize}
\item one-place verbs: walks
\item two-place verbs: loves Mary
\item three-place verbs: gives her a kiss
\item sentence-complement verbs: says that it is cold
\item VP-complement verbs: wants to give her a kiss
\item one-place verbs: "walks"
\item two-place verbs: "loves Mary"
\item three-place verbs: "gives her a kiss"
\item sentence-complement verbs: "says that it is cold"
\item VP-complement verbs: "wants to give her a kiss"
\end{itemize}
A special verb is the copula, "be" in English but not even realized
@@ -600,22 +627,24 @@ by a verb in all languages.
A copula can take different kinds of complement:
\begin{itemize}
\item an adjectival phrase: (John is) old
\item an adverb: (John is) here
\item a noun phrase: (John is) a man
\item an adjectival phrase: "(John is) old"
\item an adverb: "(John is) here"
\item a noun phrase: "(John is) a man"
\end{itemize}
Adjective: How to constuct APs. The main ways are
\textbf{Adjective}.
How to constuct \texttt{AP}s. The main ways are
\begin{itemize}
\item positive forms of adjectives: old
\item comparative forms with object of comparison: older than John
\item positive forms of adjectives: "old"
\item comparative forms with object of comparison: "older than John"
\end{itemize}
Adverb: How to construct Advs. The main ways are
\textbf{Adverb}.
How to construct \texttt{Adv}s. The main ways are
\begin{itemize}
\item from adjectives: slowly
\item from adjectives: "slowly"
\end{itemize}
\subsection{Modules and their names}
@@ -624,27 +653,28 @@ and they can be roughly classified by the "level" or "size" of expressions that
formed in them:
\begin{itemize}
\item Larger than sentence: Text, Phrase
\item Same level as sentence: Sentence, Question, Relative
\item Parts of sentence: Adjective, Adverb, Noun, Verb
\item Cross-cut: Conjunction
\item Larger than sentence: \texttt{Text}, \texttt{Phrase}
\item Same level as sentence: \texttt{Sentence}, \texttt{Question}, \texttt{Relative}
\item Parts of sentence: \texttt{Adjective}, \texttt{Adverb}, \texttt{Noun}, \texttt{Verb}
\item Cross-cut (coordination): \texttt{Conjunction}
\end{itemize}
Because of mutual recursion such as in embedded sentences, this classification is
not a complete order. However, no mutual dependence is needed between the
modules in a formal sense - they can all be compiled separately. This is due
to the module Cat, which defines the type system common to the other modules.
For instance, the types NP and VP are defined in Cat, and the module Verb only
needs to know what is given in Cat, not what is given in Noun. To implement
to the module \texttt{Cat}, which defines the type system common to the other modules.
For instance, the types \texttt{NP} and \texttt{VP} are defined in \texttt{Cat}, and the module \texttt{Verb} only
needs to know what is given in \texttt{Cat}, not what is given in \texttt{Noun}. To implement
a rule such as
\begin{verbatim}
Verb.ComplV2 : V2 -> NP -> VP
\end{verbatim}
it is enough to know the linearization type of NP (as well as those of V2 and VP, all
given in Cat). It is not necessary to know what
ways there are to build NPs (given in Noun), since all these ways must
conform to the linearization type defined in Cat. Thus the format of
it is enough to know the linearization type of \texttt{NP}
(as well as those of \texttt{V2} and \texttt{VP}, all
given in \texttt{Cat}). It is not necessary to know what
ways there are to build \texttt{NP}s (given in \texttt{Noun}), since all these ways must
conform to the linearization type defined in \texttt{Cat}. Thus the format of
category-specific modules is as follows:
\begin{verbatim}
@@ -654,32 +684,33 @@ category-specific modules is as follows:
\end{verbatim}
\subsection{Top-level grammar and lexicon}
The module Grammar collects all the category-specific modules into
The module \texttt{Grammar} collects all the category-specific modules into
a complete grammar:
\begin{verbatim}
abstract Grammar =
Adjective, Noun, Verb, ..., Structural, Idiom
\end{verbatim}
The module Structural is a lexicon of structural words (function words),
The module \texttt{Structural} is a lexicon of structural words (function words),
such as determiners.
The module Idiom is a collection of idiomatic structures whose
The module \texttt{Idiom} is a collection of idiomatic structures whose
implementation is very language-dependent. An example is existential
structures ("there is", "es gibt", "il y a", etc).
The module Lang combines Grammar with a Lexicon of ca. 350 content words:
The module \texttt{Lang} combines \texttt{Grammar} with a \texttt{Lexicon} of ca. 350 content words:
\begin{verbatim}
abstract Lang = Grammar, Lexicon
\end{verbatim}
Using Lang instead of Grammar as a library may give the advantage of prociding
Using \texttt{Lang} instead of \texttt{Grammar} as a library may give
for free some words needed in an application. But its main purpose is to
help testing the resource library. It does not seem possible to maintain
a general-purpose multilingual lexicon, and this is the form that the module
Lexicon has.
\texttt{Lexicon} has.
\subsection{Language-specific syntactic structures}
The API collected in Grammar has been designed to be implementable for
The API collected in \texttt{Grammar} has been designed to be implementable for
all languages in the resource package. It does contain some rules that
are strange or superfluous in some languages; for instance, the distinction
between definite and indefinite articles does not apply to Finnish and Russian.
@@ -693,27 +724,28 @@ rules. The top level of each languages looks as follows (with English as example
\begin{verbatim}
abstract English = Grammar, ExtraEngAbs, DictEngAbs
\end{verbatim}
where ExtraEngAbs is a collection of syntactic structures specific to English,
and DictEngAbs is an English dictionary (at the moment, it consists of IrregEngAbs,
where \texttt{ExtraEngAbs} is a collection of syntactic structures specific to English,
and \texttt{DictEngAbs} is an English dictionary (at the moment, it consists of \texttt{IrregEngAbs},
the irregular verbs of English). Each of these language-specific grammars has
the potential to grow into a full-scale grammar of the language. These grammar
can also be used as libraries, but the possibility of using functors is lost.
To give a better overview of language-specific structures, modules like ExtraEngAbs
are built from a language-independent module ExtraAbs by restricted inheritance:
To give a better overview of language-specific structures, modules like \texttt{ExtraEngAbs}
are built from a language-independent module \texttt{ExtraAbs} by restricted inheritance:
\begin{verbatim}
abstract ExtraEngAbs = Extra [f,g,...]
\end{verbatim}
Thus any category and function in Extra may be shared by a subset of all
languages. One can see this set-up as a matrix, which tells what Extra structures
are implemented in what languages. For the common API in Grammar, the matrix
Thus any category and function in \texttt{Extra} may be shared by a subset of all
languages. One can see this set-up as a matrix, which tells what \texttt{Extra} structures
are implemented in what languages. For the common API in \texttt{Grammar}, the matrix
is filled with 1's (everything is implemented in every language).
Language-specific extensions and the use of restricted
inheritance is a recent addition to the resource grammar library, and
has only been exploited in a very small scale so far.
\section{API Documentation}
\subsection{Top-level modules}

246
doc/resource.txt
View File

@@ -15,7 +15,7 @@ module system, knowledge that can be acquired e.g. from the GF
tutorial. We start with an introduction to the library, and proceed to
details with the aim of covering all that one needs to know
in order to use the library.
How to write one's own resource grammar (i.e. implement the API for
How to write one's own resource grammar (i.e. to implement the API for
a new language), is covered by a separate Resource-HOWTO document.
@@ -26,15 +26,32 @@ The GF Resource Grammar Library contains grammar rules for
is to make these rules available for application programmers,
who can thereby concentrate on the semantic and stylistic
aspects of their grammars, without having to think about
grammaticality. The level of a typical application grammarian
is skilled programmer, without knowledge linguistics, but with
grammaticality. The targeted level of application grammarians
is skilled programmer without knowledge linguistics, but with
a good knowledge of the target languages. Such a combination of
skilles is typical of a programmer who wants to localize a piece
of software to a new language.
skills is typical of programmers who want to localize
software to new languages.
To give an example, an application dealing with
music players may have a semantical category ``Kind``, examples
of Kinds being Song and Artist. In German, for instance, Song
The current resource languages are
-``Dan``ish
-``Eng``lish
-``Fin``nish
-``Fre``nch
-``Ger``man
-``Ita``lian
-``Nor``wegian
-``Rus``sian
-``Spa``nish
-``Swe``dish
The first three letters (``Dan`` etc) are used in grammar module names.
To give an example application, consider
music playing devices. In the application,
we may have a semantical category ``Kind``, examples
of ``Kind``s being ``Song`` and ``Artist``. In German, for instance, ``Song``
is linearized into the noun "Lied", but knowing this is not
enough to make the application work, because the noun must be
produced in both singular and plural, and in four different
@@ -45,13 +62,13 @@ write
```
and the eight forms are correctly generated. The resource grammar
library contains a complete set of inflectional paradigms (such as
regN2 here), enabling the definition of any lexical items.
``regN2`` here), enabling the definition of any lexical items.
The resource grammar library is not only about inflectional paradigms - it
also has syntax rules. The music player application
might also want to modify songs with properties, such as "American",
"old", "good". The German grammar for adjectival modifications is
particularly complex, because the adjectives have to agree in gender,
particularly complex, because adjectives have to agree in gender,
number, and case, and also depend on what determiner is used
("ein Amerikanisches Lied" vs. "das Amerikanische Lied"). All this
variation is taken care of by the resource grammar function
@@ -69,18 +86,18 @@ given that
lincat Kind = CN
```
The resource library API is devided into language-specific and language-independet
parts. To put is roughly,
- lexicon is language-specific
- syntax is language-independent
parts. To put it roughly,
- the lexicon API is language-specific
- the syntax API is language-independent
Thus, to render the above example in French instead of German, we need to
pick a different linearization of Song,
pick a different linearization of ``Song``,
```
lin Song = regGenN "chanson" feminine
```
But to linearize PropKind, we can use the very same rule as in German.
The resource function AdjCN has different implementations in the two
But to linearize ``PropKind``, we can use the very same rule as in German.
The resource function ``AdjCN`` has different implementations in the two
languages, but the application programmer need not care about the difference.
@@ -101,7 +118,7 @@ The abstract syntax defines a "domain ontology":
}
```
The concrete syntax is defined independently of language, by opening
two interfaces: the resource Grammar and an application lexicon.
two interfaces: the resource ``Grammar`` and an application lexicon.
```
incomplete concrete MusicI of Music = open Grammar, MusicLex in {
lincat
@@ -113,8 +130,8 @@ two interfaces: the resource Grammar and an application lexicon.
American = PositA american_A ;
}
```
The application lexicon MusicLex has an abstract syntax, that extends
the resource category system Cat.
The application lexicon ``MusicLex`` has an abstract syntax that extends
the resource category system ``Cat``.
```
abstract MusicLex = Cat ** {
fun
@@ -137,8 +154,8 @@ module for that language:
american_A = regA "américain" ;
}
```
The top-level Music grammars are obtained by instantiating the two interfaces
of MusicI:
The top-level ``Music`` grammars are obtained by instantiating the two interfaces
of ``MusicI``:
```
concrete MusicGer of Music = MusicI with
(Grammar = GrammarGer),
@@ -148,12 +165,19 @@ of MusicI:
(Grammar = GrammarFre),
(MusicLex = MusicLexFre) ;
```
To localize the system to a new language, all that is needed is two modules,
one implementing MusicLex and the other instantiating Music. The latter is
Both of these files can use the same ``path``, defined as
```
--# -path=.:present:prelude
```
The ``present`` category contains the compiled resources, restricted to
present tense; ``alltenses`` has the full resources.
To localize the music player system to a new language, all that is needed is two modules,
one implementing ``MusicLex`` and the other instantiating ``Music``. The latter is
completely trivial, whereas the former one involves the choice of correct
vocabulary and inflectional paradigms. For instance, Finnish is added as follows:
```
concrete MusicLexFin of MusicLex = CatFre ** open ParadigmsFin in {
concrete MusicLexFin of MusicLex = CatFin ** open ParadigmsFin in {
lin
song_N = regN "kappale" ;
american_A = regA "amerikkalainen" ;
@@ -171,7 +195,7 @@ for the sake of argument, that adjectival modification does not sound good in
English, but that a relative clause would be preferrable. One can then start as
before,
```
concrete MusicLexEng of MusicLex = CatFre ** open ParadigmsEng in {
concrete MusicLexEng of MusicLex = CatEng ** open ParadigmsEng in {
lin
song_N = regN "song" ;
american_A = regA "American" ;
@@ -181,9 +205,10 @@ before,
(Grammar = GrammarEng),
(MusicLex = MusicLexEng) ;
```
The module MusicEng0 would not be used on the top level, however, but
The module ``MusicEng0`` would not be used on the top level, however, but
another module would be built on top of it, with a restricted import from
MusicEng0. MusicEng inherits everything from MusicEng0 except PropKind, and
``MusicEng0``. ``MusicEng`` inherits everything from ``MusicEng0``
except ``PropKind``, and
gives its own definition of this function:
```
concrete MusicEng of Music = MusicEng0 - [PropKind] ** open GrammarEng in {
@@ -217,13 +242,13 @@ All of these problems should be solved in application grammars.
The task of resource grammars is just to take care of low-level linguistic
details such as inflection, agreement, and word order.
For the same reasons, resource grammars are not adequate for parsing.
It is for the same reasons that resource grammars are not adequate for translation.
That the syntax API is implemented for different languages of course makes
it possible to translate via it - but there is no guarantee of translation
equivalence. Of course, the use of parametrized implementations such as MusicI
equivalence. Of course, the use of parametrized implementations such as ``MusicI``
above only extends to those cases where the syntax API does give translation
equivalence - but this must be seen as a limiting case, and real applications
will often use only restricted inheritance of MusicI.
will often use only restricted inheritance of ``MusicI``.
@@ -231,8 +256,8 @@ will often use only restricted inheritance of MusicI.
===Inflection paradigms===
Inflection paradigms are defined separately for each language L
in the module ParadigmsL. To test them, the command cc (= compute_concrete)
Inflection paradigms are defined separately for each language //L//
in the module ``Paradigms``//L//. To test them, the command ``cc`` (= ``compute_concrete``)
can be used:
```
> i -retain german/ParadigmsGer.gf
@@ -266,13 +291,13 @@ For the sake of convenience, every language implements these four paradigms:
```
It is often possible to initialize a lexicon by just using these functions,
and later revise it by using the more involved paradigms. For instance, in
German we cannot use regN "Lied" for Song, because the result would be a
Masculine noun with the plural form "Liede". The individual Paradigms modules
German we cannot use ``regN "Lied"`` for ``Song``, because the result would be a
Masculine noun with the plural form ``"Liede"``. The individual ``Paradigms`` modules
tell what cases are covered by the regular heuristics.
As a limiting case, one could even initialize the lexicon for a new language
by copying the English (or some other already existing) lexicon. This will
produce language with correct grammar but content words directly borrowed from
produce language with correct grammar but with content words directly borrowed from
English.
@@ -282,9 +307,9 @@ English.
Syntax rules should be looked for in the abstract modules defining the
API. There are around 10 such modules, each defining constructors for
a group of one or more related categories. For instance, the module
Noun defines how to construct common nouns, noun phrases, and determiners.
``Noun`` defines how to construct common nouns, noun phrases, and determiners.
Thus the proper place to find out how nouns are modified with adjectives
is Noun, because the result of the construction is again a common noun.
is ``Noun``, because the result of the construction is again a common noun.
Browsing the libraries is helped by the gfdoc-generated HTML pages.
However, this is still not easy, and the most efficient way is
@@ -330,12 +355,12 @@ which uses ParadigmsIta.regGenN.
===Example-based grammar writing===
The technique of parsing with the resource grammar can be used in GF source files,
endowed with the suffix .gfe ("GF examples"). The suffix tells GF to preprocess
endowed with the suffix ``.gfe`` ("GF examples"). The suffix tells GF to preprocess
the file by replacing all expressions of the form
```
in Module.Cat "example string"
```
by the syntax trees obtained by parsing "example string" in Cat in Module.
by the syntax trees obtained by parsing "example string" in ``Cat`` in ``Module``.
For instance,
```
lin IamHungry =
@@ -356,7 +381,7 @@ However, the technique of example-based grammar writing has some limitations:
it may not be the intended one. The other parses are shown in a comment, from
where they must/can be picked manually.
- Lexicality. The arguments of a function must be atomic identifiers, and are thus
not available for categories that have no lexical items. For instance, the PropKind
not available for categories that have no lexical items. For instance, the ``PropKind``
rule above gives the result
```
lin
@@ -369,7 +394,7 @@ all those categories that can be used as arguments, for instance,
cat_CN : CN ;
old_AP : AP ;
```
and then use this lexicon instead of the standard one included in Lang.
and then use this lexicon instead of the standard one included in ``Lang``.
@@ -387,8 +412,9 @@ develop their own macro packages. The same applies to GF resource grammars:
the application grammarian might not need all the choises that the resource
provides, but would prefer less writing and higher-level programming.
To this end, application grammarians may want to write their own views on the
resource grammar. An example of this is already provided, in mathematical/Predication.
Instead of the NP-VP structure, it permits clause construction directly from
resource grammar. An example of this is already provided, in
``mathematical/Predication``.
Instead of the ``NP-VP`` structure, it permits clause construction directly from
verbs and adjectives and their arguments:
```
predV : V -> NP -> Cl ; -- "x converges"
@@ -398,7 +424,7 @@ verbs and adjectives and their arguments:
predA : A -> NP -> Cl ; -- "x is even"
predA2 : A2 -> NP -> NP -> Cl ; -- "x is divisible by y"
```
The implementation of this module is the functor PredicationI:
The implementation of this module is the functor ``PredicationI``:
```
predV v x = PredVP x (UseV v) ;
predV2 v x y = PredVP x (ComplV2 v y) ;
@@ -407,15 +433,15 @@ The implementation of this module is the functor PredicationI:
predA a x = PredVP x (UseComp (CompAP (PositA a))) ;
predA2 a x y = PredVP x (UseComp (CompAP (ComplA2 a y))) ;
```
Of course, Predication can be opened together with Grammar, but using
Of course, ``Predication`` can be opened together with ``Grammar``, but using
the resulting grammar for parsing can be frustrating, since having both
ways of building clauses simultaneously available will produce spurious
ambiguities. Using Predication without Verb for parsing is a better idea,
since parsing is also made more efficient without the VP category.
ambiguities. Using ``Predication`` without ``Verb`` for parsing is a better idea,
since parsing is also made more efficient without rules for the ``VP`` category.
The use of special-purpose APIs is to some extent to be seen as an alternative
The use of special-purpose APIs is to some extent just an alternative
to grammar writing by parsing, and its importance may decrease as parsing
with the resource grammars gets more efficient.
with resource grammars gets more efficient.
@@ -425,14 +451,14 @@ with the resource grammars gets more efficient.
===Texts. phrases, and utterances===
The outermost linguistic structure is Text. Texts are composed
from Phrases followed by punctuation marks - either of ".", "?" or
The outermost linguistic structure is ``Text``. ``Text``s are composed
from Phrases (``Phr``) followed by punctuation marks - either of ".", "?" or
"!" (with their proper variants in Spanish and Arabic). Here is an
example of a Text.
example of a ``Text`` string.
```
John walks. Why? He doesn't want to sleep!
```
Phrases are mostly built from Utterances, which in turn are
Phrases are mostly built from Utterances (``Utt``), which in turn are
declarative sentences, questions, or imperatives - but there
are also "one-word utterances" consisting of noun phrases
or other subsentential phrases. Some Phrases are atomic,
@@ -461,8 +487,8 @@ and an optional tailing vocative ("John", "please").
===Sentences and clauses===
The richest of the categories below Utterance is S, Sentence. A Sentence
is formed from a Clause, by fixing its Tense, Anteriority, and Polarity.
The richest of the categories below Utterance is ``S``, Sentence. A Sentence
is formed from a Clause (``Cl``), by fixing its Tense, Anteriority, and Polarity.
The difference between Sentence and Clause is thus also rather technical.
For example, each of the following strings has a distinct syntax tree
in the category Sentence:
@@ -530,13 +556,13 @@ many constructors:
The linguistic phenomena mostly discussed in both traditional grammars and modern
syntax belong to the level of Clauses, that is, lines 9-13, and occasionally
to Sentences, lines 5-13. At this level, the major categories are
NP (Noun Phrase) and VP (Verb Phrase). A Clause typically consists of just an
NP and a VP. The internal structure of both NP and VP can be very complex,
and these categories are mutually recursive: not only can a VP contain an NP,
``NP`` (Noun Phrase) and ``VP`` (Verb Phrase). A Clause typically consists of just an
``NP`` and a ``VP``. The internal structure of both ``NP`` and ``VP`` can be very complex,
and these categories are mutually recursive: not only can a ``VP`` contain an ``NP``,
```
[VP loves [NP Mary]]
```
but an NP can also contain a VP
but also an ``NP`` can contain a ``VP``
```
[NP every man [RS who [VP walks]]]
```
@@ -546,43 +572,47 @@ a GF syntax tree, but still a useful device of exposition).
Most of the resource modules thus define functions that are used inside
NPs and VPs. Here is a brief overview:
Noun: How to construct NPs. The main three mechanisms
**Noun**. How to construct NPs. The main three mechanisms
for constructing NPs are
- from proper names: John
- from pronouns: we
- from common nouns by determiners: this man
- from proper names: "John"
- from pronouns: "we"
- from common nouns by determiners: "this man"
The Noun module also defines the construction of common nouns. The most frequent ways are
- lexical noun items: man
- adjectival modification: old man
- relative clause modification: man who sleeps
- application of relational nouns: successor of the number
The ``Noun`` module also defines the construction of common nouns.
The most frequent ways are
- lexical noun items: "man"
- adjectival modification: "old man"
- relative clause modification: "man who sleeps"
- application of relational nouns: "successor of the number"
Verb: How to construct VPs. The main mechanism is verbs with their arguments, for instance,
- one-place verbs: walks
- two-place verbs: loves Mary
- three-place verbs: gives her a kiss
- sentence-complement verbs: says that it is cold
- VP-complement verbs: wants to give her a kiss
**Verb**.
How to construct VPs. The main mechanism is verbs with their arguments, for instance,
- one-place verbs: "walks"
- two-place verbs: "loves Mary"
- three-place verbs: "gives her a kiss"
- sentence-complement verbs: "says that it is cold"
- VP-complement verbs: "wants to give her a kiss"
A special verb is the copula, "be" in English but not even realized
by a verb in all languages.
A copula can take different kinds of complement:
- an adjectival phrase: (John is) old
- an adverb: (John is) here
- a noun phrase: (John is) a man
- an adjectival phrase: "(John is) old"
- an adverb: "(John is) here"
- a noun phrase: "(John is) a man"
Adjective: How to constuct APs. The main ways are
- positive forms of adjectives: old
- comparative forms with object of comparison: older than John
**Adjective**.
How to constuct ``AP``s. The main ways are
- positive forms of adjectives: "old"
- comparative forms with object of comparison: "older than John"
Adverb: How to construct Advs. The main ways are
- from adjectives: slowly
**Adverb**.
How to construct ``Adv``s. The main ways are
- from adjectives: "slowly"
@@ -591,26 +621,27 @@ Adverb: How to construct Advs. The main ways are
The resource modules are named after the kind of phrases that are constructed in them,
and they can be roughly classified by the "level" or "size" of expressions that are
formed in them:
- Larger than sentence: Text, Phrase
- Same level as sentence: Sentence, Question, Relative
- Parts of sentence: Adjective, Adverb, Noun, Verb
- Cross-cut: Conjunction
- Larger than sentence: ``Text``, ``Phrase``
- Same level as sentence: ``Sentence``, ``Question``, ``Relative``
- Parts of sentence: ``Adjective``, ``Adverb``, ``Noun``, ``Verb``
- Cross-cut (coordination): ``Conjunction``
Because of mutual recursion such as in embedded sentences, this classification is
not a complete order. However, no mutual dependence is needed between the
modules in a formal sense - they can all be compiled separately. This is due
to the module Cat, which defines the type system common to the other modules.
For instance, the types NP and VP are defined in Cat, and the module Verb only
needs to know what is given in Cat, not what is given in Noun. To implement
to the module ``Cat``, which defines the type system common to the other modules.
For instance, the types ``NP`` and ``VP`` are defined in ``Cat``, and the module ``Verb`` only
needs to know what is given in ``Cat``, not what is given in ``Noun``. To implement
a rule such as
```
Verb.ComplV2 : V2 -> NP -> VP
```
it is enough to know the linearization type of NP (as well as those of V2 and VP, all
given in Cat). It is not necessary to know what
ways there are to build NPs (given in Noun), since all these ways must
conform to the linearization type defined in Cat. Thus the format of
it is enough to know the linearization type of ``NP``
(as well as those of ``V2`` and ``VP``, all
given in ``Cat``). It is not necessary to know what
ways there are to build ``NP``s (given in ``Noun``), since all these ways must
conform to the linearization type defined in ``Cat``. Thus the format of
category-specific modules is as follows:
```
abstract Adjective = Cat ** {...}
@@ -621,33 +652,34 @@ category-specific modules is as follows:
===Top-level grammar and lexicon===
The module Grammar collects all the category-specific modules into
The module ``Grammar`` collects all the category-specific modules into
a complete grammar:
```
abstract Grammar =
Adjective, Noun, Verb, ..., Structural, Idiom
```
The module Structural is a lexicon of structural words (function words),
The module ``Structural`` is a lexicon of structural words (function words),
such as determiners.
The module Idiom is a collection of idiomatic structures whose
The module ``Idiom`` is a collection of idiomatic structures whose
implementation is very language-dependent. An example is existential
structures ("there is", "es gibt", "il y a", etc).
The module Lang combines Grammar with a Lexicon of ca. 350 content words:
The module ``Lang`` combines ``Grammar`` with a ``Lexicon`` of ca. 350 content words:
```
abstract Lang = Grammar, Lexicon
```
Using Lang instead of Grammar as a library may give the advantage of prociding
Using ``Lang`` instead of ``Grammar`` as a library may give
for free some words needed in an application. But its main purpose is to
help testing the resource library. It does not seem possible to maintain
a general-purpose multilingual lexicon, and this is the form that the module
Lexicon has.
``Lexicon`` has.
===Language-specific syntactic structures===
The API collected in Grammar has been designed to be implementable for
The API collected in ``Grammar`` has been designed to be implementable for
all languages in the resource package. It does contain some rules that
are strange or superfluous in some languages; for instance, the distinction
between definite and indefinite articles does not apply to Finnish and Russian.
@@ -660,20 +692,20 @@ rules. The top level of each languages looks as follows (with English as example
```
abstract English = Grammar, ExtraEngAbs, DictEngAbs
```
where ExtraEngAbs is a collection of syntactic structures specific to English,
and DictEngAbs is an English dictionary (at the moment, it consists of IrregEngAbs,
where ``ExtraEngAbs`` is a collection of syntactic structures specific to English,
and ``DictEngAbs`` is an English dictionary (at the moment, it consists of ``IrregEngAbs``,
the irregular verbs of English). Each of these language-specific grammars has
the potential to grow into a full-scale grammar of the language. These grammar
can also be used as libraries, but the possibility of using functors is lost.
To give a better overview of language-specific structures, modules like ExtraEngAbs
are built from a language-independent module ExtraAbs by restricted inheritance:
To give a better overview of language-specific structures, modules like ``ExtraEngAbs``
are built from a language-independent module ``ExtraAbs`` by restricted inheritance:
```
abstract ExtraEngAbs = Extra [f,g,...]
```
Thus any category and function in Extra may be shared by a subset of all
languages. One can see this set-up as a matrix, which tells what Extra structures
are implemented in what languages. For the common API in Grammar, the matrix
Thus any category and function in ``Extra`` may be shared by a subset of all
languages. One can see this set-up as a matrix, which tells what ``Extra`` structures
are implemented in what languages. For the common API in ``Grammar``, the matrix
is filled with 1's (everything is implemented in every language).
Language-specific extensions and the use of restricted

9
doc/tutorial/music/Music.gf Normal file
View File

@@ -0,0 +1,9 @@
abstract Music = {
cat
Kind ;
Property ;
fun
PropKind : Kind -> Property -> Kind ;
Song : Kind ;
American : Property ;
}

7
doc/tutorial/music/MusicEng.gf Normal file
View File

@@ -0,0 +1,7 @@
--# -path=.:present:prelude
concrete MusicEng of Music = MusicEng0 - [PropKind] ** open GrammarEng in {
lin
PropKind k p =
RelCN k (UseRCl TPres ASimul PPos (RelVP IdRP (UseComp (CompAP p)))) ;
}

3
doc/tutorial/music/MusicEng0.gf Normal file
View File

@@ -0,0 +1,3 @@
concrete MusicEng0 of Music = MusicI with
(Grammar = GrammarEng),
(MusicLex = MusicLexEng) ;

5
doc/tutorial/music/MusicFin.gf Normal file
View File

@@ -0,0 +1,5 @@
--# -path=.:present:prelude
concrete MusicFin of Music = MusicI with
(Grammar = GrammarFin),
(MusicLex = MusicLexFin) ;

6
doc/tutorial/music/MusicFre.gf Normal file
View File

@@ -0,0 +1,6 @@
--# -path=.:present:prelude
concrete MusicFre of Music = MusicI with
(Grammar = GrammarFre),
(MusicLex = MusicLexFre) ;

6
doc/tutorial/music/MusicGer.gf Normal file
View File

@@ -0,0 +1,6 @@
--# -path=.:present:prelude
concrete MusicGer of Music = MusicI with
(Grammar = GrammarGer),
(MusicLex = MusicLexGer) ;

9
doc/tutorial/music/MusicI.gf Normal file
View File

@@ -0,0 +1,9 @@
incomplete concrete MusicI of Music = open Grammar, MusicLex in {
lincat
Kind = CN ;
Property = AP ;
lin
PropKind k p = AdjCN p k ;
Song = UseN song_N ;
American = PositA american_A ;
}

5
doc/tutorial/music/MusicLex.gf Normal file
View File

@@ -0,0 +1,5 @@
abstract MusicLex = Cat ** {
fun
song_N : N ;
american_A : A ;
}

5
doc/tutorial/music/MusicLexEng.gf Normal file
View File

@@ -0,0 +1,5 @@
concrete MusicLexEng of MusicLex = CatEng ** open ParadigmsEng in {
lin
song_N = regN "song" ;
american_A = regA "American" ;
}

6
doc/tutorial/music/MusicLexFin.gf Normal file
View File

@@ -0,0 +1,6 @@
concrete MusicLexFin of MusicLex = CatFin ** open ParadigmsFin in {
lin
song_N = regN "kappale" ;
american_A = regA "amerikkalainen" ;
}

5
doc/tutorial/music/MusicLexFre.gf Normal file
View File

@@ -0,0 +1,5 @@
concrete MusicLexFre of MusicLex = CatFre ** open ParadigmsFre in {
lin
song_N = regGenN "chanson" feminine ;
american_A = regA "américain" ;
}

5
doc/tutorial/music/MusicLexGer.gf Normal file
View File

@@ -0,0 +1,5 @@
concrete MusicLexGer of MusicLex = CatGer ** open ParadigmsGer in {
lin
song_N = reg2N "Lied" "Lieder" neuter ;
american_A = regA "amerikanisch" ;
}