gf-core/lib/doc/index.txt

GF Resource Grammar Library v. 1.2
Author: Aarne Ranta <aarne (at) cs.chalmers.se>
Last update: %%date(%c)

% NOTE: this is a txt2tags file.
% Create an html file from this file using:
% txt2tags --toc -thtml index.txt

%!target:html

%!postproc(html): #BCEN <center>
%!postproc(html): #ECEN </center>


#BCEN

[10lang-large.png]

#ECEN


The GF Resource Grammar Library defines the basic grammar of
ten languages: 
Danish, English, Finnish, French, German,
Italian, Norwegian, Russian, Spanish, Swedish.
Still incomplete implementations for Arabic and Catalan are also
included.

**New** in December 2007: Browsing the library by syntax editor 
[directly on the web ../../../demos/resource-api/editor.html].




==Authors==

Inger Andersson and Therese Soderberg (Spanish morphology),
Nicolas Barth and Sylvain Pogodalla (French verb list),
Ali El Dada (Arabic modules),
Magda Gerritsen and Ulrich Real (Russian paradigms and lexicon),
Janna Khegai (Russian modules),
Bjorn Bringert (many Swadesh lexica),
Carlos Gonzalía (Spanish cardinals), 
Harald Hammarström (German morphology),
Patrik Jansson (Swedish cardinals),
Andreas Priesnitz (German lexicon),
Aarne Ranta,
Jordi Saludes (Catalan modules),
Henning Thielemann (German lexicon).


We are grateful for contributions and 
comments to several other people who have used this and 
the previous versions of the resource library, including
Ludmilla Bogavac,
Ana Bove,
David Burke,
Lauri Carlson,
Gloria Casanellas,
Karin Cavallin,
Robin Cooper,
Hans-Joachim Daniels,
Elisabet Engdahl,
Markus Forsberg,
Kristofer Johannisson,
Anni Laine,
Hans Leiß,
Peter Ljunglöf,
Saara Myllyntausta,
Wanjiku Ng'ang'a,
Nadine Perera,
Jordi Saludes.


==License==

The GF Resource Grammar Library is open-source software licensed under
GNU Lesser General Public License (LGPL). See the file [LICENSE ../LICENSE] for more
details.


==Scope==

Coverage, for each language:
- complete morphology
- lexicon of the ca. 100 most important structural words
- test lexicon of ca. 300 content words (rough equivalents in each language)
- list of irregular verbs (separately for each language)
- representative fragment of syntax (cf. CLE (Core Language Engine))
- rather flat semantics (cf. Quasi-Logical Form of CLE)


Organization:
- top-level (API) modules
- Ground API + special-purpose APIs
- "school grammar" concepts rather than advanced linguistic theory


Presentation:
- tool ``gfdoc`` for generating HTML from grammars
- example collections


==Location==

Assuming you have installed the libraries, you will find the precompiled
``gfc`` and ``gfr`` files directly under ``$GF_LIB_PATH``, whose default
value is ``/usr/local/share/GF/``. The precompiled subdirectories are
```
  alltenses
  mathematical
  multimodal
  present
```
Do for instance
```
  cd $GF_LIB_PATH
  gf alltenses/langs.gfcm
    
  > p -cat=S -lang=LangEng "this grammar is too big" | tb
```
For more details, see the [Synopsis synopsis.html].


==Compilation==

If you want to compile the library from scratch, use ``make`` in the root of
the source directory:
```
  cd GF/lib/resource-1.0
  make
```
The ``make`` procedure does not by default make Arabic and Catalan, but you
can uncomment the relevant lines in ``Makefile`` to compile them.


==Encoding==

Finnish, German, Romance, and Scandinavian languages are in isolatin-1.

Arabic and Russian are in UTF-8.

English is in pure ASCII.

The different encodings imply, unfortunately, that it is hard to get
a nice view of all languages simultaneously. The easiest way to achieve this is
to use ``gfeditor``, which automatically converts grammars to UTF-8.


==Using the resource as library==

This API is accessible by both ``present`` and ``alltenses``. The modules you most often need are
- ``Syntax``, the interface to syntactic structures
- ``Syntax``//L//, the implementations of ``Syntax`` for each language //L//
- ``Paradigms``//L//, the morphological paradigms for each language //L//


The [Synopsis synopsis.html] gives examples on the typical usage of these 
modules.


==Using the resource as top level grammar==

The following modules can be used for parsing and linearization. They are accessible from both
``present`` and ``alltenses``.
- ``Lang``//L// for each language //L//, implementing a common abstract syntax ``Lang``
- ``Danish``, ``English``, etc, implementing ``Lang`` with language-specific extensions


In addition, there is in both ``present`` and ``alltenses`` the file
- ``langs.gfcm``, a package with precompiled  ``Lang``//L// grammars


A way to test and view the resource grammar is to load ``langs.gfcm`` either into ``gfeditor``
or into the ``gf`` shell and perform actions such as syntax editing and treebank generation.
For instance, the command
```
  > p -lang=LangEng -cat=S "this grammar is too big" | tb
```
creates a treebank entry with translations of this sentence.

For parsing, currently only English and the Scandinavian languages are within the limits ofr
reasonable resources. For other languages //L//, parsing with ``Lang``//L// will probably eat
up the computer resources before finishing the parser generation.



==Accessing the lower level ground API==

The ``Syntax`` API is implemented in terms a bunch of ``abstract`` modules, which
as of version 1.2 are mainly interesting for implementors of the resource.
See the [documentation for version 1.1 index-1.1.html] for more details.


==Known bugs and missing components==

Danish
- the lexicon and chosen inflections are only partially verified


English 


Finnish
- wrong cases in some passive constructions


French
- multiple clitics (with V3) not always right
- third person pronominal questions with inverted word order
  have wrong forms if "t" is required e.g.
  (e.g. "comment fera-t-il" becomes "comment fera il")


German


Italian
- multiple clitics (with V3) not always right


Norwegian
- the lexicon and chosen inflections are only partially verified


Russian
- some functions missing
- some regular paradigms are missing


Spanish
- multiple clitics (with V3) not always right
- missing contractions with imperatives and clitics


Swedish




==More reading==

[Synopsis synopsis.html]. The concise guide to API v. 1.2.

[Grammars as Software Libraries gslt-sem-2006.html]. Slides
with background and motivation for the resource grammar library.

[GF Resource Grammar Library Version 1.0 clt2006.html]. Slides
giving an overview of the library and practical hints on its use.

[How to write resource grammars  Resource-HOWTO.html]. Helps you
start if you want to add another language to the library.

[Parametrized modules for Romance languages http://www.cs.chalmers.se/~aarne/geocal2006.pdf]. 
Slides explaining some ideas in the implementation of
French, Italian, and Spanish.

[Grammar writing by examples http://www.cs.chalmers.se/~aarne/slides/webalt-2005.pdf].
Slides showing how linearization rules are written as strings parsable by the resource grammar.

[Multimodal Resource Grammars http://www.cs.chalmers.se/~aarne/slides/talk-edin2005.pdf].
Slides showing how to use the multimodal resource library. N.B. the library
examples are from ``multimodal/old``, which is a reduced-size API.

[GF Resource Grammar Library ../../../doc/resource.pdf] (pdf).
Printable user manual with API documentation, for version 1.0.