cline_interface.pl -- NSM-DALIA, command line version

This module provides aliases to call the main NSM-DALIA predicates, (contained in modules nsmdalia.pl and file_reader.pl), and the running interface.

To be done: - wl/0 implementation; - re-implementation of pw

l(+Lang) is det

Loads a language module into the program, and sets the current_language flag to the loaded language. From now on, parse commands will be taken to refer to that language as the parsing-from language.

See also

- nsmdalia:load(Lang)

The "Lang" argument is either a language code (without quotation marks), or a full language name (a PROLOG string, between double quotation marks).

The "load" command works like this: if a compiled module is found in the lang_bin directory, it will be loaded. If not, the directory lang_src will be searched for the source files, compiled and loaded. (The latter procedure is slower).

The user who is not also writing a grammar need only know which languages are available, and what their code is (the "list" command displays all the installed languages with their code). As the only language available in NSM-DALIA v. 0.8 is English (code: eng), you can load the English grammar with:

DALIA> l(eng).

or:

DALIA> l("English").

A note for gramamr writers: please remember that, once a grammar is saved in the lang_bin directory ("s" command), l(Lang) will read from that directory, NOT from the source file. If, after you modified your source file, and you are wondering why on earth that sentence still do not parse, though you have fixed the bug in your grammar -- that's because you have loaded the file with "l", not with "b", so the program does not read the new grammar, but the old compiled one! When you are developing a grammar, better not to compile it ("c" command") until you are finished. (However, see "b" command below).

l2(+Lang) is det

This command loads a grammar module, just like l(Lang), with the difference that the language loaded is set as l2 (second language). Translation commands like t(Sentence) will translate into l2.

ldb(+FileName:string) is det

Loads a text database. Text is stored in a series of mtext/3 facts (see nsmfiles.txt for details).

b(+Lang) is det

Full form: build(+Lang). Loads a language module into the program, and sets the current_language flag to the loaded language. Files are searched for exclusively in the lang_src directory. Useful for grammar writers. The "Lang" argument is a language code or a string (full language name).

sl1(+Lang) is det

Sets Lang as the current language.

sl2(+Lang) is det

Full form: set_current_l2(Lang).

Used in automatic translation. Once current_lang and current_l2 are set, the "translation" commands translate NSM sentences from current_lang into current_l2. Lang = language code.

sl

Full form: switch_languages.

Useful to switch between various languages when you have loaded more than one grammar. The "Lang" argument is a language code.

sm(+MarkUp) is det

Full form: set_markup(+MarkUp).

MarkUp is a term referring to one of the supported markup schemes (see NSM-files.txt and grammars.txt).

smf(+Format) is det

Format is one of the terms line_by_line or whole_text. The NSM-file parser has an option for translating an NSM text and displaying the translation together with the original.

Note that, for example, the rtf markup scheme requires a line_by_line option which, however, result in the texts being displayed in a two_column table, one column for the original, the other for the translation, and with corresponding lines aligned.

stt(+TableId:integer) is det

For languages having more than one transcription possibility, sets the transcription table number TableId as active.

pw(+W) is det

Full form: parse_and_write_word(W).

You won't need this very much, because "ps" can parse single words, as well as whole sentences. When you are developing a grammar, however, this command can be useful, because it also outputs a morpheme split-up of the word. The argument W is a double-quoted string, which is analyzed as a word of the last current language set. Ex.

DALIA> pw("things").

NSM-DALIA answers:

Morphemic String: thing-s
ct(n(n), sp(e, e, plur(e), [], something(thing))).

ps(+S) is det

Full form: parse_and_write_sentence(S). Like "pw", but it parses a whole sentence. Ex.

DALIA> ps("something good is happening now").

NSM-DALIA answers with the corresponding NSM-PROLOG formula:

ct(s, s(e, time(e, now, e), e, e, e,
        p(i(happen),
          [o:sp(e, e, sing(e),[good], something(thing)),
           d:e]),
        e, e)).

pst(+S) is det

Full form: parse_and_write_tabular_sentence(S,min). Like "ps", but delivers the analysis in tabular fashion.

Prints only the "minimal" table.

DALIA> ps("I do something good")

NSM-DALIA answers with the corresponding NSM-PROLOG formula:

s: _____
. pred: _____
. . v:do
. . a: _____
. . . pers:loc(me)
. . . n:me
. . o: _____
. . . a: _____
. . . . eval: _____
. . . . . a:good
. . . n:something(thing)

pstf(+S) is det

Full form: parse_and_write_tabular_sentence(S,full). Like "ps", but delivers the analysis in tabular fashion.

Prints only the full table.

DALIA> ps("I do something good")

NSM-DALIA answers with the corresponding NSM-PROLOG formula:

s: _____
. c: _____
. . compl:e
. . top:e
. . int:e
. . top2:e
. . foc:e
. . pol:e
. mod: _____
. . speech_act:e
. . eval:e
. . evid:e
. . epist:e
. f: _____
. . top3:e
. . finite:e
. t: _____
. . past:e
. . fut:e
. m: _____
. . irrealis:e
. . necess:e
. . possib:e
. . vol:e
. . oblig:e
. . allow:e
. asp: _____
. . hab:e
. . rep:e
. . freq:e
. . celer:e
. . ant:e
. . term:e
. . cont:e
. . perf:e
. . retro:e
. . pross:e
. . dur:e
. . prog:e
. . prosp:e
. . compl_sg:e
. . compl_pl:e
. vo: _____
. . v_1:e
. . v_2:e
. . v_3:e
. ak: _____
. . celer:e
. . comp:e
. . rep:e
. . freq:e
. pred: _____
. . v:do
. . a: _____
. . . det:e
. . . alt:e
. . . q:e
. . . pers:loc(me)
. . . a: _____
. . . . eval:e
. . . . size:e
. . . . length:e
. . . . height:e
. . . . speed:e
. . . . width:e
. . . . weight:e
. . . . temp:
. . . . age:e
. . . . shape:e
. . . . colour:e
. . . . origin:e
. . . . material:e
. . . dem:e
. . . poss:e
. . . class:e
. . . n:me
. . o: _____
. . . det:e
. . . alt:e
. . . q:e
. . . pers:e
. . . a: _____
. . . . eval: _____
. . . . . int:e
. . . . . a:good
. . . . size:e
. . . . length:e
. . . . height:e
. . . . speed:e
. . . . width:e
. . . . weight:e
. . . . temp:e
. . . . age:e
. . . . shape:e
. . . . colour:e
. . . . origin:e
. . . . material:e
. . . dem:e
. . . poss:e
. . . class:e
. . . n:something(thing)
. . d:e
. . e:e
. . c:e
. . i:e
. . b:e
. . l:e
. . m:e

gs(+LF) is det

Full form: gen_and_write_sentence(LF). The "LF" argument is an NSM-PROLOG formula; the output is the corresponding sentence in the current language (or the formula itself, if NSM-DALIA fails generation). Ex.

DALIA> gs(ct(s, s(e, before(now), e, e, e,
                  p(do, [a:sp(e, e, sing(e), [], me),
                         o:sp(e, e, sing(e), [good],
                              something(thing)),
                         d:e, c:e, i:e]),
                  e, e))).

Answer:

I did something good

t(+Sentence) is det

Full form: translate_and_write_sentence(Sentence). Translates a sentence from current_language to current_l2. Sentence argument is a string, between double quotation marks.

so(+Filename) is det

Full form: set_output_file(Filename). Redirects the output to the file named Filename (in some operating systems, you will need to give a full name with extension -- NSM-DALIA adds none). "Filename" is double-quoted string. Ex.

DALIA> so("pippo.txt").
  * Output file set to pippo.txt

You will perhaps use this mostly before a "pf" command.

ro(+F) is det

Used to select an output file if the user wants it to be overwritten by the new data. so/1 appends the output file if it exists, and does not rewrite it, as ro/1 does.

co

Full form: close_output_file

Users will rarely use this command, because the output file is automatically closed (and the output redirected again to the console) when the parsing of an NSM-input file is finished.

pf(+F) is det

Full form: parse_file(F). Parses and outputs a text file with particular tags (see the documentation file "NSM-files.txt"). You can write a whole file of NSM texts in, say, English NSM, and then have it parsed or translated automatically. This will probably be one of the most used commands, when other language modules are available.

tm(+SWITCH) is det

Full form: trace_morphology(SWITCH). To switch on verbose mode for morphology parsing, say:

tm(1).

To switch it off, say

tm(0)

Verbose modes are useful in grammar development. Turning tracing morphology on, the "parse" and "generate" commands will display information about the grammar rules used during the parsing process, in the morphophonemic component of the grammar. This will help you to find out why the morphophonemic component of your grammar is not doing what you intended it to do.

ts(+SWITCH) is det