1 .TH lt-proc 1 2006-03-23 "" ""
3 lt-proc \- This application is part of the lexical processing modules
8 This tool is part of the apertium machine translation
9 architecture: \fBhttp://apertium.sf.net\fR.
21 ] fst_file [input_file [output_file]]
26 .B \-\-generation \fR|
27 .B \-\-non-marked-gen \fR|
28 .B \-\-debugged-gen \fR|
29 .B \-\-post-generation \fR|
33 ] fst_file [input_file [output_file]]
36 is the application responsible of providing the four lexical
37 processing functionalities
40 \(bu \fImorphological analyser\fR ( option \fB\-a\fR )
42 \(bu \fImorphological generator\fR ( option \fB\-g\fR )
44 \(bu \fImorphological generator\fR (without marks) ( option \fB\-n\fR )
46 \(bu \fImorphological generator\fR (with debugging information) ( option \fB\-d\fR )
48 \(bu \fIpost-generator\fR ( option \fB\-p\fR )
52 It accomplishes these tasks by reading binary files containing a
53 compact and efficient representation of dictionaries (a class of
54 finite-state transducers called augmented letter transducers). These
55 files are generated by \fBlt\-comp(1)\fR.
57 It is worth to mention that some characters
58 (`\fB[\fR', `\fB]\fR', `\fB$\fR', `\fB^\fR', `\fB/\fR', `\fB+\fR') are
59 \fIspecial\fR chars used for format and encapsulation. They should be
60 escaped if they have to be used literally, for
61 instance: `\fB[\fR'...`\fB]\fR' are ignored and the format of a
62 \fIlinefeed\fR is `\fB^\fR...\fB$\fR'.
66 Tokenizes the text in surface forms (lexical units as they appear in
67 texts) and delivers, for each surface form, one or more lexical forms
68 consisting of lemma, lexical category and morphological inflection
69 information. Tokenization is not straightforward due to the existence,
70 on the one hand, of contractions, and, on the other hand, of
71 multi-word lexical units. For contractions, the system reads in a
72 single surface form and delivers the corresponding sequence of lexical
73 forms. Multi-word surface forms are analysed in a left-to-right,
74 longest-match fashion. Multi-word surface forms may be invariable
75 (such as a multi-word preposition or conjunction) or inflected (for
76 example, in es, \fI"echaban de menos"\fR, \(dqthey missed\(dq, is a
77 form of the imperfect indicative tense of the verb \fI"echar de
78 menos"\fR, \(dqto miss\(dq). Limited support for some kinds of
79 discontinuous multi-word units is also available. Single-word surface
80 forms analysis produces output like the one in these examples:
81 \fI"cantar"\fR \-> `\fI^cantar/cantar<vblex><inf>$\fR' or
83 `\fI^cantaba/cantar<vblex><pii><p1><sg>/cantar<vblex><pii><p3><sg>$\fR'.
85 .B \-g, \-\-generation
86 Delivers a target-language surface form for each target-language
87 lexical form, by suitably inflecting it.
89 .B \-n, \-\-non-marked-gen
90 Morphological generation (like \fB-g\fR) but without unknown word
93 .B \-d, \-\-debugged-gen
94 Morphological generation (like \fB-g\fR) but also with tags information
95 for errors in the transfer and generator dictionaries.
97 .B \-p, \-\-post-generation
98 Performs orthographical operations such as contractions and
99 apostrophations. The post-generator is usually \fIdormant\fR (just
100 copies the input to the output) until a special \fIalarm\fR symbol
101 contained in some target-language surface forms \fIwakes\fR it up to
102 perform a particular string transformation if necessary; then it goes
106 Input processing is in \fIorthoepikon\fR (previously `\fIsao\fR')
107 annotation system format: \fBhttp://orthoepikon.sf.net\fR.
110 Display the version number.
116 The input compiled dictionary.
120 .I apertium-tagger\fR(1),
121 .I apertium-translator\fR(1),
124 Lots of...lurking in the dark and waiting for you!
126 (c) 2005,2006 Universitat d'Alacant / Universidad de Alicante. All rights