Debian Codegraph @

manual/cgglossary.xml

raw
1 <?xml version="1.0" encoding="utf-8"?>
2 <!DOCTYPE chapter SYSTEM "docbook-dtd-45/docbookx.dtd">
3
4 <chapter id="cgglossary">
5 <title>Constraint Grammar Glossary</title>
6
7 <para>
8 The terms in this glossary are specific to the workings and concepts of constraint grammar;
9 internal and otherwise.
10 </para>
11
12 <section id="cg-baseform">
13 <title>Baseform</title>
14 <indexterm>
15 <primary>baseform</primary>
16 <seealso>wordform</seealso>
17 </indexterm>
18 <para>
19 The first part of a reading. Every reading must have one of these. It is usually the
20 base form of the word in the containing cohort. Other than that, it is a tag like any other.
21 </para>
22 <screen>
23 "defy"
24 </screen>
25 </section>
26
27 <section id="cg-cohort">
28 <title>Cohort</title>
29 <indexterm>
30 <primary>cohort</primary>
31 </indexterm>
32 <para>
33 A single word with all its readings.
34 </para>
35 <screen>
36 "&lt;defiable&gt;"
37 &nbsp; "defy" adjective
38 &nbsp; "defy" adverb
39 &nbsp; "defy" verb plural
40 </screen>
41 </section>
42
43 <section id="cg-contextual-target">
44 <title>Contextual Target</title>
45 <indexterm>
46 <primary>contextual target</primary>
47 <secondary>target</secondary>
48 </indexterm>
49 <para>
50 The part of a dependency rule that locates a suitable cohort to attach to.
51 It is no different from a normal contextual test and can have links, barriers, etc.
52 <emphasis>
53 While you technically can NEGATE the entire contextual target,
54 it is probably not a good idea to do so.
55 </emphasis>
56 </para>
57 <screen>
58 (-1* Verb LINK 1* Noun)
59 </screen>
60 </section>
61
62 <section id="cg-contextual-test">
63 <title>Contextual Test</title>
64 <indexterm>
65 <primary>contextual test</primary>
66 <secondary>context</secondary>
67 <tertiary>test</tertiary>
68 </indexterm>
69 <para>
70 The part of a disambiguation rule that looks at words surrounding the active cohort.
71 A rule will not act unless all its contextual tests are satisfied.
72 </para>
73 <screen>
74 (-1* Verb LINK 1* Noun)
75 </screen>
76 </section>
77
78 <section id="cg-dependency">
79 <title>Dependency</title>
80 <indexterm>
81 <primary>dependency</primary>
82 </indexterm>
83 <indexterm>
84 <primary>attachment</primary>
85 </indexterm>
86 <para>
87 A tag in #X-&gt;Y or #X→Y format denoting that this cohort is the child of another cohort.
88 </para>
89 </section>
90
91 <section id="cg-disambiguation-window">
92 <title>Disambiguation Window</title>
93 <indexterm>
94 <primary>disambiguation window</primary>
95 <secondary>window</secondary>
96 </indexterm>
97 <para>
98 An array of cohorts as provided by the input stream. Usually the last cohort in a window is
99 one from the DELIMITERS definition.
100 CG-3 can keep several windows in memory
101 at any time in order to fascilitate cross-window scanning from contextual tests. It is also
102 possible to break a window into smaller windows on-the-fly with a DELIMIT rule.
103 </para>
104 </section>
105
106 <section id="cg-mapping-tag">
107 <title>Mapping Tag</title>
108 <indexterm>
109 <primary>mapping tag</primary>
110 <secondary>mapping</secondary>
111 <seealso>mapping prefix</seealso>
112 <seealso>MAP</seealso>
113 </indexterm>
114 <para>
115 A special type of tag, as defined by the mapping prefix.
116 If a reading contains more than one mapping tag, the reading is multiplied
117 into several readings with one of the mapping tags each and all the normal tags copied.
118 </para>
119 </section>
120
121 <section id="cg-mapping-prefix">
122 <title>Mapping Prefix</title>
123 <indexterm>
124 <primary>mapping prefix</primary>
125 <secondary>prefix</secondary>
126 </indexterm>
127 <para>
128 A single unicode character. If a tag begins with this character, it is considered a mapping tag.
129 The character can be changed with the grammar keyword MAPPING-PREFIX or the command line flag
130 --prefix. Defaults to @.
131 </para>
132 </section>
133
134 <section id="cg-reading">
135 <title>Reading</title>
136 <indexterm>
137 <primary>reading</primary>
138 <seealso>cohort</seealso>
139 </indexterm>
140 <para>
141 Part of a cohort. Defines one variant of the word in question.
142 A reading must begin with a baseform.
143 Readings are the whole
144 basis for how CG operates; they are what we disambiguate between and test against.
145 The goal is to exit with one per cohort.
146 </para>
147 <screen>
148 &nbsp; "defy" adjective
149 </screen>
150 </section>
151
152 <section id="cg-rule">
153 <title>Rule</title>
154 <indexterm>
155 <primary>rule</primary>
156 </indexterm>
157 <para>
158 The primary workhorses of CG. They can add, remove, alter readings, cohorts, and windows based on
159 tests and type.
160 </para>
161 <screen>
162 ADD (tags) targetset (-1* ("someword")) ;
163 SELECT targetset (-1* ("someword") BARRIER (tag)) ;
164 DELIMIT targetset (-1* ("someword")) ;
165 </screen>
166 </section>
167
168 <section id="cg-set">
169 <title>Set</title>
170 <indexterm>
171 <primary>set</primary>
172 </indexterm>
173 <para>
174 A list of tags or a combination of other sets. Used as targets for rules and contextual tests.
175 </para>
176 <screen>
177 LIST SomeSet = tags moretags (etc tags) ;
178 LIST OtherSet = moretags (etc tags) ;
179 SET CombiSet = SomeSet + OtherSet - (tag) ;
180 </screen>
181 </section>
182
183 <section id="cg-tag">
184 <title>Tag</title>
185 <indexterm>
186 <primary>tag</primary>
187 </indexterm>
188 <para>
189 Any simple string of text. Readings and sets are built from tags.
190 </para>
191 <screen>
192 "plumber" noun @jobtitle
193 </screen>
194 </section>
195
196 <section id="cg-wordform">
197 <title>Wordform</title>
198 <indexterm>
199 <primary>wordform</primary>
200 <seealso>baseform</seealso>
201 </indexterm>
202 <para>
203 The container word of a cohort. Every cohort has exactly one of these. It is usually
204 the original word of the input text.
205 </para>
206 <screen>
207 "&lt;defiable&gt;"
208 </screen>
209 </section>
210
211 </chapter>