<!DOCTYPE File PUBLIC "-//Liam Quin//DTD C API Documentation v1.1//EN" "doc.dtd"><File>
<Entry dir="liblqtext">
<Function File="../../src/liblqtext/phrall.c">
<Name>LQT_AllPhrasesOfLengthNOrMore</Name>
<Class>Retrieval/Matching, Retrieval/Phrases
<Purpose>
<P>Finds all sequences of N or more words which occur in the data.
For example, given the phrase `the barefooted boy was very slender',
and supposing `the' to be the only word for which LQT_WordIsInStopList
returns true, LQT_AllPhrasesOfLengthNOrMore might find
`barefooted boy' and `boy was very' and `very slender' as
sub-phrases that occur; if the entire phrase occurs, it will
be returned.</P>
<P>If a phrase of M words matches, all phrases of lengths from N to
M inclusive will also be returned.</P>
<P>It is the caller's responsibility to deallocate the returned array
and its elements.</P>
<Returns>
an array of t_PhraseElement structures, and the number of distinct
phrases found in *Countp.
<Notes>
This function is experimental.  It has not been optimised, and
is currently unusable for long phrases as a result.
</Function>
<Decl>
API t_PhraseElement *
LQT_AllPhrasesOfLengthNOrMore(db, N, theQuery, Countp)
    t_LQTEXT_Database *db;
    int N;
    char *theQuery;
    long *Countp;
</Decl>
</Entry>
