2 /* Make header file work when included from C++ */
8 typedef unsigned char sb_symbol;
10 /* FIXME - should be able to get a version number for each stemming
11 * algorithm (which will be incremented each time the output changes). */
13 /** Returns an array of the names of the available stemming algorithms.
14 * Note that these are the canonical names - aliases (ie, other names for
15 * the same algorithm) will not be included in the list.
16 * The list is terminated with a null pointer.
18 * The list must not be modified in any way.
20 const char ** sb_stemmer_list(void);
22 /** Create a new stemmer object, using the specified algorithm, for the
23 * specified character encoding.
25 * All algorithms will usually be available in UTF-8, but may also be
26 * available in other character encodings.
28 * @param algorithm The algorithm name. This is either the english
29 * name of the algorithm, or the 2 or 3 letter ISO 639 codes for the
30 * language. Note that case is significant in this parameter - the
31 * value should be supplied in lower case.
33 * @param charenc The character encoding. NULL may be passed as
34 * this value, in which case UTF-8 encoding will be assumed. Otherwise,
35 * the argument may be one of "UTF_8", "ISO_8859_1" (ie, Latin 1),
36 * "CP850" (ie, MS-DOS Latin 1) or "KOI8_R" (Russian). Note that
37 * case is significant in this parameter.
39 * @return NULL if the specified algorithm is not recognised, or the
40 * algorithm is not available for the requested encoding. Otherwise,
41 * returns a pointer to a newly created stemmer for the requested algorithm.
42 * The returned pointer must be deleted by calling sb_stemmer_delete().
44 * @note NULL will also be returned if an out of memory error occurs.
46 struct sb_stemmer * sb_stemmer_new(const char * algorithm, const char * charenc);
48 /** Delete a stemmer object.
50 * This frees all resources allocated for the stemmer. After calling
51 * this function, the supplied stemmer may no longer be used in any way.
53 * It is safe to pass a null pointer to this function - this will have
56 void sb_stemmer_delete(struct sb_stemmer * stemmer);
60 * The return value is owned by the stemmer - it must not be freed or
61 * modified, and it will become invalid when the stemmer is called again,
62 * or if the stemmer is freed.
64 * The length of the return value can be obtained using sb_stemmer_length().
66 * If an out-of-memory error occurs, this will return NULL.
68 const sb_symbol * sb_stemmer_stem(struct sb_stemmer * stemmer,
69 const sb_symbol * word, int size);
71 /** Get the length of the result of the last stemmed word.
72 * This should not be called before sb_stemmer_stem() has been called.
74 int sb_stemmer_length(struct sb_stemmer * stemmer);