michael@0: /* Hyphen - hyphenation library using converted TeX hyphenation patterns
michael@0:  *
michael@0:  * (C) 1998 Raph Levien
michael@0:  * (C) 2001 ALTLinux, Moscow
michael@0:  * (C) 2006, 2007, 2008 László Németh
michael@0:  *
michael@0:  * This was part of libHnj library by Raph Levien.
michael@0:  *
michael@0:  * Peter Novodvorsky from ALTLinux cut hyphenation part from libHnj
michael@0:  * to use it in OpenOffice.org.
michael@0:  *
michael@0:  * Non-standard and compound word hyphenation support by László Németh.
michael@0:  * 
michael@0:  * License is the original LibHnj license:
michael@0:  *
michael@0:  * LibHnj is dual licensed under LGPL and MPL. Boilerplate for both
michael@0:  * licenses follows.
michael@0:  */
michael@0: 
michael@0: /* LibHnj - a library for high quality hyphenation and justification
michael@0:  * Copyright (C) 1998 Raph Levien
michael@0:  *
michael@0:  * This library is free software; you can redistribute it and/or
michael@0:  * modify it under the terms of the GNU Library General Public
michael@0:  * License as published by the Free Software Foundation; either
michael@0:  * version 2 of the License, or (at your option) any later version.
michael@0:  *
michael@0:  * This library is distributed in the hope that it will be useful,
michael@0:  * but WITHOUT ANY WARRANTY; without even the implied warranty of
michael@0:  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
michael@0:  * Library General Public License for more details.
michael@0:  *
michael@0:  * You should have received a copy of the GNU Library General Public
michael@0:  * License along with this library; if not, write to the 
michael@0:  * Free Software Foundation, Inc., 59 Temple Place - Suite 330, 
michael@0:  * Boston, MA  02111-1307  USA.
michael@0: */
michael@0: 
michael@0: /*
michael@0:  * The contents of this file are subject to the Mozilla Public License
michael@0:  * Version 1.0 (the "MPL"); you may not use this file except in
michael@0:  * compliance with the MPL.  You may obtain a copy of the MPL at
michael@0:  * http://www.mozilla.org/MPL/
michael@0:  *
michael@0:  * Software distributed under the MPL is distributed on an "AS IS" basis,
michael@0:  * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the MPL
michael@0:  * for the specific language governing rights and limitations under the
michael@0:  * MPL.
michael@0:  *
michael@0:  */
michael@0: #ifndef __HYPHEN_H__
michael@0: #define __HYPHEN_H__
michael@0: 
michael@0: #ifdef __cplusplus
michael@0: extern "C" {
michael@0: #endif /* __cplusplus */
michael@0: 
michael@0: typedef struct _HyphenDict HyphenDict;
michael@0: typedef struct _HyphenState HyphenState;
michael@0: typedef struct _HyphenTrans HyphenTrans;
michael@0: #define MAX_CHARS 100
michael@0: #define MAX_NAME 20
michael@0: 
michael@0: struct _HyphenDict {
michael@0:   /* user options */
michael@0:   char lhmin;    /* lefthyphenmin: min. hyph. distance from the left side */
michael@0:   char rhmin;    /* righthyphenmin: min. hyph. distance from the right side */
michael@0:   char clhmin;   /* min. hyph. distance from the left compound boundary */
michael@0:   char crhmin;   /* min. hyph. distance from the right compound boundary */
michael@0:   char * nohyphen; /* comma separated list of characters or character
michael@0:                     sequences with forbidden hyphenation */
michael@0:   int nohyphenl; /* count of elements in nohyphen */
michael@0:   /* system variables */
michael@0:   int num_states;
michael@0:   char cset[MAX_NAME];
michael@0:   int utf8;
michael@0:   HyphenState *states;
michael@0:   HyphenDict *nextlevel;
michael@0: };
michael@0: 
michael@0: struct _HyphenState {
michael@0:   char *match;
michael@0:   char *repl;
michael@0:   signed char replindex;
michael@0:   signed char replcut;
michael@0:   int fallback_state;
michael@0:   int num_trans;
michael@0:   HyphenTrans *trans;
michael@0: };
michael@0: 
michael@0: struct _HyphenTrans {
michael@0:   char ch;
michael@0:   int new_state;
michael@0: };
michael@0: 
michael@0: HyphenDict *hnj_hyphen_load (const char *fn);
michael@0: void hnj_hyphen_free (HyphenDict *dict);
michael@0: 
michael@0: /* obsolete, use hnj_hyphen_hyphenate2() or *hyphenate3() functions) */
michael@0: int hnj_hyphen_hyphenate (HyphenDict *dict,
michael@0: 			   const char *word, int word_size,
michael@0: 			   char *hyphens);
michael@0: 
michael@0: /*
michael@0: 
michael@0:  int hnj_hyphen_hyphenate2(): non-standard hyphenation.
michael@0: 
michael@0:  (It supports Catalan, Dutch, German, Hungarian, Norwegian, Swedish
michael@0:   etc. orthography, see documentation.)
michael@0:  
michael@0:  input data:
michael@0:  word:      input word
michael@0:  word_size: byte length of the input word
michael@0:  
michael@0:  hyphens:   allocated character buffer (size = word_size + 5)
michael@0:  hyphenated_word: allocated character buffer (size ~ word_size * 2) or NULL
michael@0:  rep, pos, cut: pointers (point to the allocated and _zeroed_ buffers
michael@0:                 (size=word_size) or with NULL value) or NULL
michael@0: 
michael@0:  output data:
michael@0:  hyphens:   hyphenation vector (hyphenation points signed with odd numbers)
michael@0:  hyphenated_word: hyphenated input word (hyphens signed with `='),
michael@0:                   optional (NULL input)
michael@0:  rep:       NULL (only standard hyph.), or replacements (hyphenation points
michael@0:             signed with `=' in replacements);
michael@0:  pos:       NULL, or difference of the actual position and the beginning
michael@0:             positions of the change in input words;
michael@0:  cut:       NULL, or counts of the removed characters of the original words
michael@0:             at hyphenation,
michael@0: 
michael@0:  Note: rep, pos, cut are complementary arrays to the hyphens, indexed with the
michael@0:        character positions of the input word.
michael@0: 
michael@0:  For example:
michael@0:  Schiffahrt -> Schiff=fahrt,
michael@0:  pattern: f1f/ff=f,1,2
michael@0:  output: rep[5]="ff=f", pos[5] = 1, cut[5] = 2
michael@0: 
michael@0:  Note: hnj_hyphen_hyphenate2() can allocate rep, pos, cut (word_size
michael@0:        length arrays):
michael@0: 
michael@0:  char ** rep = NULL;
michael@0:  int * pos = NULL;
michael@0:  int * cut = NULL;
michael@0:  char hyphens[MAXWORDLEN];
michael@0:  hnj_hyphen_hyphenate2(dict, "example", 7, hyphens, NULL, &rep, &pos, &cut);
michael@0:  
michael@0:  See example in the source distribution.
michael@0: 
michael@0: */
michael@0: 
michael@0: int hnj_hyphen_hyphenate2 (HyphenDict *dict,
michael@0:         const char *word, int word_size, char * hyphens,
michael@0:         char *hyphenated_word, char *** rep, int ** pos, int ** cut);
michael@0: 
michael@0: /* like hnj_hyphen_hyphenate2, but with hyphenmin parameters */
michael@0: /* lhmin: lefthyphenmin
michael@0:  * rhmin: righthyphenmin
michael@0:  * clhmin: compoundlefthyphemin
michael@0:  * crhmin: compoundrighthyphenmin
michael@0:  * (see documentation) */
michael@0: 
michael@0: int hnj_hyphen_hyphenate3 (HyphenDict *dict,
michael@0: 	const char *word, int word_size, char * hyphens,
michael@0: 	char *hyphword, char *** rep, int ** pos, int ** cut,
michael@0: 	int lhmin, int rhmin, int clhmin, int crhmin);
michael@0: 
michael@0: #ifdef __cplusplus
michael@0: }
michael@0: #endif /* __cplusplus */
michael@0: 
michael@0: #endif /* __HYPHEN_H__ */