/* ***** BEGIN LICENSE BLOCK *****
* Version: MPL 1.1/GPL 2.0/LGPL 2.1
*
* Copyright (C) 2002-2022 Németh László
*
* The contents of this file are subject to the Mozilla Public License Version
* 1.1 (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
* http://www.mozilla.org/MPL/
*
* Software distributed under the License is distributed on an "AS IS" basis,
* WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
* for the specific language governing rights and limitations under the
* License.
*
* Hunspell is based on MySpell which is Copyright (C) 2002 Kevin Hendricks.
*
* Contributor(s): David Einstein, Davide Prina, Giuseppe Modugno,
* Gianluca Turconi, Simon Brouwer, Noll János, Bíró Árpád,
* Goldman Eleonóra, Sarlós Tamás, Bencsáth Boldizsár, Halácsy Péter,
* Dvornik László, Gefferth András, Nagy Viktor, Varga Dániel, Chris Halls,
* Rene Engelhard, Bram Moolenaar, Dafydd Jones, Harri Pitkänen
*
* Alternatively, the contents of this file may be used under the terms of
* either the GNU General Public License Version 2 or later (the "GPL"), or
* the GNU Lesser General Public License Version 2.1 or later (the "LGPL"),
* in which case the provisions of the GPL or the LGPL are applicable instead
* of those above. If you wish to allow use of your version of this file only
* under the terms of either the GPL or the LGPL, and not to allow others to
* use your version of this file under the terms of the MPL, indicate your
* decision by deleting the provisions above and replace them with the notice
* and other provisions required by the GPL or the LGPL. If you do not delete
* the provisions above, a recipient may use your version of this file under
* the terms of any one of the MPL, the GPL or the LGPL.
*
* ***** END LICENSE BLOCK ***** */
/*
* Copyright 2002 Kevin B. Hendricks, Stratford, Ontario, Canada
* And Contributors. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
*
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* 3. All modifications to the source code must be clearly marked as
* such. Binary redistributions based on modified source code
* must be clearly marked as modified versions in the documentation
* and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY KEVIN B. HENDRICKS AND CONTRIBUTORS
* ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
* FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL
* KEVIN B. HENDRICKS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
* BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
* LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <stdlib.h>
#include <string.h>
#include <stdio.h>
#include <ctype.h>
#include <time.h>
#include <algorithm>
#include <limits>
#include <string>
#include <vector>
#include "affixmgr.hxx"
#include "affentry.hxx"
#include "langnum.hxx"
#include "csutil.hxx"
AffixMgr::AffixMgr(
const char* affpath,
const std::vector<HashMgr*>& ptr,
const char* key)
: alldic(ptr)
, pHMgr(ptr[0]) {
// register hash manager and load affix data from aff file
csconv = NULL;
utf8 = 0;
complexprefixes = 0;
parsedmaptable =
false;
parsedbreaktable =
false;
iconvtable = NULL;
oconvtable = NULL;
// allow simplified compound forms (see 3rd field of CHECKCOMPOUNDPATTERN)
simplifiedcpd = 0;
parsedcheckcpd =
false;
parseddefcpd =
false;
phone = NULL;
compoundflag = FLAG_NULL;
// permits word in compound forms
compoundbegin = FLAG_NULL;
// may be first word in compound forms
compoundmiddle = FLAG_NULL;
// may be middle word in compound forms
compoundend = FLAG_NULL;
// may be last word in compound forms
compoundroot = FLAG_NULL;
// compound word signing flag
compoundpermitflag = FLAG_NULL;
// compound permitting flag for suffixed word
compoundforbidflag = FLAG_NULL;
// compound fordidden flag for suffixed word
compoundmoresuffixes = 0;
// allow more suffixes within compound words
checkcompounddup = 0;
// forbid double words in compounds
checkcompoundrep = 0;
// forbid bad compounds (may be non-compound word with
// a REP substitution)
checkcompoundcase =
0;
// forbid upper and lowercase combinations at word bounds
checkcompoundtriple = 0;
// forbid compounds with triple letters
simplifiedtriple = 0;
// allow simplified triple letters in compounds
// (Schiff+fahrt -> Schiffahrt)
forbiddenword = FORBIDDENWORD;
// forbidden word signing flag
nosuggest = FLAG_NULL;
// don't suggest words signed with NOSUGGEST flag
nongramsuggest = FLAG_NULL;
langnum = 0;
// language code (see http://l10n.openoffice.org/languages.html)
needaffix = FLAG_NULL;
// forbidden root, allowed only with suffixes
cpdwordmax = -1;
// default: unlimited wordcount in compound words
cpdmin = -1;
// undefined
cpdmaxsyllable = 0;
// default: unlimited syllablecount in compound words
pfxappnd = NULL;
// previous prefix for counting syllables of the prefix BUG
sfxappnd = NULL;
// previous suffix for counting syllables of the suffix BUG
sfxextra = 0;
// modifier for syllable count of sfxappnd BUG
checknum = 0;
// checking numbers, and word with numbers
havecontclass = 0;
// flags of possible continuing classes (double affix)
// LEMMA_PRESENT: not put root into the morphological output. Lemma presents
// in morhological description in dictionary file. It's often combined with
// PSEUDOROOT.
lemma_present = FLAG_NULL;
circumfix = FLAG_NULL;
onlyincompound = FLAG_NULL;
maxngramsugs = -1;
// undefined
maxdiff = -1;
// undefined
onlymaxdiff = 0;
maxcpdsugs = -1;
// undefined
nosplitsugs = 0;
sugswithdots = 0;
keepcase = 0;
forceucase = 0;
warn = 0;
forbidwarn = 0;
checksharps = 0;
substandard = FLAG_NULL;
fullstrip = 0;
sfx = NULL;
pfx = NULL;
for (
int i = 0; i < SETSIZE; i++) {
pStart[i] = NULL;
sStart[i] = NULL;
pFlag[i] = NULL;
sFlag[i] = NULL;
}
for (
int j = 0; j < CONTSIZE; j++) {
contclasses[j] = 0;
}
if (parse_file(affpath, key)) {
HUNSPELL_WARNING(stderr,
"Failure loading aff file %s\n", affpath);
}
if (cpdmin == -1)
cpdmin = MINCPDLEN;
}
AffixMgr::~AffixMgr() {
// pass through linked prefix entries and clean up
for (
int i = 0; i < SETSIZE; i++) {
pFlag[i] = NULL;
PfxEntry* ptr = pStart[i];
PfxEntry* nptr = NULL;
while (ptr) {
nptr = ptr->getNext();
delete (ptr);
ptr = nptr;
nptr = NULL;
}
}
// pass through linked suffix entries and clean up
for (
int j = 0; j < SETSIZE; j++) {
sFlag[j] = NULL;
SfxEntry* ptr = sStart[j];
SfxEntry* nptr = NULL;
while (ptr) {
nptr = ptr->getNext();
delete (ptr);
ptr = nptr;
nptr = NULL;
}
sStart[j] = NULL;
}
delete iconvtable;
delete oconvtable;
delete phone;
FREE_FLAG(compoundflag);
FREE_FLAG(compoundbegin);
FREE_FLAG(compoundmiddle);
FREE_FLAG(compoundend);
FREE_FLAG(compoundpermitflag);
FREE_FLAG(compoundforbidflag);
FREE_FLAG(compoundroot);
FREE_FLAG(forbiddenword);
FREE_FLAG(nosuggest);
FREE_FLAG(nongramsuggest);
FREE_FLAG(needaffix);
FREE_FLAG(lemma_present);
FREE_FLAG(circumfix);
FREE_FLAG(onlyincompound);
cpdwordmax = 0;
pHMgr = NULL;
cpdmin = 0;
cpdmaxsyllable = 0;
free_utf_tbl();
checknum = 0;
#ifdef MOZILLA_CLIENT
delete[] csconv;
#endif
}
void AffixMgr::finishFileMgr(FileMgr* afflst) {
delete afflst;
// convert affix trees to sorted list
process_pfx_tree_to_list();
process_sfx_tree_to_list();
}
// read in aff file and build up prefix and suffix entry objects
int AffixMgr::parse_file(
const char* affpath,
const char* key) {
// checking flag duplication
char dupflags[CONTSIZE];
char dupflags_ini = 1;
// first line indicator for removing byte order mark
int firstline = 1;
// open the affix file
FileMgr* afflst =
new FileMgr(affpath, key);
if (!afflst) {
HUNSPELL_WARNING(
stderr,
"error: could not open affix description file %s\n", affpath);
return 1;
}
// step one is to parse the affix file building up the internal
// affix data structures
// read in each line ignoring any that do not
// start with a known line type indicator
std::string line;
while (afflst->getline(line)) {
mychomp(line);
/* remove byte order mark */
if (firstline) {
firstline = 0;
// Affix file begins with byte order mark: possible incompatibility with
// old Hunspell versions
if (line.compare(0, 3,
"\xEF\xBB\xBF", 3) == 0) {
line.erase(0, 3);
}
}
/* parse in the keyboard string */
if (line.compare(0, 3,
"KEY", 3) == 0) {
if (!parse_string(line, keystring, afflst->getlinenum())) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the try string */
if (line.compare(0, 3,
"TRY", 3) == 0) {
if (!parse_string(line, trystring, afflst->getlinenum())) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the name of the character set used by the .dict and .aff */
if (line.compare(0, 3,
"SET", 3) == 0) {
if (!parse_string(line, encoding, afflst->getlinenum())) {
finishFileMgr(afflst);
return 1;
}
if (encoding ==
"UTF-8") {
utf8 = 1;
#ifndef OPENOFFICEORG
#ifndef MOZILLA_CLIENT
initialize_utf_tbl();
#endif
#endif
}
}
/* parse COMPLEXPREFIXES for agglutinative languages with right-to-left
* writing system */
if (line.compare(0, 15,
"COMPLEXPREFIXES", 15) == 0)
complexprefixes = 1;
/* parse in the flag used by the controlled compound words */
if (line.compare(0, 12,
"COMPOUNDFLAG", 12) == 0) {
if (!parse_flag(line, &compoundflag, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the flag used by compound words */
if (line.compare(0, 13,
"COMPOUNDBEGIN", 13) == 0) {
if (complexprefixes) {
if (!parse_flag(line, &compoundend, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
else {
if (!parse_flag(line, &compoundbegin, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
}
/* parse in the flag used by compound words */
if (line.compare(0, 14,
"COMPOUNDMIDDLE", 14) == 0) {
if (!parse_flag(line, &compoundmiddle, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the flag used by compound words */
if (line.compare(0, 11,
"COMPOUNDEND", 11) == 0) {
if (complexprefixes) {
if (!parse_flag(line, &compoundbegin, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
else {
if (!parse_flag(line, &compoundend, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
}
/* parse in the data used by compound_check() method */
if (line.compare(0, 15,
"COMPOUNDWORDMAX", 15) == 0) {
if (!parse_num(line, &cpdwordmax, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the flag sign compounds in dictionary */
if (line.compare(0, 12,
"COMPOUNDROOT", 12) == 0) {
if (!parse_flag(line, &compoundroot, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the flag used by compound_check() method */
if (line.compare(0, 18,
"COMPOUNDPERMITFLAG", 18) == 0) {
if (!parse_flag(line, &compoundpermitflag, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the flag used by compound_check() method */
if (line.compare(0, 18,
"COMPOUNDFORBIDFLAG", 18) == 0) {
if (!parse_flag(line, &compoundforbidflag, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
if (line.compare(0, 20,
"COMPOUNDMORESUFFIXES", 20) == 0) {
compoundmoresuffixes = 1;
}
if (line.compare(0, 16,
"CHECKCOMPOUNDDUP", 16) == 0) {
checkcompounddup = 1;
}
if (line.compare(0, 16,
"CHECKCOMPOUNDREP", 16) == 0) {
checkcompoundrep = 1;
}
if (line.compare(0, 19,
"CHECKCOMPOUNDTRIPLE", 19) == 0) {
checkcompoundtriple = 1;
}
if (line.compare(0, 16,
"SIMPLIFIEDTRIPLE", 16) == 0) {
simplifiedtriple = 1;
}
if (line.compare(0, 17,
"CHECKCOMPOUNDCASE", 17) == 0) {
checkcompoundcase = 1;
}
if (line.compare(0, 9,
"NOSUGGEST", 9) == 0) {
if (!parse_flag(line, &nosuggest, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
if (line.compare(0, 14,
"NONGRAMSUGGEST", 14) == 0) {
if (!parse_flag(line, &nongramsuggest, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the flag used by forbidden words */
if (line.compare(0, 13,
"FORBIDDENWORD", 13) == 0) {
if (!parse_flag(line, &forbiddenword, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the flag used by forbidden words (is deprecated) */
if (line.compare(0, 13,
"LEMMA_PRESENT", 13) == 0) {
if (!parse_flag(line, &lemma_present, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the flag used by circumfixes */
if (line.compare(0, 9,
"CIRCUMFIX", 9) == 0) {
if (!parse_flag(line, &circumfix, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the flag used by fogemorphemes */
if (line.compare(0, 14,
"ONLYINCOMPOUND", 14) == 0) {
if (!parse_flag(line, &onlyincompound, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the flag used by `needaffixs' (is deprecated) */
if (line.compare(0, 10,
"PSEUDOROOT", 10) == 0) {
if (!parse_flag(line, &needaffix, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the flag used by `needaffixs' */
if (line.compare(0, 9,
"NEEDAFFIX", 9) == 0) {
if (!parse_flag(line, &needaffix, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the minimal length for words in compounds */
if (line.compare(0, 11,
"COMPOUNDMIN", 11) == 0) {
if (!parse_num(line, &cpdmin, afflst)) {
finishFileMgr(afflst);
return 1;
}
if (cpdmin < 1)
cpdmin = 1;
}
/* parse in the max. words and syllables in compounds */
if (line.compare(0, 16,
"COMPOUNDSYLLABLE", 16) == 0) {
if (!parse_cpdsyllable(line, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the flag used by compound_check() method */
if (line.compare(0, 11,
"SYLLABLENUM", 11) == 0) {
if (!parse_string(line, cpdsyllablenum, afflst->getlinenum())) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the flag used by the controlled compound words */
if (line.compare(0, 8,
"CHECKNUM", 8) == 0) {
checknum = 1;
}
/* parse in the extra word characters */
if (line.compare(0, 9,
"WORDCHARS", 9) == 0) {
if (!parse_array(line, wordchars, wordchars_utf16,
utf8, afflst->getlinenum())) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the ignored characters (for example, Arabic optional diacretics
* charachters */
if (line.compare(0, 6,
"IGNORE", 6) == 0) {
if (!parse_array(line, ignorechars, ignorechars_utf16,
utf8, afflst->getlinenum())) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the input conversion table */
if (line.compare(0, 5,
"ICONV", 5) == 0) {
if (!parse_convtable(line, afflst, &iconvtable,
"ICONV")) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the output conversion table */
if (line.compare(0, 5,
"OCONV", 5) == 0) {
if (!parse_convtable(line, afflst, &oconvtable,
"OCONV")) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the phonetic translation table */
if (line.compare(0, 5,
"PHONE", 5) == 0) {
if (!parse_phonetable(line, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the checkcompoundpattern table */
if (line.compare(0, 20,
"CHECKCOMPOUNDPATTERN", 20) == 0) {
if (!parse_checkcpdtable(line, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the defcompound table */
if (line.compare(0, 12,
"COMPOUNDRULE", 12) == 0) {
if (!parse_defcpdtable(line, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the related character map table */
if (line.compare(0, 3,
"MAP", 3) == 0) {
if (!parse_maptable(line, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the word breakpoints table */
if (line.compare(0, 5,
"BREAK", 5) == 0) {
if (!parse_breaktable(line, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the language for language specific codes */
if (line.compare(0, 4,
"LANG", 4) == 0) {
if (!parse_string(line, lang, afflst->getlinenum())) {
finishFileMgr(afflst);
return 1;
}
langnum = get_lang_num(lang);
}
if (line.compare(0, 7,
"VERSION", 7) == 0) {
size_t startpos = line.find_first_not_of(
" \t", 7);
if (startpos != std::string::npos) {
version = line.substr(startpos);
}
}
if (line.compare(0, 12,
"MAXNGRAMSUGS", 12) == 0) {
if (!parse_num(line, &maxngramsugs, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
if (line.compare(0, 11,
"ONLYMAXDIFF", 11) == 0)
onlymaxdiff = 1;
if (line.compare(0, 7,
"MAXDIFF", 7) == 0) {
if (!parse_num(line, &maxdiff, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
if (line.compare(0, 10,
"MAXCPDSUGS", 10) == 0) {
if (!parse_num(line, &maxcpdsugs, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
if (line.compare(0, 11,
"NOSPLITSUGS", 11) == 0) {
nosplitsugs = 1;
}
if (line.compare(0, 9,
"FULLSTRIP", 9) == 0) {
fullstrip = 1;
}
if (line.compare(0, 12,
"SUGSWITHDOTS", 12) == 0) {
sugswithdots = 1;
}
/* parse in the flag used by forbidden words */
if (line.compare(0, 8,
"KEEPCASE", 8) == 0) {
if (!parse_flag(line, &keepcase, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the flag used by `forceucase' */
if (line.compare(0, 10,
"FORCEUCASE", 10) == 0) {
if (!parse_flag(line, &forceucase, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
/* parse in the flag used by `warn' */
if (line.compare(0, 4,
"WARN", 4) == 0) {
if (!parse_flag(line, &warn, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
if (line.compare(0, 10,
"FORBIDWARN", 10) == 0) {
forbidwarn = 1;
}
/* parse in the flag used by the affix generator */
if (line.compare(0, 11,
"SUBSTANDARD", 11) == 0) {
if (!parse_flag(line, &substandard, afflst)) {
finishFileMgr(afflst);
return 1;
}
}
if (line.compare(0, 11,
"CHECKSHARPS", 11) == 0) {
checksharps = 1;
}
/* parse this affix: P - prefix, S - suffix */
// affix type
char ft =
' ';
if (line.compare(0, 3,
"PFX", 3) == 0)
ft = complexprefixes ?
'S' :
'P';
if (line.compare(0, 3,
"SFX", 3) == 0)
ft = complexprefixes ?
'P' :
'S';
if (ft !=
' ') {
if (dupflags_ini) {
memset(dupflags, 0,
sizeof(dupflags));
dupflags_ini = 0;
}
if (!parse_affix(line, ft, afflst, dupflags)) {
finishFileMgr(afflst);
return 1;
}
}
}
finishFileMgr(afflst);
// affix trees are sorted now
// now we can speed up performance greatly taking advantage of the
// relationship between the affixes and the idea of "subsets".
// View each prefix as a potential leading subset of another and view
// each suffix (reversed) as a potential trailing subset of another.
// To illustrate this relationship if we know the prefix "ab" is found in the
// word to examine, only prefixes that "ab" is a leading subset of need be
// examined.
// Furthermore is "ab" is not present then none of the prefixes that "ab" is
// is a subset need be examined.
// The same argument goes for suffix string that are reversed.
// Then to top this off why not examine the first char of the word to quickly
// limit the set of prefixes to examine (i.e. the prefixes to examine must
// be leading supersets of the first character of the word (if they exist)
// To take advantage of this "subset" relationship, we need to add two links
// from entry. One to take next if the current prefix is found (call it
// nexteq)
// and one to take next if the current prefix is not found (call it nextne).
// Since we have built ordered lists, all that remains is to properly
// initialize
// the nextne and nexteq pointers that relate them
process_pfx_order();
process_sfx_order();
/* get encoding for CHECKCOMPOUNDCASE */
if (!utf8) {
csconv = get_current_cs(get_encoding());
for (
int i = 0; i <= 255; i++) {
if ((csconv[i].cupper != csconv[i].clower) &&
(wordchars.find((
char)i) == std::string::npos)) {
wordchars.push_back((
char)i);
}
}
}
// default BREAK definition
if (!parsedbreaktable) {
breaktable.push_back(
"-");
breaktable.push_back(
"^-");
breaktable.push_back(
"-$");
parsedbreaktable =
true;
}
return 0;
}
// we want to be able to quickly access prefix information
// both by prefix flag, and sorted by prefix string itself
// so we need to set up two indexes
int AffixMgr::build_pfxtree(PfxEntry* pfxptr) {
PfxEntry* ptr;
PfxEntry* pptr;
PfxEntry* ep = pfxptr;
// get the right starting points
const char* key = ep->getKey();
const unsigned char flg = (
unsigned char)(ep->getFlag() & 0x00FF);
// first index by flag which must exist
ptr = pFlag[flg];
ep->setFlgNxt(ptr);
pFlag[flg] = ep;
// handle the special case of null affix string
if (strlen(key) == 0) {
// always inset them at head of list at element 0
ptr = pStart[0];
ep->setNext(ptr);
pStart[0] = ep;
return 0;
}
// now handle the normal case
ep->setNextEQ(NULL);
ep->setNextNE(NULL);
unsigned char sp = *((
const unsigned char*)key);
ptr = pStart[sp];
// handle the first insert
if (!ptr) {
pStart[sp] = ep;
return 0;
}
// otherwise use binary tree insertion so that a sorted
// list can easily be generated later
pptr = NULL;
for (;;) {
pptr = ptr;
if (strcmp(ep->getKey(), ptr->getKey()) <= 0) {
ptr = ptr->getNextEQ();
if (!ptr) {
pptr->setNextEQ(ep);
break;
}
}
else {
ptr = ptr->getNextNE();
if (!ptr) {
pptr->setNextNE(ep);
break;
}
}
}
return 0;
}
// we want to be able to quickly access suffix information
// both by suffix flag, and sorted by the reverse of the
// suffix string itself; so we need to set up two indexes
int AffixMgr::build_sfxtree(SfxEntry* sfxptr) {
sfxptr->initReverseWord();
SfxEntry* ptr;
SfxEntry* pptr;
SfxEntry* ep = sfxptr;
/* get the right starting point */
const char* key = ep->getKey();
const unsigned char flg = (
unsigned char)(ep->getFlag() & 0x00FF);
// first index by flag which must exist
ptr = sFlag[flg];
ep->setFlgNxt(ptr);
sFlag[flg] = ep;
// next index by affix string
// handle the special case of null affix string
if (strlen(key) == 0) {
// always inset them at head of list at element 0
ptr = sStart[0];
ep->setNext(ptr);
sStart[0] = ep;
return 0;
}
// now handle the normal case
ep->setNextEQ(NULL);
ep->setNextNE(NULL);
unsigned char sp = *((
const unsigned char*)key);
ptr = sStart[sp];
// handle the first insert
if (!ptr) {
sStart[sp] = ep;
return 0;
}
// otherwise use binary tree insertion so that a sorted
// list can easily be generated later
pptr = NULL;
for (;;) {
pptr = ptr;
if (strcmp(ep->getKey(), ptr->getKey()) <= 0) {
ptr = ptr->getNextEQ();
if (!ptr) {
pptr->setNextEQ(ep);
break;
}
}
else {
ptr = ptr->getNextNE();
if (!ptr) {
pptr->setNextNE(ep);
break;
}
}
}
return 0;
}
// convert from binary tree to sorted list
int AffixMgr::process_pfx_tree_to_list() {
for (
int i = 1; i < SETSIZE; i++) {
pStart[i] = process_pfx_in_order(pStart[i], NULL);
}
return 0;
}
PfxEntry* AffixMgr::process_pfx_in_order(PfxEntry* ptr, PfxEntry* nptr) {
if (ptr) {
nptr = process_pfx_in_order(ptr->getNextNE(), nptr);
ptr->setNext(nptr);
nptr = process_pfx_in_order(ptr->getNextEQ(), ptr);
}
return nptr;
}
// convert from binary tree to sorted list
int AffixMgr::process_sfx_tree_to_list() {
for (
int i = 1; i < SETSIZE; i++) {
sStart[i] = process_sfx_in_order(sStart[i], NULL);
}
return 0;
}
SfxEntry* AffixMgr::process_sfx_in_order(SfxEntry* ptr, SfxEntry* nptr) {
if (ptr) {
nptr = process_sfx_in_order(ptr->getNextNE(), nptr);
ptr->setNext(nptr);
nptr = process_sfx_in_order(ptr->getNextEQ(), ptr);
}
return nptr;
}
// reinitialize the PfxEntry links NextEQ and NextNE to speed searching
// using the idea of leading subsets this time
int AffixMgr::process_pfx_order() {
PfxEntry* ptr;
// loop through each prefix list starting point
for (
int i = 1; i < SETSIZE; i++) {
ptr = pStart[i];
// look through the remainder of the list
// and find next entry with affix that
// the current one is not a subset of
// mark that as destination for NextNE
// use next in list that you are a subset
// of as NextEQ
for (; ptr != NULL; ptr = ptr->getNext()) {
PfxEntry* nptr = ptr->getNext();
for (; nptr != NULL; nptr = nptr->getNext()) {
if (!isSubset(ptr->getKey(), nptr->getKey()))
break;
}
ptr->setNextNE(nptr);
ptr->setNextEQ(NULL);
if ((ptr->getNext()) &&
isSubset(ptr->getKey(), (ptr->getNext())->getKey()))
ptr->setNextEQ(ptr->getNext());
}
// now clean up by adding smart search termination strings:
// if you are already a superset of the previous prefix
// but not a subset of the next, search can end here
// so set NextNE properly
ptr = pStart[i];
for (; ptr != NULL; ptr = ptr->getNext()) {
PfxEntry* nptr = ptr->getNext();
PfxEntry* mptr = NULL;
for (; nptr != NULL; nptr = nptr->getNext()) {
if (!isSubset(ptr->getKey(), nptr->getKey()))
break;
mptr = nptr;
}
if (mptr)
mptr->setNextNE(NULL);
}
}
return 0;
}
// initialize the SfxEntry links NextEQ and NextNE to speed searching
// using the idea of leading subsets this time
int AffixMgr::process_sfx_order() {
SfxEntry* ptr;
// loop through each prefix list starting point
for (
int i = 1; i < SETSIZE; i++) {
ptr = sStart[i];
// look through the remainder of the list
// and find next entry with affix that
// the current one is not a subset of
// mark that as destination for NextNE
// use next in list that you are a subset
// of as NextEQ
for (; ptr != NULL; ptr = ptr->getNext()) {
SfxEntry* nptr = ptr->getNext();
for (; nptr != NULL; nptr = nptr->getNext()) {
if (!isSubset(ptr->getKey(), nptr->getKey()))
break;
}
ptr->setNextNE(nptr);
ptr->setNextEQ(NULL);
if ((ptr->getNext()) &&
isSubset(ptr->getKey(), (ptr->getNext())->getKey()))
ptr->setNextEQ(ptr->getNext());
}
// now clean up by adding smart search termination strings:
// if you are already a superset of the previous suffix
// but not a subset of the next, search can end here
// so set NextNE properly
ptr = sStart[i];
for (; ptr != NULL; ptr = ptr->getNext()) {
SfxEntry* nptr = ptr->getNext();
SfxEntry* mptr = NULL;
for (; nptr != NULL; nptr = nptr->getNext()) {
if (!isSubset(ptr->getKey(), nptr->getKey()))
break;
mptr = nptr;
}
if (mptr)
mptr->setNextNE(NULL);
}
}
return 0;
}
// add flags to the result for dictionary debugging
std::string& AffixMgr::debugflag(std::string& result,
unsigned short flag) {
char* st = encode_flag(flag);
result.push_back(MSEP_FLD);
result.append(MORPH_FLAG);
if (st) {
result.append(st);
free(st);
}
return result;
}
// calculate the character length of the condition
int AffixMgr::condlen(
const char* st) {
int l = 0;
bool group =
false;
for (; *st; st++) {
if (*st ==
'[') {
group =
true;
l++;
}
else if (*st ==
']')
group =
false;
else if (!group && (!utf8 || (!(*st & 0x80) || ((*st & 0xc0) == 0x80))))
l++;
}
return l;
}
int AffixMgr::encodeit(AffEntry& entry,
const char* cs) {
if (strcmp(cs,
".") != 0) {
entry.numconds = (
char)condlen(cs);
const size_t cslen = strlen(cs);
const size_t short_part = std::min<size_t>(MAXCONDLEN, cslen);
memcpy(entry.c.conds, cs, short_part);
if (short_part < MAXCONDLEN) {
//blank out the remaining space
memset(entry.c.conds + short_part, 0, MAXCONDLEN - short_part);
}
else if (cs[MAXCONDLEN]) {
//there is more conditions than fit in fixed space, so its
//a long condition
entry.opts |= aeLONGCOND;
entry.c.l.conds2 = mystrdup(cs + MAXCONDLEN_1);
if (!entry.c.l.conds2)
return 1;
}
}
else {
entry.numconds = 0;
entry.c.conds[0] =
'\0';
}
return 0;
}
// return 1 if s1 is a leading subset of s2 (dots are for infixes)
inline int AffixMgr::isSubset(
const char* s1,
const char* s2) {
while (((*s1 == *s2) || (*s1 ==
'.')) && (*s1 !=
'\0')) {
s1++;
s2++;
}
return (*s1 ==
'\0');
}
// check word for prefixes
struct hentry* AffixMgr::prefix_check(
const char* word,
int len,
char in_compound,
const FLAG needflag) {
struct hentry* rv = NULL;
pfx = NULL;
pfxappnd = NULL;
sfxappnd = NULL;
sfxextra = 0;
// first handle the special case of 0 length prefixes
PfxEntry* pe = pStart[0];
while (pe) {
if (
// fogemorpheme
((in_compound != IN_CPD_NOT) ||
!(pe->getCont() &&
(TESTAFF(pe->getCont(), onlyincompound, pe->getContLen())))) &&
// permit prefixes in compounds
((in_compound != IN_CPD_END) ||
(pe->getCont() &&
(TESTAFF(pe->getCont(), compoundpermitflag, pe->getContLen()))))) {
// check prefix
rv = pe->checkword(word, len, in_compound, needflag);
if (rv) {
pfx = pe;
// BUG: pfx not stateless
return rv;
}
}
pe = pe->getNext();
}
// now handle the general case
unsigned char sp = *((
const unsigned char*)word);
PfxEntry* pptr = pStart[sp];
while (pptr) {
if (isSubset(pptr->getKey(), word)) {
if (
// fogemorpheme
((in_compound != IN_CPD_NOT) ||
!(pptr->getCont() &&
(TESTAFF(pptr->getCont(), onlyincompound, pptr->getContLen())))) &&
// permit prefixes in compounds
((in_compound != IN_CPD_END) ||
(pptr->getCont() && (TESTAFF(pptr->getCont(), compoundpermitflag,
pptr->getContLen()))))) {
// check prefix
rv = pptr->checkword(word, len, in_compound, needflag);
if (rv) {
pfx = pptr;
// BUG: pfx not stateless
return rv;
}
}
pptr = pptr->getNextEQ();
}
else {
pptr = pptr->getNextNE();
}
}
return NULL;
}
// check word for prefixes and two-level suffixes
struct hentry* AffixMgr::prefix_check_twosfx(
const char* word,
int len,
char in_compound,
const FLAG needflag) {
struct hentry* rv = NULL;
pfx = NULL;
sfxappnd = NULL;
sfxextra = 0;
// first handle the special case of 0 length prefixes
PfxEntry* pe = pStart[0];
while (pe) {
rv = pe->check_twosfx(word, len, in_compound, needflag);
if (rv)
return rv;
pe = pe->getNext();
}
// now handle the general case
unsigned char sp = *((
const unsigned char*)word);
PfxEntry* pptr = pStart[sp];
while (pptr) {
if (isSubset(pptr->getKey(), word)) {
rv = pptr->check_twosfx(word, len, in_compound, needflag);
if (rv) {
pfx = pptr;
return rv;
}
pptr = pptr->getNextEQ();
}
else {
pptr = pptr->getNextNE();
}
}
return NULL;
}
// check word for prefixes and morph
std::string AffixMgr::prefix_check_morph(
const char* word,
int len,
char in_compound,
const FLAG needflag) {
std::string result;
pfx = NULL;
sfxappnd = NULL;
sfxextra = 0;
// first handle the special case of 0 length prefixes
PfxEntry* pe = pStart[0];
while (pe) {
std::string st = pe->check_morph(word, len, in_compound, needflag);
if (!st.empty()) {
result.append(st);
}
pe = pe->getNext();
}
// now handle the general case
unsigned char sp = *((
const unsigned char*)word);
PfxEntry* pptr = pStart[sp];
while (pptr) {
if (isSubset(pptr->getKey(), word)) {
std::string st = pptr->check_morph(word, len, in_compound, needflag);
if (!st.empty()) {
// fogemorpheme
if ((in_compound != IN_CPD_NOT) ||
!((pptr->getCont() && (TESTAFF(pptr->getCont(), onlyincompound,
pptr->getContLen()))))) {
result.append(st);
pfx = pptr;
}
}
pptr = pptr->getNextEQ();
}
else {
pptr = pptr->getNextNE();
}
}
return result;
}
// check word for prefixes and morph and two-level suffixes
std::string AffixMgr::prefix_check_twosfx_morph(
const char* word,
int len,
char in_compound,
const FLAG needflag) {
std::string result;
pfx = NULL;
sfxappnd = NULL;
sfxextra = 0;
// first handle the special case of 0 length prefixes
PfxEntry* pe = pStart[0];
while (pe) {
std::string st = pe->check_twosfx_morph(word, len, in_compound, needflag);
if (!st.empty()) {
result.append(st);
}
pe = pe->getNext();
}
// now handle the general case
unsigned char sp = *((
const unsigned char*)word);
PfxEntry* pptr = pStart[sp];
while (pptr) {
if (isSubset(pptr->getKey(), word)) {
std::string st = pptr->check_twosfx_morph(word, len, in_compound, needflag);
if (!st.empty()) {
result.append(st);
pfx = pptr;
}
pptr = pptr->getNextEQ();
}
else {
pptr = pptr->getNextNE();
}
}
return result;
}
// Is word a non-compound with a REP substitution (see checkcompoundrep)?
int AffixMgr::cpdrep_check(
const char* word,
int wl) {
if ((wl < 2) || get_reptable().empty())
return 0;
for (size_t i = 0; i < get_reptable().size(); ++i) {
// use only available mid patterns
if (!get_reptable()[i].outstrings[0].empty()) {
const char* r = word;
const size_t lenp = get_reptable()[i].pattern.size();
// search every occurence of the pattern in the word
while ((r = strstr(r, get_reptable()[i].pattern.c_str())) != NULL) {
std::string candidate(word);
candidate.replace(r - word, lenp, get_reptable()[i].outstrings[0]);
if (candidate_check(candidate.c_str(), candidate.size()))
return 1;
++r;
// search for the next letter
}
}
}
return 0;
}
// forbid compound words, if they are in the dictionary as a
// word pair separated by space
int AffixMgr::cpdwordpair_check(
const char * word,
int wl) {
if (wl > 2) {
std::string candidate(word);
for (size_t i = 1; i < candidate.size(); i++) {
// go to end of the UTF-8 character
if (utf8 && ((word[i] & 0xc0) == 0x80))
continue;
candidate.insert(i, 1,
' ');
if (candidate_check(candidate.c_str(), candidate.size()))
return 1;
candidate.erase(i, 1);
}
}
return 0;
}
// forbid compoundings when there are special patterns at word bound
int AffixMgr::cpdpat_check(
const char* word,
int pos,
hentry* r1,
hentry* r2,
const char /*affixed*/) {
for (size_t i = 0; i < checkcpdtable.size(); ++i) {
size_t len;
if (isSubset(checkcpdtable[i].pattern2.c_str(), word + pos) &&
(!r1 || !checkcpdtable[i].cond ||
(r1->astr && TESTAFF(r1->astr, checkcpdtable[i].cond, r1->alen))) &&
(!r2 || !checkcpdtable[i].cond2 ||
(r2->astr && TESTAFF(r2->astr, checkcpdtable[i].cond2, r2->alen))) &&
// zero length pattern => only TESTAFF
// zero pattern (0/flag) => unmodified stem (zero affixes allowed)
(checkcpdtable[i].pattern.empty() ||
((checkcpdtable[i].pattern[0] ==
'0' && r1->blen <= pos &&
strncmp(word + pos - r1->blen, r1->word, r1->blen) == 0) ||
(checkcpdtable[i].pattern[0] !=
'0' &&
((len = checkcpdtable[i].pattern.size()) != 0) &&
strncmp(word + pos - len, checkcpdtable[i].pattern.c_str(), len) == 0)))) {
return 1;
}
}
return 0;
}
// forbid compounding with neighbouring upper and lower case characters at word
// bounds
int AffixMgr::cpdcase_check(
const char* word,
int pos) {
if (utf8) {
const char* p;
for (p = word + pos - 1; (*p & 0xc0) == 0x80; p--)
;
std::string pair(p);
std::vector<w_char> pair_u;
u8_u16(pair_u, pair);
unsigned short a = pair_u.size() > 1 ? ((pair_u[1].h << 8) + pair_u[1].l) : 0;
unsigned short b = !pair_u.empty() ? ((pair_u[0].h << 8) + pair_u[0].l) : 0;
if (((unicodetoupper(a, langnum) == a) ||
(unicodetoupper(b, langnum) == b)) &&
(a !=
'-') && (b !=
'-'))
return 1;
}
else {
unsigned char a = *(word + pos - 1);
unsigned char b = *(word + pos);
if ((csconv[a].ccase || csconv[b].ccase) && (a !=
'-') && (b !=
'-'))
return 1;
}
return 0;
}
struct metachar_data {
signed short btpp;
// metacharacter (*, ?) position for backtracking
signed short btwp;
// word position for metacharacters
int btnum;
// number of matched characters in metacharacter
};
// check compound patterns
int AffixMgr::defcpd_check(hentry*** words,
short wnum,
hentry* rv,
hentry** def,
char all) {
int w = 0;
if (!*words) {
w = 1;
*words = def;
}
if (!*words) {
return 0;
}
std::vector<metachar_data> btinfo(1);
short bt = 0;
(*words)[wnum] = rv;
// has the last word COMPOUNDRULE flag?
if (rv->alen == 0) {
(*words)[wnum] = NULL;
if (w)
*words = NULL;
return 0;
}
int ok = 0;
for (size_t i = 0; i < defcpdtable.size(); ++i) {
for (size_t j = 0; j < defcpdtable[i].size(); ++j) {
if (defcpdtable[i][j] !=
'*' && defcpdtable[i][j] !=
'?' &&
TESTAFF(rv->astr, defcpdtable[i][j], rv->alen)) {
ok = 1;
break;
}
}
}
if (ok == 0) {
(*words)[wnum] = NULL;
if (w)
*words = NULL;
return 0;
}
for (size_t i = 0; i < defcpdtable.size(); ++i) {
size_t pp = 0;
// pattern position
signed short wp = 0;
// "words" position
int ok2;
ok = 1;
ok2 = 1;
do {
while ((pp < defcpdtable[i].size()) && (wp <= wnum)) {
if (((pp + 1) < defcpdtable[i].size()) &&
((defcpdtable[i][pp + 1] ==
'*') ||
(defcpdtable[i][pp + 1] ==
'?'))) {
int wend = (defcpdtable[i][pp + 1] ==
'?') ? wp : wnum;
ok2 = 1;
pp += 2;
btinfo[bt].btpp = pp;
btinfo[bt].btwp = wp;
while (wp <= wend) {
if (!(*words)[wp]->alen ||
!TESTAFF((*words)[wp]->astr, defcpdtable[i][pp - 2],
(*words)[wp]->alen)) {
ok2 = 0;
break;
}
wp++;
}
if (wp <= wnum)
ok2 = 0;
btinfo[bt].btnum = wp - btinfo[bt].btwp;
if (btinfo[bt].btnum > 0) {
++bt;
btinfo.resize(bt+1);
}
if (ok2)
break;
}
else {
ok2 = 1;
if (!(*words)[wp] || !(*words)[wp]->alen ||
!TESTAFF((*words)[wp]->astr, defcpdtable[i][pp],
(*words)[wp]->alen)) {
ok = 0;
break;
}
pp++;
wp++;
if ((defcpdtable[i].size() == pp) && !(wp > wnum))
ok = 0;
}
}
if (ok && ok2) {
size_t r = pp;
while ((defcpdtable[i].size() > r) && ((r + 1) < defcpdtable[i].size()) &&
((defcpdtable[i][r + 1] ==
'*') ||
(defcpdtable[i][r + 1] ==
'?')))
r += 2;
if (defcpdtable[i].size() <= r)
return 1;
}
// backtrack
if (bt)
do {
ok = 1;
btinfo[bt - 1].btnum--;
pp = btinfo[bt - 1].btpp;
wp = btinfo[bt - 1].btwp + (
signed short)btinfo[bt - 1].btnum;
}
while ((btinfo[bt - 1].btnum < 0) && --bt);
}
while (bt);
if (ok && ok2 && (!all || (defcpdtable[i].size() <= pp)))
return 1;
// check zero ending
while (ok && ok2 && (defcpdtable[i].size() > pp) &&
((pp + 1) < defcpdtable[i].size()) &&
((defcpdtable[i][pp + 1] ==
'*') ||
(defcpdtable[i][pp + 1] ==
'?')))
pp += 2;
if (ok && ok2 && (defcpdtable[i].size() <= pp))
return 1;
}
(*words)[wnum] = NULL;
if (w)
*words = NULL;
return 0;
}
inline int AffixMgr::candidate_check(
const char* word,
int len) {
struct hentry* rv = lookup(word);
if (rv)
return 1;
// rv = prefix_check(word,len,1);
// if (rv) return 1;
rv = affix_check(word, len);
if (rv)
return 1;
return 0;
}
// calculate number of syllable for compound-checking
short AffixMgr::get_syllable(
const std::string& word) {
if (cpdmaxsyllable == 0)
return 0;
short num = 0;
if (!utf8) {
for (size_t i = 0; i < word.size(); ++i) {
if (std::binary_search(cpdvowels.begin(), cpdvowels.end(),
word[i])) {
++num;
}
}
}
else if (!cpdvowels_utf16.empty()) {
std::vector<w_char> w;
u8_u16(w, word);
for (size_t i = 0; i < w.size(); ++i) {
if (std::binary_search(cpdvowels_utf16.begin(),
cpdvowels_utf16.end(),
w[i])) {
++num;
}
}
}
return num;
}
void AffixMgr::setcminmax(
int* cmin,
int* cmax,
const char* word,
int len) {
if (utf8) {
int i;
for (*cmin = 0, i = 0; (i < cpdmin) && *cmin < len; i++) {
for ((*cmin)++; *cmin < len && (word[*cmin] & 0xc0) == 0x80; (*cmin)++)
;
}
for (*cmax = len, i = 0; (i < (cpdmin - 1)) && *cmax >= 0; i++) {
for ((*cmax)--; *cmax >= 0 && (word[*cmax] & 0xc0) == 0x80; (*cmax)--)
;
}
}
else {
*cmin = cpdmin;
*cmax = len - cpdmin + 1;
}
}
// check if compound word is correctly spelled
// hu_mov_rule = spec. Hungarian rule (XXX)
struct hentry* AffixMgr::compound_check(
const std::string& word,
short wordnum,
short numsyllable,
short maxwordnum,
short wnum,
hentry** words = NULL,
hentry** rwords = NULL,
char hu_mov_rule = 0,
char is_sug = 0,
int* info = NULL) {
int i;
short oldnumsyllable, oldnumsyllable2, oldwordnum, oldwordnum2;
struct hentry* rv = NULL;
struct hentry* rv_first;
std::string st;
char ch =
'\0';
int cmin;
int cmax;
int striple = 0;
size_t scpd = 0;
int soldi = 0;
int oldcmin = 0;
int oldcmax = 0;
int oldlen = 0;
int checkedstriple = 0;
char affixed = 0;
hentry** oldwords = words;
size_t len = word.size();
int checked_prefix;
// add a time limit to handle possible
// combinatorical explosion of the overlapping words
HUNSPELL_THREAD_LOCAL clock_t timelimit;
if (wordnum == 0) {
// get the start time, seeing as we're reusing this set to 0
// to flag timeout, use clock() + 1 to avoid start clock()
// of 0 as being a timeout
timelimit = clock() + 1;
}
else if (timelimit != 0 && (clock() > timelimit + TIMELIMIT)) {
timelimit = 0;
}
setcminmax(&cmin, &cmax, word.c_str(), len);
st.assign(word);
for (i = cmin; i < cmax; i++) {
// go to end of the UTF-8 character
if (utf8) {
for (; (st[i] & 0xc0) == 0x80; i++)
;
if (i >= cmax)
return NULL;
}
words = oldwords;
int onlycpdrule = (words) ? 1 : 0;
do {
// onlycpdrule loop
oldnumsyllable = numsyllable;
oldwordnum = wordnum;
checked_prefix = 0;
do {
// simplified checkcompoundpattern loop
if (timelimit == 0)
return 0;
if (scpd > 0) {
for (; scpd <= checkcpdtable.size() &&
(checkcpdtable[scpd - 1].pattern3.empty() ||
strncmp(word.c_str() + i, checkcpdtable[scpd - 1].pattern3.c_str(),
checkcpdtable[scpd - 1].pattern3.size()) != 0);
scpd++)
;
if (scpd > checkcpdtable.size())
break;
// break simplified checkcompoundpattern loop
st.replace(i, std::string::npos, checkcpdtable[scpd - 1].pattern);
soldi = i;
i += checkcpdtable[scpd - 1].pattern.size();
st.replace(i, std::string::npos, checkcpdtable[scpd - 1].pattern2);
st.replace(i + checkcpdtable[scpd - 1].pattern2.size(), std::string::npos,
word.substr(soldi + checkcpdtable[scpd - 1].pattern3.size()));
oldlen = len;
len += checkcpdtable[scpd - 1].pattern.size() +
checkcpdtable[scpd - 1].pattern2.size() -
checkcpdtable[scpd - 1].pattern3.size();
oldcmin = cmin;
oldcmax = cmax;
setcminmax(&cmin, &cmax, st.c_str(), len);
cmax = len - cpdmin + 1;
}
ch = st[i];
st[i] =
'\0';
sfx = NULL;
pfx = NULL;
// FIRST WORD
affixed = 1;
rv = lookup(st.c_str());
// perhaps without prefix
// forbid dictionary stems with COMPOUNDFORBIDFLAG in
// compound words, overriding the effect of COMPOUNDPERMITFLAG
if ((rv) && compoundforbidflag &&
TESTAFF(rv->astr, compoundforbidflag, rv->alen) && !hu_mov_rule)
continue;
// search homonym with compound flag
while ((rv) && !hu_mov_rule &&
((needaffix && TESTAFF(rv->astr, needaffix, rv->alen)) ||
!((compoundflag && !words && !onlycpdrule &&
TESTAFF(rv->astr, compoundflag, rv->alen)) ||
(compoundbegin && !wordnum && !onlycpdrule &&
TESTAFF(rv->astr, compoundbegin, rv->alen)) ||
(compoundmiddle && wordnum && !words && !onlycpdrule &&
TESTAFF(rv->astr, compoundmiddle, rv->alen)) ||
(!defcpdtable.empty() && onlycpdrule &&
((!words && !wordnum &&
defcpd_check(&words, wnum, rv, rwords, 0)) ||
(words &&
defcpd_check(&words, wnum, rv, rwords, 0))))) ||
(scpd != 0 && checkcpdtable[scpd - 1].cond != FLAG_NULL &&
!TESTAFF(rv->astr, checkcpdtable[scpd - 1].cond, rv->alen)))) {
rv = rv->next_homonym;
}
if (rv)
affixed = 0;
if (!rv) {
if (onlycpdrule)
break;
if (compoundflag &&
!(rv = prefix_check(st.c_str(), i,
hu_mov_rule ? IN_CPD_OTHER : IN_CPD_BEGIN,
compoundflag))) {
if (((rv = suffix_check(
st.c_str(), i, 0, NULL, FLAG_NULL, compoundflag,
hu_mov_rule ? IN_CPD_OTHER : IN_CPD_BEGIN)) ||
(compoundmoresuffixes &&
(rv = suffix_check_twosfx(st.c_str(), i, 0, NULL, compoundflag)))) &&
!hu_mov_rule && sfx->getCont() &&
((compoundforbidflag &&
TESTAFF(sfx->getCont(), compoundforbidflag,
sfx->getContLen())) ||
(compoundend &&
TESTAFF(sfx->getCont(), compoundend, sfx->getContLen())))) {
rv = NULL;
}
}
if (rv ||
(((wordnum == 0) && compoundbegin &&
((rv = suffix_check(
st.c_str(), i, 0, NULL, FLAG_NULL, compoundbegin,
hu_mov_rule ? IN_CPD_OTHER : IN_CPD_BEGIN)) ||
(compoundmoresuffixes &&
(rv = suffix_check_twosfx(
st.c_str(), i, 0, NULL,
compoundbegin))) ||
// twofold suffixes + compound
(rv = prefix_check(st.c_str(), i,
hu_mov_rule ? IN_CPD_OTHER : IN_CPD_BEGIN,
compoundbegin)))) ||
((wordnum > 0) && compoundmiddle &&
((rv = suffix_check(
st.c_str(), i, 0, NULL, FLAG_NULL, compoundmiddle,
hu_mov_rule ? IN_CPD_OTHER : IN_CPD_BEGIN)) ||
(compoundmoresuffixes &&
(rv = suffix_check_twosfx(
st.c_str(), i, 0, NULL,
compoundmiddle))) ||
// twofold suffixes + compound
(rv = prefix_check(st.c_str(), i,
hu_mov_rule ? IN_CPD_OTHER : IN_CPD_BEGIN,
compoundmiddle))))))
checked_prefix = 1;
// else check forbiddenwords and needaffix
}
else if (rv->astr && (TESTAFF(rv->astr, forbiddenword, rv->alen) ||
TESTAFF(rv->astr, needaffix, rv->alen) ||
TESTAFF(rv->astr, ONLYUPCASEFLAG, rv->alen) ||
(is_sug && nosuggest &&
TESTAFF(rv->astr, nosuggest, rv->alen)))) {
st[i] = ch;
// continue;
break;
}
// check non_compound flag in suffix and prefix
if ((rv) && !hu_mov_rule &&
((pfx && pfx->getCont() &&
TESTAFF(pfx->getCont(), compoundforbidflag, pfx->getContLen())) ||
(sfx && sfx->getCont() &&
TESTAFF(sfx->getCont(), compoundforbidflag,
sfx->getContLen())))) {
rv = NULL;
}
// check compoundend flag in suffix and prefix
if ((rv) && !checked_prefix && compoundend && !hu_mov_rule &&
((pfx && pfx->getCont() &&
TESTAFF(pfx->getCont(), compoundend, pfx->getContLen())) ||
(sfx && sfx->getCont() &&
TESTAFF(sfx->getCont(), compoundend, sfx->getContLen())))) {
rv = NULL;
}
// check compoundmiddle flag in suffix and prefix
if ((rv) && !checked_prefix && (wordnum == 0) && compoundmiddle &&
!hu_mov_rule &&
((pfx && pfx->getCont() &&
TESTAFF(pfx->getCont(), compoundmiddle, pfx->getContLen())) ||
(sfx && sfx->getCont() &&
TESTAFF(sfx->getCont(), compoundmiddle, sfx->getContLen())))) {
rv = NULL;
}
// check forbiddenwords
if ((rv) && (rv->astr) &&
(TESTAFF(rv->astr, forbiddenword, rv->alen) ||
TESTAFF(rv->astr, ONLYUPCASEFLAG, rv->alen) ||
(is_sug && nosuggest && TESTAFF(rv->astr, nosuggest, rv->alen)))) {
return NULL;
}
// increment word number, if the second root has a compoundroot flag
if ((rv) && compoundroot &&
(TESTAFF(rv->astr, compoundroot, rv->alen))) {
wordnum++;
}
// first word is acceptable in compound words?
if (((rv) &&
(checked_prefix || (words && words[wnum]) ||
(compoundflag && TESTAFF(rv->astr, compoundflag, rv->alen)) ||
((oldwordnum == 0) && compoundbegin &&
TESTAFF(rv->astr, compoundbegin, rv->alen)) ||
((oldwordnum > 0) && compoundmiddle &&
TESTAFF(rv->astr, compoundmiddle, rv->alen))
// LANG_hu section: spec. Hungarian rule
|| ((langnum == LANG_hu) && hu_mov_rule &&
(TESTAFF(
rv->astr,
'F',
rv->alen) ||
// XXX hardwired Hungarian dictionary codes
TESTAFF(rv->astr,
'G', rv->alen) ||
TESTAFF(rv->astr,
'H', rv->alen)))
// END of LANG_hu section
) &&
(
// test CHECKCOMPOUNDPATTERN conditions
scpd == 0 || checkcpdtable[scpd - 1].cond == FLAG_NULL ||
TESTAFF(rv->astr, checkcpdtable[scpd - 1].cond, rv->alen)) &&
!((checkcompoundtriple && scpd == 0 &&
!words &&
// test triple letters
(word[i - 1] == word[i]) &&
(((i > 1) && (word[i - 1] == word[i - 2])) ||
((word[i - 1] == word[i + 1]))
// may be word[i+1] == '\0'
)) ||
(checkcompoundcase && scpd == 0 && !words &&
cpdcase_check(word.c_str(), i))))
// LANG_hu section: spec. Hungarian rule
|| ((!rv) && (langnum == LANG_hu) && hu_mov_rule &&
(rv = affix_check(st.c_str(), i)) &&
(sfx && sfx->getCont() &&
(
// XXX hardwired Hungarian dic. codes
TESTAFF(sfx->getCont(), (
unsigned short)
'x',
sfx->getContLen()) ||
TESTAFF(
sfx->getCont(), (
unsigned short)
'%',
sfx->getContLen()))))) {
// first word is ok condition
// LANG_hu section: spec. Hungarian rule
if (langnum == LANG_hu) {
// calculate syllable number of the word
numsyllable += get_syllable(st.substr(0, i));
// + 1 word, if syllable number of the prefix > 1 (hungarian
// convention)
if (pfx && (get_syllable(pfx->getKey()) > 1))
wordnum++;
}
// END of LANG_hu section
// NEXT WORD(S)
rv_first = rv;
st[i] = ch;
do {
// striple loop
// check simplifiedtriple
if (simplifiedtriple) {
if (striple) {
checkedstriple = 1;
i--;
// check "fahrt" instead of "ahrt" in "Schiffahrt"
}
else if (i > 2 && word[i - 1] == word[i - 2])
striple = 1;
}
rv = lookup(st.c_str() + i);
// perhaps without prefix
// search homonym with compound flag
while ((rv) &&
((needaffix && TESTAFF(rv->astr, needaffix, rv->alen)) ||
!((compoundflag && !words &&
TESTAFF(rv->astr, compoundflag, rv->alen)) ||
(compoundend && !words &&
TESTAFF(rv->astr, compoundend, rv->alen)) ||
(!defcpdtable.empty() && words &&
defcpd_check(&words, wnum + 1, rv, NULL, 1))) ||
(scpd != 0 && checkcpdtable[scpd - 1].cond2 != FLAG_NULL &&
!TESTAFF(rv->astr, checkcpdtable[scpd - 1].cond2,
rv->alen)))) {
rv = rv->next_homonym;
}
// check FORCEUCASE
if (rv && forceucase && (rv) &&
(TESTAFF(rv->astr, forceucase, rv->alen)) &&
!(info && *info & SPELL_ORIGCAP))
rv = NULL;
if (rv && words && words[wnum + 1])
return rv_first;
oldnumsyllable2 = numsyllable;
oldwordnum2 = wordnum;
// LANG_hu section: spec. Hungarian rule, XXX hardwired dictionary
// code
if ((rv) && (langnum == LANG_hu) &&
(TESTAFF(rv->astr,
'I', rv->alen)) &&
!(TESTAFF(rv->astr,
'J', rv->alen))) {
numsyllable--;
}
// END of LANG_hu section
// increment word number, if the second root has a compoundroot flag
if ((rv) && (compoundroot) &&
(TESTAFF(rv->astr, compoundroot, rv->alen))) {
wordnum++;
}
// check forbiddenwords
if ((rv) && (rv->astr) &&
(TESTAFF(rv->astr, forbiddenword, rv->alen) ||
TESTAFF(rv->astr, ONLYUPCASEFLAG, rv->alen) ||
(is_sug && nosuggest &&
TESTAFF(rv->astr, nosuggest, rv->alen))))
return NULL;
// second word is acceptable, as a root?
// hungarian conventions: compounding is acceptable,
// when compound forms consist of 2 words, or if more,
// then the syllable number of root words must be 6, or lesser.
if ((rv) &&
((compoundflag && TESTAFF(rv->astr, compoundflag, rv->alen)) ||
(compoundend && TESTAFF(rv->astr, compoundend, rv->alen))) &&
(((cpdwordmax == -1) || (wordnum + 1 < cpdwordmax)) ||
((cpdmaxsyllable != 0) &&
(numsyllable + get_syllable(std::string(HENTRY_WORD(rv), rv->blen)) <=
cpdmaxsyllable))) &&
(
// test CHECKCOMPOUNDPATTERN
checkcpdtable.empty() || scpd != 0 ||
!cpdpat_check(word.c_str(), i, rv_first, rv, 0)) &&
((!checkcompounddup || (rv != rv_first)))
// test CHECKCOMPOUNDPATTERN conditions
&&
(scpd == 0 || checkcpdtable[scpd - 1].cond2 == FLAG_NULL ||
TESTAFF(rv->astr, checkcpdtable[scpd - 1].cond2, rv->alen))) {
// forbid compound word, if it is a non-compound word with typical
// fault
if ((checkcompoundrep && cpdrep_check(word.c_str(), len)) ||
cpdwordpair_check(word.c_str(), len))
return NULL;
return rv_first;
}
numsyllable = oldnumsyllable2;
wordnum = oldwordnum2;
// perhaps second word has prefix or/and suffix
sfx = NULL;
sfxflag = FLAG_NULL;
rv = (compoundflag && !onlycpdrule)
? affix_check((word.c_str() + i), strlen(word.c_str() + i), compoundflag,
IN_CPD_END)
: NULL;
if (!rv && compoundend && !onlycpdrule) {
sfx = NULL;
pfx = NULL;
rv = affix_check((word.c_str() + i), strlen(word.c_str() + i), compoundend,
IN_CPD_END);
}
if (!rv && !defcpdtable.empty() && words) {
rv = affix_check((word.c_str() + i), strlen(word.c_str() + i), 0, IN_CPD_END);
if (rv && defcpd_check(&words, wnum + 1, rv, NULL, 1))
return rv_first;
rv = NULL;
}
// test CHECKCOMPOUNDPATTERN conditions (allowed forms)
if (rv &&
!(scpd == 0 || checkcpdtable[scpd - 1].cond2 == FLAG_NULL ||
TESTAFF(rv->astr, checkcpdtable[scpd - 1].cond2, rv->alen)))
rv = NULL;
// test CHECKCOMPOUNDPATTERN conditions (forbidden compounds)
if (rv && !checkcpdtable.empty() && scpd == 0 &&
cpdpat_check(word.c_str(), i, rv_first, rv, affixed))
rv = NULL;
// check non_compound flag in suffix and prefix
if ((rv) && ((pfx && pfx->getCont() &&
TESTAFF(pfx->getCont(), compoundforbidflag,
pfx->getContLen())) ||
(sfx && sfx->getCont() &&
TESTAFF(sfx->getCont(), compoundforbidflag,
sfx->getContLen())))) {
rv = NULL;
}
// check FORCEUCASE
if (rv && forceucase && (rv) &&
(TESTAFF(rv->astr, forceucase, rv->alen)) &&
!(info && *info & SPELL_ORIGCAP))
rv = NULL;
// check forbiddenwords
if ((rv) && (rv->astr) &&
(TESTAFF(rv->astr, forbiddenword, rv->alen) ||
TESTAFF(rv->astr, ONLYUPCASEFLAG, rv->alen) ||
(is_sug && nosuggest &&
TESTAFF(rv->astr, nosuggest, rv->alen))))
return NULL;
// pfxappnd = prefix of word+i, or NULL
// calculate syllable number of prefix.
// hungarian convention: when syllable number of prefix is more,
// than 1, the prefix+word counts as two words.
if (langnum == LANG_hu) {
// calculate syllable number of the word
numsyllable += get_syllable(word.c_str() + i);
// - affix syllable num.
// XXX only second suffix (inflections, not derivations)
if (sfxappnd) {
std::string tmp(sfxappnd);
reverseword(tmp);
numsyllable -=
short(get_syllable(tmp) + sfxextra);
}
else {
numsyllable -=
short(sfxextra);
}
// + 1 word, if syllable number of the prefix > 1 (hungarian
// convention)
if (pfx && (get_syllable(pfx->getKey()) > 1))
wordnum++;
// increment syllable num, if last word has a SYLLABLENUM flag
// and the suffix is beginning `s'
if (!cpdsyllablenum.empty()) {
switch (sfxflag) {
case 'c': {
numsyllable += 2;
break;
}
case 'J': {
numsyllable += 1;
break;
}
case 'I': {
if (rv && TESTAFF(rv->astr,
'J', rv->alen))
numsyllable += 1;
break;
}
}
}
}
// increment word number, if the second word has a compoundroot flag
if ((rv) && (compoundroot) &&
(TESTAFF(rv->astr, compoundroot, rv->alen))) {
wordnum++;
}
// second word is acceptable, as a word with prefix or/and suffix?
// hungarian conventions: compounding is acceptable,
// when compound forms consist 2 word, otherwise
--> --------------------
--> maximum size reached
--> --------------------