/* morph.c - WordNet search code morphology functions */ #include #include #include #include #ifdef HAVE_CONFIG_H #include "config.h" #endif #include "wn.h" #ifdef _WINDOWS #include #include #define EXCFILE "%s\\%s.exc" #else #define EXCFILE "%s/%s.exc" #endif __FBSDID("$Id: morph.c,v 1.67 2006/11/14 21:00:23 wn Exp $"); static const char *sufx[] ={ /* Noun suffixes */ "s", "ses", "xes", "zes", "ches", "shes", "men", "ies", /* Verb suffixes */ "s", "ies", "es", "es", "ed", "ed", "ing", "ing", /* Adjective suffixes */ "er", "est", "er", "est" }; static const char *addr[] ={ /* Noun endings */ "", "s", "x", "z", "ch", "sh", "man", "y", /* Verb endings */ "", "y", "e", "", "e", "", "e", "", /* Adjective endings */ "", "", "e", "e" }; static int offsets[NUMPARTS] = { 0, 0, 8, 16 }; static int cnts[NUMPARTS] = { 0, 8, 8, 4 }; static char msgbuf[256]; static struct { const char *str; int strlen; } prepositions[] = { { "to", 2 }, { "at", 2 }, { "of", 2 }, { "on", 2 }, { "off", 3 }, { "in", 2 }, { "out", 3 }, { "up", 2 }, { "down", 4 }, { "from", 4 }, { "with", 4 }, { "into", 4 }, { "for", 3 }, { "about", 5 }, { "between", 7 } }; #define NUMPREPS sizeof(prepositions)/sizeof(prepositions[0]) static FILE *exc_fps[NUMPARTS + 1]; static int do_init(void); static int strend(const char *, const char *); static const char *wordbase(const char *, int); static int hasprep(const char *, unsigned int); static const char *exc_lookup(const char *, int); static const char *morphprep(const char *); /* Open exception list files */ int morphinit(void) { static int done = 0; static int openerr = 0; if (!done) { if (OpenDB) { /* make sure WN database files are open */ if (!(openerr = do_init())) done = 1; } else openerr = -1; } return(openerr); } /* Close exception list files and reopen */ int re_morphinit(void) { int i; for (i = 1; i <= NUMPARTS; i++) { if (exc_fps[i] != NULL) { fclose(exc_fps[i]); exc_fps[i] = NULL; } } return(OpenDB ? do_init() : -1); } static int do_init(void) { int i, openerr; #ifdef _WINDOWS HKEY hkey; DWORD dwType, dwSize; #else char *env; #endif char searchdir[256], fname[256]; openerr = 0; /* Find base directory for database. If set, use WNSEARCHDIR. If not set, check for WNHOME/dict, otherwise use DEFAULTPATH. */ #ifdef _WINDOWS if (RegOpenKeyEx(HKEY_LOCAL_MACHINE, TEXT("Software\\WordNet\\3.0"), 0, KEY_READ, &hkey) == ERROR_SUCCESS) { dwSize = sizeof(searchdir); RegQueryValueEx(hkey, TEXT("WNHome"), NULL, &dwType, searchdir, &dwSize); RegCloseKey(hkey); strcat(searchdir, DICTDIR); } else if (RegOpenKeyEx(HKEY_CURRENT_USER, TEXT("Software\\WordNet\\3.0"), 0, KEY_READ, &hkey) == ERROR_SUCCESS) { dwSize = sizeof(searchdir); RegQueryValueEx(hkey, TEXT("WNHome"), NULL, &dwType, searchdir, &dwSize); RegCloseKey(hkey); strcat(searchdir, DICTDIR); } else sprintf(searchdir, DEFAULTPATH); #else if ((env = getenv("WNSEARCHDIR")) != NULL) strcpy(searchdir, env); else if ((env = getenv("WNHOME")) != NULL) sprintf(searchdir, "%s%s", env, DICTDIR); else strcpy(searchdir, DEFAULTPATH); #endif for (i = 1; i <= NUMPARTS; i++) { sprintf(fname, EXCFILE, searchdir, partnames[i]); if ((exc_fps[i] = fopen(fname, "r")) == NULL) { sprintf(msgbuf, "WordNet library error: Can't open exception file(%s)\n\n", fname); display_message(msgbuf); openerr = -1; } } return(openerr); } /* Try to find baseform (lemma) of word or collocation in POS. Works like strtok() - first call is with string, subsequent calls with NULL argument return additional baseforms for original string. */ const char * morphstr(const char *origstr, int pos) { static char searchstr[WORDBUF], str[WORDBUF]; static int svcnt, svprep; char word[WORDBUF]; const char *tmp; int cnt, st_idx = 0, end_idx; int prep; const char *end_idx1, *end_idx2; const char *append; if (pos == SATELLITE) pos = ADJ; /* First time through for this string */ if (origstr != NULL) { /* Assume string hasn't had spaces substitued with '_' */ strtolower(strsubst(strcpy(str, origstr), ' ', '_')); searchstr[0] = '\0'; cnt = cntwords(str, '_'); svprep = 0; /* first try exception list */ if ((tmp = exc_lookup(str, pos)) && strcmp(tmp, str)) { svcnt = 1; /* force next time to pass NULL */ return(tmp); } /* Then try simply morph on original string */ if (pos != VERB && (tmp = morphword(str, pos)) && strcmp(tmp, str)) return(tmp); if (pos == VERB && cnt > 1 && (prep = hasprep(str, cnt))) { /* assume we have a verb followed by a preposition */ svprep = prep; return(morphprep(str)); } else { svcnt = cnt = cntwords(str, '-'); while (origstr && --cnt) { end_idx1 = strchr(str + st_idx, '_'); end_idx2 = strchr(str + st_idx, '-'); if (end_idx1 && end_idx2) { if (end_idx1 < end_idx2) { end_idx = (int)(end_idx1 - str); append = "_"; } else { end_idx = (int)(end_idx2 - str); append = "-"; } } else { if (end_idx1) { end_idx = (int)(end_idx1 - str); append = "_"; } else { end_idx = (int)(end_idx2 - str); append = "-"; } } if (end_idx < 0) return(NULL); /* shouldn't do this */ strncpy(word, str + st_idx, end_idx - st_idx); word[end_idx - st_idx] = '\0'; tmp = morphword(word, pos); if(tmp) strcat(searchstr,tmp); else strcat(searchstr,word); strcat(searchstr, append); st_idx = end_idx + 1; } tmp = morphword(strcpy(word, str + st_idx), pos); if(tmp) strcat(searchstr,tmp); else strcat(searchstr,word); if(strcmp(searchstr, str) && is_defined(searchstr,pos)) return(searchstr); else return(NULL); } } else { /* subsequent call on string */ if (svprep) { /* if verb has preposition, no more morphs */ svprep = 0; return(NULL); } else if (svcnt == 1) return(exc_lookup(NULL, pos)); else { svcnt = 1; if ((tmp = exc_lookup(str, pos)) && strcmp(tmp, str)) return(tmp); else return(NULL); } } } /* Try to find baseform (lemma) of individual word in POS */ const char * morphword(const char *word, int pos) { int offset, cnt; int i; static char retval[WORDBUF]; char tmpbuf[WORDBUF]; const char *tmp, *end; retval[0] = tmpbuf[0] = '\0'; end = ""; if(word == NULL) return(NULL); /* first look for word on exception list */ if((tmp = exc_lookup(word, pos)) != NULL) return(tmp); /* found it in exception list */ if (pos == ADV) { /* only use exception list for adverbs */ return(NULL); } if (pos == NOUN) { if (strend(word, "ful")) { cnt = strrchr(word, 'f') - word; strncat(tmpbuf, word, cnt); end = "ful"; } else /* check for noun ending with 'ss' or short words */ if (strend(word, "ss") || (strlen(word) <= 2)) return(NULL); } /* If not in exception list, try applying rules from tables */ if (tmpbuf[0] == '\0') strcpy(tmpbuf, word); offset = offsets[pos]; cnt = cnts[pos]; for(i = 0; i < cnt; i++){ strcpy(retval, wordbase(tmpbuf, (i + offset))); if(strcmp(retval, tmpbuf) && is_defined(retval, pos)) { strcat(retval, end); return(retval); } } return(NULL); } static int strend(const char *str1, const char *str2) { const char *pt1; if(strlen(str2) >= strlen(str1)) return(0); else { pt1=str1; pt1=strchr(str1,0); pt1=pt1-strlen(str2); return(!strcmp(pt1,str2)); } } static const char * wordbase(const char *word, int ender) { char *pt1; static char copy[WORDBUF]; strcpy(copy, word); if(strend(copy,sufx[ender])) { pt1=strchr(copy,'\0'); pt1 -= strlen(sufx[ender]); *pt1='\0'; strcat(copy,addr[ender]); } return(copy); } static int hasprep(const char *s, unsigned int wdcnt) { /* Find a preposition in the verb string and return its corresponding word number. */ unsigned int i, wdnum; for (wdnum = 2; wdnum <= wdcnt; wdnum++) { s = strchr(s, '_'); for (s++, i = 0; i < NUMPREPS; i++) if (!strncmp(s, prepositions[i].str, prepositions[i].strlen) && (s[prepositions[i].strlen] == '_' || s[prepositions[i].strlen] == '\0')) return(wdnum); } return(0); } static const char * exc_lookup(const char *word, int pos) { static char line[WORDBUF], *beglp, *endlp; const char *excline; if (exc_fps[pos] == NULL) return(NULL); /* first time through load line from exception file */ if(word != NULL){ if ((excline = bin_search(word, exc_fps[pos])) != NULL) { strcpy(line, excline); endlp = strchr(line,' '); } else endlp = NULL; } if(endlp && *(endlp + 1) != ' '){ beglp = endlp + 1; while(*beglp && *beglp == ' ') beglp++; endlp = beglp; while(*endlp && *endlp != ' ' && *endlp != '\n') endlp++; if(endlp != beglp){ *endlp='\0'; return(beglp); } } beglp = NULL; endlp = NULL; return(NULL); } static const char * morphprep(const char *s) { const char *rest, *exc_word, *lastwd = NULL, *last; int i, offset, cnt; char word[WORDBUF], end[WORDBUF]; static char retval[WORDBUF]; /* Assume that the verb is the first word in the phrase. Strip it off, check for validity, then try various morphs with the rest of the phrase tacked on, trying to find a match. */ rest = strchr(s, '_'); last = strrchr(s, '_'); if (rest != last) { /* more than 2 words */ lastwd = morphword(last + 1, NOUN); if (lastwd) { strncpy(end, rest, last - rest + 1); end[last-rest+1] = '\0'; strcat(end, lastwd); } } strncpy(word, s, rest - s); word[rest - s] = '\0'; for (i = 0, cnt = strlen(word); i < cnt; i++) if (!isalnum((unsigned char)(word[i]))) return(NULL); offset = offsets[VERB]; cnt = cnts[VERB]; /* First try to find the verb in the exception list */ if ((exc_word = exc_lookup(word, VERB)) && strcmp(exc_word, word)) { sprintf(retval, "%s%s", exc_word, rest); if(is_defined(retval, VERB)) return(retval); else if (lastwd) { sprintf(retval, "%s%s", exc_word, end); if(is_defined(retval, VERB)) return(retval); } } for (i = 0; i < cnt; i++) { if ((exc_word = wordbase(word, (i + offset))) && strcmp(word, exc_word)) { /* ending is different */ sprintf(retval, "%s%s", exc_word, rest); if(is_defined(retval, VERB)) return(retval); else if (lastwd) { sprintf(retval, "%s%s", exc_word, end); if(is_defined(retval, VERB)) return(retval); } } } sprintf(retval, "%s%s", word, rest); if (strcmp(s, retval)) return(retval); if (lastwd) { sprintf(retval, "%s%s", word, end); if (strcmp(s, retval)) return(retval); } return(NULL); } /* * Revision 1.1 91/09/25 15:39:47 wn * Initial revision * */