pscan.c

来自「<B>Digital的Unix操作系统VAX 4.2源码</B>」· C语言 代码 · 共 965 行 · 第 1/2 页

#ifndef lint
static char sccsid[] = "@(#)pscan.c	4.2	(Berkeley)	82/11/06";
#endif not lint

#include "names.h"
#include "conp.h"
#define isadv(c)	(c == ADV || c == ADJ_ADV)
#define isadj(c)	(c == ADJ || c == NOUN_ADJ || c == ADJ_ADV || c == UNK)
#define notnoun(c)	(c != NOUN && c != UNK && c != NOUN_VERB && c != NV_PL && c != PNOUN)
#define issing(c)	(c == UNK || c == NOUN_VERB || c == NOUN || c == NOM || c == NOUN_ADJ)
#define isnoun(c)	(c == NOUN || c == NOUN_VERB || c == NV_PL || c == UNK || c == PNOUN)
#define nounnom(c)	(c == NOUN || c == NOM)
char lastc,nextc;
int savsub;
scan(si,ce,command)	/*scan a phrase */
int si; 
char ce;
{
	int savi;
	char c;
	i=si;
	if(command == 1)subty = PLURAL;
	else subty = 0;
	if(sent[i].cc==ADJ_ADV && sent[i+1].cc != NOUN && sent[i+1].cc != UNK && sent[i+1].cc != ADJ)
		sent[i++].cc = ADV;
	done=0;
	verb = 0;
	verb=getv(i,ce);	/*get verb if obvious*/
	if(command == 0)j=getnoun(i,ce);		/*get subject*/
	else j = si;
	if(i != j || ((i==j) && sent[i].cc == NOUN)){
		i = j+1;
	}
	for(;((c=sent[i].cc) != ce) && (c != END)  && (c != '\0');i++){
		nextc=sent[i+1].cc;
		if(i>0){
			lastc=sent[i-1].cc;
			if(lastc==BE)be=1;
			else{
				if(lastc != ADV)be=0;
			}
		}
		else lastc=0;
		if(verb==1)question=0;
		switch(c){
		case '"':
			if(nextc==ED || nextc == MVERB){
				verb=1;
				sent[++i].cc=VERB;
				subty=SING;
				continue;
			}
			subty=0;
			verb=getv(++i,ce);
			i=getnoun(i,ce);
			continue;
		case MVERB:
			sent[i].cc = VERB;
			if(i < nsleng*.5){
				verb = getv(++i,ce);
				i = getnoun(i,ce);
				continue;
			}
			continue;
		case ART:
		case ADJ:
		case POS:	
		case ING:
			aflg=1;
			i=getnoun(i,ce);
			aflg=0;
			continue;
		case PREP:	
			if(nextc == ce){
				sent[i].cc=ADV;
				goto sdone;
			}
			prep=1;
			i=getnoun(++i,ce);
			prep=0;
			continue;
		case VERB_ADJ:
			if(verb==0){
				sent[i].cc=VERB;
				verb=1;
				continue;
			}
		case NOUN_ADJ:
			if(be==1){
				sent[i].cc=ADJ;
				continue;
			}
		case PRONP_ADJ:
		case PRONS_ADJ:
			i=getnoun(i,ce);
			continue;
		case NOUN_ADV:
			if(verb == 1 && iverb == i+1){
				sent[i].cc = NOUN;
			}
			else {
				sent[i].cc = ADV;
			}
			continue;
		case ADJ_ADV:
			if(be == 1){
				if (isadj(nextc)) {
					sent[i].cc = ADV;
					continue;
				}
				if(nextc == ',' && isadv(sent[i+2].cc)) {
					sent[i++].cc = ADV;
					sent[++i].cc = ADV;
					comma--;
					continue;
				}
				sent[i].cc = ADJ;
				continue;
			}
			if(lastc == NOUN && (nextc == ',' || nextc == END)){
				sent[i].cc=ADJ;
				continue;
			}
			if(notnoun(nextc)){
				sent[i].cc=ADV;
				continue;
			}
			sent[i].cc=ADJ;
			continue;
		case WHO:
			i=who(i,ce);
			continue;
		case PRONP:
			subty=PLURAL;
			continue;
		case NOUN:
			if(nextc==NOUN){
				sent[i].cc=ADJ;
				continue;
			}
		case PRONS:
			subty=SING;
			continue;
		case PNOUN:
			sent[i].cc = NOUN;
			if(subty == 0)subty=PLURAL;
			continue;
		case CONJ:
			if(nextc==UNK || nextc == NOUN_VERB || nextc == NV_PL){
				if(lastc == ADJ)sent[++i].cc = ADJ;
				else sent[++i].cc = VERB;
			}
			if(nextc == ED){
				sent[++i].cc = VERB;
			}
			continue;
		case AUXP:
		case AUXS:
		case BES:
		case BEP:
		case AUXV:
		case AUX:
		case HAS:
		case HAVE:
		case HAD:
			verb=getv(i,ce);
			continue;
		case ADV:
		case AUXX:
		case VERB:
		case BE:
		case INTER:
			continue;
		case THAT:
			if(nextc==SUBCONJ){
				sent[i].cc=PRONP;
				goto subc;
			}
			verb=getv(i+1,ce);
			j = i+1;
			while(isadv(sent[j].cc)) {
				j++;
			}
			nextc = sent[j].cc;
			if((verb==1 && iverb== j) || nextc==ED || nextc == VERB_ADJ){
				sent[i].cc=PRONP;
				while(i+1 < j)sent[++i].cc = ADV;
			}
			else if(verb==0 && nextc==NV_PL && ((lastc==NOUN
				&& issing(sent[i-1].ic)) || lastc == ',')) {
				sent[i].cc=PRONP;
				subty=SING;
				while(i+1 < j)sent[++i].cc = ADV;
			}
			else if(verb==0 && (nextc==UNK || nextc==NOUN_VERB) && (lastc==NOUN
				&& sent[i-1].ic==NV_PL)){
				subty=PLURAL;
				sent[i].cc=PRONP;
				while(i+1 < j)sent[++i].cc = ADV;
			}
			else {
				if(i == 0)sent[i].cc=ADJ;
				else sent[i].cc=SUBCONJ;
				j=i;
				subty=0;
				i=getnoun(i+1,ce);
				if(done)sent[j].cc=ADJ;
			}
			continue;
		case ',':
			if(nextc != CONJ){	/*parenthetical*/
				continue;
			}
			i++;
			comma--;
			if(sent[i+1].cc == ED || sent[i+1].cc == MVERB){
				sent[++i].cc = VERB;
				continue;
			}
			if(command == 1){
				verb = getv(++i,ce);
				continue;
			}
		case SUBCONJ:	
subc:
			savi = i;
			if(nextc==END||(lastc==',' && nextc==',')){
				sent[i].cc=ADV;
				continue;
			}
			subty=0;
			ce=',';
			verb=getv(++i,ce);
			if(sent[savi+1].cc == VERB || sent[savi+1].cc == AUXX)
				if(sent[savi].cc == SUBCONJ)sent[savi].cc = ADV;
			i=getnoun(i,ce);
			continue;
		case PREP_ADV:
			if(sent[i+2].cc==PREP_ADV &&(sent[i+1].cc != NOUN && sent[i+1].cc != NOUN_VERB
				&& sent[i+1].cc != NV_PL && sent[i+1].cc != PNOUN)){
				sent[i].cc=ADV;
				sent[i+2].cc=CONJ;
				continue;
			}
			sav=i++;
			savsub = subty;
			pverb=getv(i,ce);
			if((pverb==1 && iverb == i) || sent[i].cc== ED){
				sent[sav].cc=PRONP;
				sent[i].cc=VERB;
				continue;
			}
			i=getnoun(i,ce);
			if(done==1){
				sent[sav].cc=ADV;
				goto sdone;
			}
			if(pverb==1 && iverb == i+1){
				sent[sav].cc=SUBCONJ;
				ce=',';
				continue;
			}
			switch(sent[i+1].cc){
			case UNK:
			case NV_PL:
			case NOUN_VERB:
			case ED:
				sent[sav].cc=SUBCONJ;
				verb=0;
				ce=',';
				continue;
			default:
				sent[sav].cc=PREP;
				subty = savsub;
				continue;
			}
		case TO:	
			savi = i;
			sent[i++].cc=VERB;
sw:
			switch(nextc){
			case UNK:
			case AUXS:
			case VERB_ADJ:
			case NOUN_VERB:	
			case VERB:
			case MVERB:
				sent[i].cc=VERB;
				continue;
			case HAVE:
				sent[i].cc = VERB;
				if(sent[i+1].cc == ED){
					sent[++i].cc = VERB;
					continue;
				}
				if(sent[i+1].cc == ADV)i++;
				if(sent[i+1].cc != BE)continue;
				i++;
			case BE:	
				sent[i].cc=VERB;
				if(sent[i+1].cc == ADV)i++;
				if(sent[i+1].cc == ED || sent[i+1].cc == ING){
					sent[++i].cc = VERB;
				}
				else if(sent[i+1].cc == UNK){
					sent[++i].cc = ADJ;
				}
				continue;
			case ADV:
				nextc = sent[++i].cc;
				goto sw;
			default:
				sent[savi].cc = PREP;
				prep=1;
				i=getnoun(i,ce);
				prep=0;
				continue;
			}
		case NV_PL: 
			if(subty==0){
				i=getnoun(i,ce);
				subty=PLURAL;
			}
			else if(verb== 0){	/*need verb*/
				sent[i].cc=VERB;
				verb=1;
			}
			else{
				i=getnoun(i,ce);
			}
			continue;
		case UNK:	
		case NOUN_VERB:
			if(verb==1){
				if(be==1 && nextc != WHO){
					sent[i].cc=ADJ;
					continue;
				}
				i=getnoun(i,ce);
			}
			else if(nextc==NV_PL && isnoun(sent[i+2].cc)) {
				sent[i].cc=NOUN;
				if(sent[i-1].cc == NOUN){
					sent[i-1].cc = ADJ;
				}
				subty=SING;
				continue;
			}
			else if(subty==0){
				i=getnoun(i,ce);
				continue;
			}
			else {		/* desparation */
				sent[i].cc=VERB;
				verb=1;
			}
			continue;
		case ED:
			if(verb == 0){
				sent[i].cc=VERB;
				verb=1;
			}
			else if(sent[i-1].cc == BE){
				sent[i].cc = VERB;
			}
			else{
				i=getnoun(i,ce);
			}
			continue;
		default:	printf("got a %c %o on %sat %d\n",sent[i].cc,sent[i].cc,sent[i].sp,i);
		}
	}
sdone:
	if(sent[i].cc==END && sent[i].ic != ';')return(-1);
	else {
		if(sent[i].ic == ';')sent[i].cc=';';
		else comma--;
		return(i);
	}
}
getv(si,ce)
int si; 
char ce;
{
	int conj;
	char c;
	int ik;
	must=0;
	bflg=0;
	hflg=0;
	verbty=0;
	for(j=si;((c=sent[j].cc) != ce) && (c != END); j++){
		iverb=j;
		switch(c){
		case ED:
			continue;
		case HAS:
			hflg=1;
			verbty=SING;
			sent[j].cc=AUXX;
			goto next;
		case HAVE:
			if(sent[j-1].cc==TO){
				continue;
			}
		case HAD:
			hflg=1;
			verbty=PLURAL;
			sent[j].cc=AUXX;
			goto next;
		case BE:
			if(sent[j].ic != BE){
				return(1);
			}
			continue;
		case VERB:	
			if(sent[j-1].cc==TO){
				continue;
			}
			return(1);
		case AUXX:
			return(1);
		case AUXP:	
			must=1;
			verbty=PLURAL;
			sent[j].cc=AUXX;
			goto next;
		case AUXS:	
			if(sent[j-1].cc==TO){
				continue;
			}
			verbty=SING;
			sent[j].cc=AUXX;
			goto next;
		case AUX:
		case AUXV:
			must=1;
			sent[j].cc=AUXX;
next:	
			if(sent[j-1].cc == CONJ && verb == 0)conj = 0;
			else conj = 1;
			if(question==1){
				j=getnoun(j+1,ce);
				question=0;
			}
			getv2(ce);
			if(sent[j].cc == AUXX){
				sent[j].cc = VERB;
			}
			if((sent[j].cc!=VERB && sent[j].cc!=BE) && sent[iverb].cc==AUXX){
				sent[iverb].cc=VERB;
				for(ik=iverb+1;ik<=j;ik++){
					if(sent[ik].ic == NOM)sent[ik].cc=NOUN;
					else sent[ik].cc=sent[ik].ic;
				}
			}
			return(conj);
		case BES:	
			verbty=SING;
			bflg=1;
			sent[j].cc=BE;
			goto next;
		case BEP:	
			verbty=PLURAL;
			bflg=1;
			sent[j].cc=BE;
			goto next;
		case SUBCONJ:
			if(sent[j-1].cc== ',' && sent[j+1].cc == ','){
				continue;
			}
		case THAT:
		case WHO:
			goto vdone;
		case ',':	if(comma==1 && sent[j+1].cc==CONJ){
					goto vdone;
				}
		}
	}
vdone:
	return(0);
}
getv2(ce)