/*
* r_reg -- display a regular file in reverse order by line.
*/
static int
r_reg(FILE *fp, enum STYLE style, off_t off, struct stat *sbp)
{
off_t start, pos, end;
int ch;
end = sbp->st_size;
if (end == 0)
return (0);
/* Position before char, ignore last char whether newline or not */
pos = end-2;
ch = EOF;
start = 0;
if (style == RBYTES && off < end)
start = end - off;
for (; pos >= start; pos--) {
/* A seek per char isn't a problem with a smart stdio */
if (fseeko(fp, pos, SEEK_SET) != 0) {
ierr();
return (0);
}
if ((ch = getc(fp)) == '\n') {
while (--end > pos)
COPYCHAR(fp, ch);
end++;
if (style == RLINES && --off == 0)
break;
}
else if (ch == EOF) {
ierr();
return (0);
}
}
if (pos < start) {
if (ch != EOF && ungetc(ch, fp) == EOF) {
ierr();
return (0);
}
while (--end >= start)
COPYCHAR(fp, ch);
}
return (0);
}
void
RS_compile(Regis *r, bool issuffix, const char *str)
{
int len = strlen(str);
int state = RS_IN_WAIT;
const char *c = str;
RegisNode *ptr = NULL;
memset(r, 0, sizeof(Regis));
r->issuffix = (issuffix) ? 1 : 0;
while (*c)
{
if (state == RS_IN_WAIT)
{
if (t_isalpha(c))
{
if (ptr)
ptr = newRegisNode(ptr, len);
else
ptr = r->node = newRegisNode(NULL, len);
COPYCHAR(ptr->data, c);
ptr->type = RSF_ONEOF;
ptr->len = pg_mblen(c);
}
else if (t_iseq(c, '['))
{
if (ptr)
ptr = newRegisNode(ptr, len);
else
ptr = r->node = newRegisNode(NULL, len);
ptr->type = RSF_ONEOF;
state = RS_IN_ONEOF;
}
else /* shouldn't get here */
elog(ERROR, "invalid regis pattern: \"%s\"", str);
}
else if (state == RS_IN_ONEOF)
{
if (t_iseq(c, '^'))
{
ptr->type = RSF_NONEOF;
state = RS_IN_NONEOF;
}
else if (t_isalpha(c))
{
COPYCHAR(ptr->data, c);
ptr->len = pg_mblen(c);
state = RS_IN_ONEOF_IN;
}
else /* shouldn't get here */
elog(ERROR, "invalid regis pattern: \"%s\"", str);
}
else if (state == RS_IN_ONEOF_IN || state == RS_IN_NONEOF)
{
if (t_isalpha(c))
{
COPYCHAR(ptr->data + ptr->len, c);
ptr->len += pg_mblen(c);
}
else if (t_iseq(c, ']'))
state = RS_IN_WAIT;
else /* shouldn't get here */
elog(ERROR, "invalid regis pattern: \"%s\"", str);
}
else
elog(ERROR, "internal error in RS_compile: state %d", state);
c += pg_mblen(c);
}
if (state != RS_IN_WAIT) /* shouldn't get here */
elog(ERROR, "invalid regis pattern: \"%s\"", str);
ptr = r->node;
while (ptr)
{
r->nchar++;
ptr = ptr->next;
}
}
/*
* Get next token from string being parsed. Returns true if successful,
* false if end of input string is reached. On success, these output
* parameters are filled in:
*
* *strval pointer to token
* *lenval length of *strval
* *pos_ptr pointer to a palloc'd array of positions and weights
* associated with the token. If the caller is not interested
* in the information, NULL can be supplied. Otherwise
* the caller is responsible for pfreeing the array.
* *poslen number of elements in *pos_ptr
* *endptr scan resumption point
*
* Pass NULL for unwanted output parameters.
*/
bool
gettoken_tsvector(TSVectorParseState state,
char **strval, int *lenval,
WordEntryPos **pos_ptr, int *poslen,
char **endptr)
{
int oldstate = 0;
char *curpos = state->word;
int statecode = WAITWORD;
/*
* pos is for collecting the comma delimited list of positions followed by
* the actual token.
*/
WordEntryPos *pos = NULL;
int npos = 0; /* elements of pos used */
int posalen = 0; /* allocated size of pos */
while (1)
{
if (statecode == WAITWORD)
{
if (*(state->prsbuf) == '\0')
return false;
else if (t_iseq(state->prsbuf, '\''))
statecode = WAITENDCMPLX;
else if (t_iseq(state->prsbuf, '\\'))
{
statecode = WAITNEXTCHAR;
oldstate = WAITENDWORD;
}
else if (state->oprisdelim && ISOPERATOR(state->prsbuf))
PRSSYNTAXERROR;
else if (!t_isspace(state->prsbuf))
{
COPYCHAR(curpos, state->prsbuf);
curpos += pg_mblen(state->prsbuf);
statecode = WAITENDWORD;
}
}
else if (statecode == WAITNEXTCHAR)
{
if (*(state->prsbuf) == '\0')
ereport(ERROR,
(errcode(ERRCODE_SYNTAX_ERROR),
errmsg("there is no escaped character: \"%s\"",
state->bufstart)));
else
{
RESIZEPRSBUF;
COPYCHAR(curpos, state->prsbuf);
curpos += pg_mblen(state->prsbuf);
Assert(oldstate != 0);
statecode = oldstate;
}
}
else if (statecode == WAITENDWORD)
{
if (t_iseq(state->prsbuf, '\\'))
{
statecode = WAITNEXTCHAR;
oldstate = WAITENDWORD;
}
else if (t_isspace(state->prsbuf) || *(state->prsbuf) == '\0' ||
(state->oprisdelim && ISOPERATOR(state->prsbuf)))
{
RESIZEPRSBUF;
if (curpos == state->word)
PRSSYNTAXERROR;
*(curpos) = '\0';
RETURN_TOKEN;
}
else if (t_iseq(state->prsbuf, ':'))
{
if (curpos == state->word)
PRSSYNTAXERROR;
*(curpos) = '\0';
if (state->oprisdelim)
RETURN_TOKEN;
else
statecode = INPOSINFO;
}
else
{
RESIZEPRSBUF;
COPYCHAR(curpos, state->prsbuf);
curpos += pg_mblen(state->prsbuf);
}
}
else if (statecode == WAITENDCMPLX)
{
if (t_iseq(state->prsbuf, '\''))
{
statecode = WAITCHARCMPLX;
}
else if (t_iseq(state->prsbuf, '\\'))
{
statecode = WAITNEXTCHAR;
oldstate = WAITENDCMPLX;
}
else if (*(state->prsbuf) == '\0')
PRSSYNTAXERROR;
else
{
RESIZEPRSBUF;
COPYCHAR(curpos, state->prsbuf);
curpos += pg_mblen(state->prsbuf);
}
}
else if (statecode == WAITCHARCMPLX)
{
if (t_iseq(state->prsbuf, '\''))
{
RESIZEPRSBUF;
COPYCHAR(curpos, state->prsbuf);
curpos += pg_mblen(state->prsbuf);
statecode = WAITENDCMPLX;
}
else
{
RESIZEPRSBUF;
*(curpos) = '\0';
if (curpos == state->word)
PRSSYNTAXERROR;
if (state->oprisdelim)
{
/* state->prsbuf+=pg_mblen(state->prsbuf); */
RETURN_TOKEN;
}
else
statecode = WAITPOSINFO;
continue; /* recheck current character */
}
}
else if (statecode == WAITPOSINFO)
{
if (t_iseq(state->prsbuf, ':'))
statecode = INPOSINFO;
else
RETURN_TOKEN;
}
else if (statecode == INPOSINFO)
{
if (t_isdigit(state->prsbuf))
{
if (posalen == 0)
{
posalen = 4;
pos = (WordEntryPos *) palloc(sizeof(WordEntryPos) * posalen);
npos = 0;
}
else if (npos + 1 >= posalen)
{
posalen *= 2;
pos = (WordEntryPos *) repalloc(pos, sizeof(WordEntryPos) * posalen);
}
npos++;
WEP_SETPOS(pos[npos - 1], LIMITPOS(atoi(state->prsbuf)));
/* we cannot get here in tsquery, so no need for 2 errmsgs */
if (WEP_GETPOS(pos[npos - 1]) == 0)
ereport(ERROR,
(errcode(ERRCODE_SYNTAX_ERROR),
errmsg("wrong position info in tsvector: \"%s\"",
state->bufstart)));
WEP_SETWEIGHT(pos[npos - 1], 0);
statecode = WAITPOSDELIM;
}
else
PRSSYNTAXERROR;
}
else if (statecode == WAITPOSDELIM)
{
if (t_iseq(state->prsbuf, ','))
statecode = INPOSINFO;
else if (t_iseq(state->prsbuf, 'a') || t_iseq(state->prsbuf, 'A') || t_iseq(state->prsbuf, '*'))
{
if (WEP_GETWEIGHT(pos[npos - 1]))
PRSSYNTAXERROR;
WEP_SETWEIGHT(pos[npos - 1], 3);
}
else if (t_iseq(state->prsbuf, 'b') || t_iseq(state->prsbuf, 'B'))
{
if (WEP_GETWEIGHT(pos[npos - 1]))
PRSSYNTAXERROR;
WEP_SETWEIGHT(pos[npos - 1], 2);
}
else if (t_iseq(state->prsbuf, 'c') || t_iseq(state->prsbuf, 'C'))
{
if (WEP_GETWEIGHT(pos[npos - 1]))
PRSSYNTAXERROR;
WEP_SETWEIGHT(pos[npos - 1], 1);
}
else if (t_iseq(state->prsbuf, 'd') || t_iseq(state->prsbuf, 'D'))
{
if (WEP_GETWEIGHT(pos[npos - 1]))
PRSSYNTAXERROR;
WEP_SETWEIGHT(pos[npos - 1], 0);
}
else if (t_isspace(state->prsbuf) ||
*(state->prsbuf) == '\0')
RETURN_TOKEN;
else if (!t_isdigit(state->prsbuf))
PRSSYNTAXERROR;
}
else /* internal error */
elog(ERROR, "unrecognized state in gettoken_tsvector: %d",
statecode);
/* get next char */
state->prsbuf += pg_mblen(state->prsbuf);
}
}
