tmpstr = POPs;
if (SvROK(tmpstr)) {
- SV *sv = SvRV(tmpstr);
+ SV * const sv = SvRV(tmpstr);
if(SvMAGICAL(sv))
mg = mg_find(sv, PERL_MAGIC_qr);
}
else {
STRLEN len;
const char *t = SvPV_const(tmpstr, len);
+ regexp * const re = PM_GETRE(pm);
/* Check against the last compiled regexp. */
- if (!PM_GETRE(pm) || !PM_GETRE(pm)->precomp ||
- PM_GETRE(pm)->prelen != (I32)len ||
- memNE(PM_GETRE(pm)->precomp, t, len))
+ if (!re || !re->precomp || re->prelen != (I32)len ||
+ memNE(re->precomp, t, len))
{
- if (PM_GETRE(pm)) {
- ReREFCNT_dec(PM_GETRE(pm));
+ if (re) {
+ ReREFCNT_dec(re);
PM_SETRE(pm, NULL); /* crucial if regcomp aborts */
}
if (PL_op->op_flags & OPf_SPECIAL)
/* set up a scratch pad */
CvPADLIST(PL_compcv) = pad_new(padnew_SAVE);
+ PL_op = NULL; /* avoid PL_op and PL_curpad referring to different CVs */
if (!PL_madskills)
const I32 gimme = GIMME_V;
int filter_has_file = 0;
PerlIO *tryrsfp = NULL;
+ SV *filter_cache = NULL;
SV *filter_state = NULL;
SV *filter_sub = NULL;
SV *hook_sv = NULL;
SP -= count - 1;
arg = SP[i++];
+ if (SvROK(arg) && (SvTYPE(SvRV(arg)) <= SVt_PVLV)
+ && !isGV_with_GP(SvRV(arg))) {
+ filter_cache = SvRV(arg);
+ SvREFCNT_inc_void_NN(filter_cache);
+
+ if (i < count) {
+ arg = SP[i++];
+ }
+ }
+
if (SvROK(arg) && SvTYPE(SvRV(arg)) == SVt_PVGV) {
arg = SvRV(arg);
}
filter_state = SP[i];
SvREFCNT_inc_simple_void(filter_state);
}
+ }
- if (!tryrsfp) {
- tryrsfp = PerlIO_open(BIT_BUCKET,
- PERL_SCRIPT_MODE);
- }
+ if (!tryrsfp && (filter_cache || filter_sub)) {
+ tryrsfp = PerlIO_open(BIT_BUCKET,
+ PERL_SCRIPT_MODE);
}
SP--;
}
}
filter_has_file = 0;
+ if (filter_cache) {
+ SvREFCNT_dec(filter_cache);
+ filter_cache = NULL;
+ }
if (filter_state) {
SvREFCNT_dec(filter_state);
filter_state = NULL;
SAVESPTR(PL_compiling.cop_io);
PL_compiling.cop_io = NULL;
- if (filter_sub) {
+ if (filter_sub || filter_cache) {
SV * const datasv = filter_add(S_run_user_filter, NULL);
IoLINES(datasv) = filter_has_file;
IoTOP_GV(datasv) = (GV *)filter_state;
IoBOTTOM_GV(datasv) = (GV *)filter_sub;
+ IoFMT_GV(datasv) = (GV *)filter_cache;
}
/* switch to eval mode */
const int filter_has_file = IoLINES(datasv);
SV * const filter_state = (SV *)IoTOP_GV(datasv);
SV * const filter_sub = (SV *)IoBOTTOM_GV(datasv);
- int len = 0;
- /* Filter API says that the filter appends to the contents of the buffer.
- Usually the buffer is "", so the details don't matter. But if it's not,
- then clearly what it contains is already filtered by this filter, so we
- don't want to pass it in a second time.
- I'm going to use a mortal in case the upstream filter croaks. */
- SV *const upstream
- = ((SvOK(buf_sv) && sv_len(buf_sv)) || SvGMAGICAL(buf_sv))
- ? sv_newmortal() : buf_sv;
+ int status = 0;
+ SV *upstream;
+ STRLEN got_len;
+ const char *got_p;
+ const char *prune_from = NULL;
+ bool read_from_cache = FALSE;
- SvUPGRADE(upstream, SVt_PV);
/* I was having segfault trouble under Linux 2.2.5 after a
parse error occured. (Had to hack around it with a test
for PL_error_count == 0.) Solaris doesn't segfault --
not sure where the trouble is yet. XXX */
- if (maxlen && IoFMT_GV(datasv)) {
+ if (IoFMT_GV(datasv)) {
SV *const cache = (SV *)IoFMT_GV(datasv);
if (SvOK(cache)) {
STRLEN cache_len;
const char *cache_p = SvPV(cache, cache_len);
- /* Running in block mode and we have some cached data already. */
- if (cache_len >= maxlen) {
- /* In fact, so much data we don't even need to call
- filter_read. */
- sv_catpvn(buf_sv, cache_p, maxlen);
- sv_chop(cache, cache_p + maxlen);
+ STRLEN take = 0;
+
+ if (maxlen) {
+ /* Running in block mode and we have some cached data already.
+ */
+ if (cache_len >= maxlen) {
+ /* In fact, so much data we don't even need to call
+ filter_read. */
+ take = maxlen;
+ }
+ } else {
+ const char *const first_nl = memchr(cache_p, '\n', cache_len);
+ if (first_nl) {
+ take = first_nl + 1 - cache_p;
+ }
+ }
+ if (take) {
+ sv_catpvn(buf_sv, cache_p, take);
+ sv_chop(cache, cache_p + take);
/* Definately not EOF */
return 1;
}
+
sv_catsv(buf_sv, cache);
- maxlen -= cache_len;
+ if (maxlen) {
+ maxlen -= cache_len;
+ }
SvOK_off(cache);
+ read_from_cache = TRUE;
}
}
+
+ /* Filter API says that the filter appends to the contents of the buffer.
+ Usually the buffer is "", so the details don't matter. But if it's not,
+ then clearly what it contains is already filtered by this filter, so we
+ don't want to pass it in a second time.
+ I'm going to use a mortal in case the upstream filter croaks. */
+ upstream = ((SvOK(buf_sv) && sv_len(buf_sv)) || SvGMAGICAL(buf_sv))
+ ? sv_newmortal() : buf_sv;
+ SvUPGRADE(upstream, SVt_PV);
if (filter_has_file) {
- len = FILTER_READ(idx+1, upstream, maxlen);
+ status = FILTER_READ(idx+1, upstream, 0);
}
- if (filter_sub && len >= 0) {
+ if (filter_sub && status >= 0) {
dSP;
int count;
DEFSV = upstream;
PUSHMARK(SP);
- PUSHs(sv_2mortal(newSViv(maxlen)));
+ PUSHs(sv_2mortal(newSViv(0)));
if (filter_state) {
PUSHs(filter_state);
}
if (count > 0) {
SV *out = POPs;
if (SvOK(out)) {
- len = SvIV(out);
+ status = SvIV(out);
}
}
LEAVE;
}
- if (maxlen) {
- /* Running in block mode. */
- STRLEN got_len;
- const char *got_p = SvPV(upstream, got_len);
-
- if (got_len > maxlen) {
- /* Oh. Too long. Stuff some in our cache. */
- SV *cache = (SV *)IoFMT_GV(datasv);
-
- if (!cache) {
- IoFMT_GV(datasv) = (GV*) (cache = newSV(got_len - maxlen));
- } else if (SvOK(cache)) {
- /* Cache should be empty. */
- assert(!SvCUR(cache));
+ if(SvOK(upstream)) {
+ got_p = SvPV(upstream, got_len);
+ if (maxlen) {
+ if (got_len > maxlen) {
+ prune_from = got_p + maxlen;
}
-
- sv_setpvn(cache, got_p + maxlen, got_len - maxlen);
- /* If you ask for block mode, you may well split UTF-8 characters.
- "If it breaks, you get to keep both parts"
- (Your code is broken if you don't put them back together again
- before something notices.) */
- if (SvUTF8(upstream)) {
- SvUTF8_on(cache);
+ } else {
+ const char *const first_nl = memchr(got_p, '\n', got_len);
+ if (first_nl && first_nl + 1 < got_p + got_len) {
+ /* There's a second line here... */
+ prune_from = first_nl + 1;
}
- SvCUR_set(upstream, maxlen);
}
}
+ if (prune_from) {
+ /* Oh. Too long. Stuff some in our cache. */
+ STRLEN cached_len = got_p + got_len - prune_from;
+ SV *cache = (SV *)IoFMT_GV(datasv);
+
+ if (!cache) {
+ IoFMT_GV(datasv) = (GV*) (cache = newSV(got_len - maxlen));
+ } else if (SvOK(cache)) {
+ /* Cache should be empty. */
+ assert(!SvCUR(cache));
+ }
+
+ sv_setpvn(cache, prune_from, cached_len);
+ /* If you ask for block mode, you may well split UTF-8 characters.
+ "If it breaks, you get to keep both parts"
+ (Your code is broken if you don't put them back together again
+ before something notices.) */
+ if (SvUTF8(upstream)) {
+ SvUTF8_on(cache);
+ }
+ SvCUR_set(upstream, got_len - cached_len);
+ /* Can't yet be EOF */
+ if (status == 0)
+ status = 1;
+ }
- if (upstream != buf_sv) {
+ /* If they are at EOF but buf_sv has something in it, then they may never
+ have touched the SV upstream, so it may be undefined. If we naively
+ concatenate it then we get a warning about use of uninitialised value.
+ */
+ if (upstream != buf_sv && (SvOK(upstream) || SvGMAGICAL(upstream))) {
sv_catsv(buf_sv, upstream);
}
- if (len <= 0) {
+ if (status <= 0) {
IoLINES(datasv) = 0;
SvREFCNT_dec(IoFMT_GV(datasv));
if (filter_state) {
}
filter_del(S_run_user_filter);
}
- return len;
+ if (status == 0 && read_from_cache) {
+ /* If we read some data from the cache (and by getting here it implies
+ that we emptied the cache) then we aren't yet at EOF, and mustn't
+ report that to our caller. */
+ return 1;
+ }
+ return status;
}
/* perhaps someone can come up with a better name for
projects / p5sagit/p5-mst-13.2.git / blobdiff