summaryrefslogtreecommitdiffabout
path: root/lib/discovery.cc
authorMichael Krelin <hacker@klever.net>2008-04-06 19:26:50 (UTC)
committer Michael Krelin <hacker@klever.net>2008-04-06 19:26:50 (UTC)
commitaa3a6821f6aaaf4ae63bafbbc76da5f414a75fd4 (patch) (side-by-side diff)
treec64231565e123d678850a7b0560b348ee82b2dcc /lib/discovery.cc
parent752e484cd2fc239bc582a88fe7d62a225880ee3b (diff)
downloadlibopkele-aa3a6821f6aaaf4ae63bafbbc76da5f414a75fd4.zip
libopkele-aa3a6821f6aaaf4ae63bafbbc76da5f414a75fd4.tar.gz
libopkele-aa3a6821f6aaaf4ae63bafbbc76da5f414a75fd4.tar.bz2
fix for discovery failing on some entities coming before the tag of interest
One of the numerous opportunities to express my gratitude by means of commit message to Joseph Smarr of Plaxo for spotting it. Thanks, Joseph! Signed-off-by: Michael Krelin <hacker@klever.net>
Diffstat (limited to 'lib/discovery.cc') (more/less context) (ignore whitespace changes)
-rw-r--r--lib/discovery.cc1
1 files changed, 1 insertions, 0 deletions
diff --git a/lib/discovery.cc b/lib/discovery.cc
index c118c80..3b90977 100644
--- a/lib/discovery.cc
+++ b/lib/discovery.cc
@@ -247,32 +247,33 @@ namespace opkele {
if(!parser_choked) {
parse(0,0,true);
}else if(xmode&xmode_html){
/* TODO: do not bother if we've seen xml */
try {
util::tidy_doc_t td = util::tidy_doc_t::create();
if(!td)
throw exception_tidy(OPKELE_CP_ "failed to create htmltidy document");
#ifndef NDEBUG
td.opt_set(TidyQuiet,false);
td.opt_set(TidyShowWarnings,false);
#endif /* NDEBUG */
td.opt_set(TidyForceOutput,true);
td.opt_set(TidyXhtmlOut,true);
td.opt_set(TidyDoctypeMode,TidyDoctypeOmit);
td.opt_set(TidyMark,false);
+ td.opt_set(TidyNumEntities,true);
if(td.parse_string(save_html)<=0)
throw exception_tidy(OPKELE_CP_ "tidy failed to parse document");
if(td.clean_and_repair()<=0)
throw exception_tidy(OPKELE_CP_ "tidy failed to clean and repair");
util::tidy_buf_t tide;
if(td.save_buffer(tide)<=0)
throw exception_tidy(OPKELE_CP_ "tidy failed to save buffer");
prepare_to_parse();
parse(tide.c_str(),tide.size(),true);
}catch(exception_tidy& et) { }
}
save_html.clear();
}
void prepare_to_parse() {
(*(expat_t*)this) = parser_create_ns();