tris/recoll
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

mh_mail: use truncate_to_word to avoid cutting an utf8 char. rcldb: logdeb text_to_word errors

Browse Source
This commit is contained in:
dockes 2009-11-18 10:26:47 +00:00
parent f5637445d6
commit dd6acb07cc
2 changed files with 9 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
src/internfile/mh_mail.cpp
View File

@ -169,7 +169,8 @@ bool MimeHandlerMail::next_document()
m_metaData[cstr_mimetype].c_str())); m_metaData[cstr_mimetype].c_str()));
const string& txt = m_metaData[cstr_content]; const string& txt = m_metaData[cstr_content];
if (m_startoftext < txt.size()) if (m_startoftext < txt.size())
m_metaData[cstr_abstract] = txt.substr(m_startoftext, 250); m_metaData[cstr_abstract] =
truncate_to_word(txt.substr(m_startoftext), 250);
} else { } else {
m_metaData[cstr_abstract] = ""; m_metaData[cstr_abstract] = "";
res = processAttach(); res = processAttach();

10
src/rcldb/rcldb.cpp
View File

@ -879,7 +879,8 @@ bool Db::addOrUpdate(const string &udi, const string &parent_udi,
// Split and index file name as document term(s) // Split and index file name as document term(s)
LOGDEB2(("Db::add: split file name [%s]\n", fn.c_str())); LOGDEB2(("Db::add: split file name [%s]\n", fn.c_str()));
splitter.text_to_words(doc.utf8fn); if (!splitter.text_to_words(doc.utf8fn))
LOGDEB(("Db::addOrUpdate: split failed for file name\n"));
splitData.basepos += splitData.curpos + 100; splitData.basepos += splitData.curpos + 100;
// Index textual metadata. These are all indexed as text with // Index textual metadata. These are all indexed as text with
@ -901,7 +902,9 @@ bool Db::addOrUpdate(const string &udi, const string &parent_udi,
meta_it->first.c_str(), pfx.c_str(), meta_it->first.c_str(), pfx.c_str(),
meta_it->second.c_str())); meta_it->second.c_str()));
splitData.setprefix(pfx); // Subject splitData.setprefix(pfx); // Subject
splitter.text_to_words(meta_it->second); if (!splitter.text_to_words(meta_it->second))
LOGDEB(("Db::addOrUpdate: split failed for %s\n",
meta_it->first.c_str()));
splitData.setprefix(string()); splitData.setprefix(string());
splitData.basepos += splitData.curpos + 100; splitData.basepos += splitData.curpos + 100;
} }
@ -914,7 +917,8 @@ bool Db::addOrUpdate(const string &udi, const string &parent_udi,
// Split and index body text // Split and index body text
LOGDEB2(("Db::add: split body\n")); LOGDEB2(("Db::add: split body\n"));
splitter.text_to_words(doc.text); if (!splitter.text_to_words(doc.text))
LOGDEB(("Db::addOrUpdate: split failed for main text\n"));
////// Special terms for other metadata. No positions for these. ////// Special terms for other metadata. No positions for these.
// Mime type // Mime type