yetangitu/recoll
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

Use readnext() method to read even 1st chunk of text files to perform appropriate end of chunk truncation to eol. Wont affect uncunked files

Browse source
This commit is contained in:
Jean-Francois Dockes 2015-01-21 16:03:26 +01:00
parent aa36cda20c
commit 5cbae3985b
1 changed files with 5 additions and 10 deletions
Download patch file Download diff file Expand all files Collapse all files

13
src/internfile/mh_text.cpp
View file

@ -71,16 +71,10 @@ bool MimeHandlerText::set_document_file(const string& mt, const string &fn)
ps *= KB; ps *= KB;
m_paging = true; m_paging = true;
} }
// Note: size_t is guaranteed unsigned, so max if ps is -1
m_pagesz = size_t(ps); m_pagesz = size_t(ps);
string reason; if (!readnext())
LOGDEB(("calling file_to_string\n"));
// file_to_string() takes pagesz == size_t(-1) to mean read all.
if (!file_to_string(fn, m_text, 0, m_pagesz, &reason)) {
LOGERR(("MimeHandlerText: can't read file: %s\n", reason.c_str()));
return false; return false;
}
LOGDEB(("file_to_string OK\n"));
m_offs = m_text.length();
} }
if (!m_forPreview) { if (!m_forPreview) {
string md5, xmd5; string md5, xmd5;
@ -178,7 +172,8 @@ bool MimeHandlerText::readnext()
} }
// If possible try to adjust the chunk to end right after a line // If possible try to adjust the chunk to end right after a line
// Don't do this for the last chunk // Don't do this for the last chunk. Last chunk of exactly the
// page size might be unduly split, no big deal
if (m_text.length() == m_pagesz) { if (m_text.length() == m_pagesz) {
string::size_type pos = m_text.find_last_of("\n\r"); string::size_type pos = m_text.find_last_of("\n\r");
if (pos != string::npos && pos != 0) { if (pos != string::npos && pos != 0) {