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

search/index: fixed a number of bad conversions to properly deal with text documents bigger than 2GB

Browse Source
This commit is contained in:
"Jean-Francois Dockes ext:(%22)
2011-07-12 08:28:09 -07:00
parent 5e59354535
commit 88685d2e64
11 changed files with 42 additions and 23 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/index/beaglequeue.cpp
+2 -2
View File
@@ -406,7 +406,7 @@ BeagleQueueIndexer::processone(const string &path,
dotdoc.fmtime = ascdate;
char cbuf[100];
sprintf(cbuf, "%ld", (long)stp->st_size);
sprintf(cbuf, OFFTPC, stp->st_size);
dotdoc.fbytes = cbuf;
// Document signature for up to date checks: none.
@@ -453,7 +453,7 @@ BeagleQueueIndexer::processone(const string &path,
doc.fmtime = ascdate;
char cbuf[100];
sprintf(cbuf, "%ld", (long)stp->st_size);
sprintf(cbuf, OFFTPC, stp->st_size);
doc.fbytes = cbuf;
// Document signature for up to date checks: none.
doc.sig = "";
src/index/fsindexer.cpp
+13 -9
View File
@@ -288,6 +288,12 @@ void FsIndexer::setlocalfields(Rcl::Doc& doc)
}
}
static void makesig(const struct stat *stp, string& out)
{
char cbuf[100];
sprintf(cbuf, OFFTPC "%ld", stp->st_size, (long)stp->RCL_STTIME);
out = cbuf;
}
/// This method gets called for every file and directory found by the
/// tree walker.
@@ -340,9 +346,8 @@ FsIndexer::processone(const std::string &fn, const struct stat *stp,
// for the uptodate check (the value computed here is checked
// against the stored one). Changing the computation forces a full
// reindex of course.
char cbuf[100];
sprintf(cbuf, "%ld%ld", (long)stp->st_size, (long)stp->RCL_STTIME);
string sig = cbuf;
string sig;
makesig(stp, sig);
string udi;
make_udi(fn, "", udi);
if (!m_db->needUpdate(udi, sig)) {
@@ -420,14 +425,14 @@ FsIndexer::processone(const std::string &fn, const struct stat *stp,
doc.utf8fn = utf8fn;
char cbuf[100];
sprintf(cbuf, "%ld", (long)stp->st_size);
sprintf(cbuf, OFFTPC, stp->st_size);
doc.fbytes = cbuf;
// Document signature for up to date checks: concatenate
// m/ctime and size. Looking for changes only, no need to
// parseback so no need for reversible formatting. Also set,
// but never used, for subdocs.
sprintf(cbuf, "%ld%ld", (long)stp->st_size, (long)stp->RCL_STTIME);
doc.sig = cbuf;
makesig(stp, doc.sig);
// If there was an error, ensure indexing will be
// retried. This is for the once missing, later installed
// filter case. It can make indexing much slower (if there are
@@ -473,11 +478,10 @@ FsIndexer::processone(const std::string &fn, const struct stat *stp,
fileDoc.url = string("file://") + fn;
char cbuf[100];
sprintf(cbuf, "%ld", (long)stp->st_size);
sprintf(cbuf, OFFTPC, stp->st_size);
fileDoc.fbytes = cbuf;
// Document signature for up to date checks.
sprintf(cbuf, "%ld%ld", (long)stp->st_size, (long)stp->RCL_STTIME);
fileDoc.sig = cbuf;
makesig(stp, fileDoc.sig);
if (!m_db->addOrUpdate(parent_udi, "", fileDoc))
return FsTreeWalker::FtwError;
}