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

cleanup repeated punctuation in snippets

Browse Source
This commit is contained in:
Jean-Francois Dockes 2018-04-10 13:07:27 +02:00
parent 8dee9054fe
commit e4e5ee35d6
1 changed files with 14 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
src/rcldb/rclabsfromtext.cpp
View File

@ -22,6 +22,7 @@
#include <unordered_map> #include <unordered_map>
#include <deque> #include <deque>
#include <algorithm> #include <algorithm>
#include <regex>
#include "log.h" #include "log.h"
#include "rcldb.h" #include "rcldb.h"
@ -45,8 +46,19 @@ using namespace std;
namespace Rcl { namespace Rcl {
//// Fragment cleanup
// Chars we turn to spaces in the Snippets // Chars we turn to spaces in the Snippets
static const string cstr_nc("\n\r\x0c\\"); static const string cstr_nc("\n\r\x0c\\");
// Things that we don't want to repeat in a displayed snippet.
// e.g. > > > > > >
static const string punctcls("[<>.-_+,#*=]");
static const string punctRE = "(" + punctcls + " *)(" + punctcls + " *)+";
static std::regex fixfrag_re(punctRE);
static string fixfrag(const string& infrag)
{
return std::regex_replace(neutchars(infrag, cstr_nc), fixfrag_re, "$2");
}
// Fragment descriptor. A fragment is a text area with one or several // Fragment descriptor. A fragment is a text area with one or several
// matched terms and some context. It is ranked according to the // matched terms and some context. It is ranked according to the
@ -383,8 +395,8 @@ int Query::Native::abstractFromText(
// main term and the page positions. // main term and the page positions.
unsigned int count = 0; unsigned int count = 0;
for (const auto& entry : result) { for (const auto& entry : result) {
string frag = neutchars( string frag(
rawtext.substr(entry.start, entry.stop - entry.start), cstr_nc); fixfrag(rawtext.substr(entry.start, entry.stop - entry.start)));
#ifdef COMPUTE_HLZONES #ifdef COMPUTE_HLZONES
// This would need to be modified to take tag parameters // This would need to be modified to take tag parameters