1
0
Fork 0
mirror of https://github.com/sbrl/Pepperminty-Wiki.git synced 2024-12-22 13:45:02 +00:00

Search System; Don't bother getting a page's id if we don't need to

This commit is contained in:
Starbeamrainbowlabs 2018-06-26 14:28:11 +01:00
parent 3d3b6c491a
commit 19e49777b2
Signed by: sbrl
GPG key ID: 1BE5172E637709C2
3 changed files with 31 additions and 16 deletions

View file

@ -397,7 +397,7 @@ if($settings->sessionprefix == "auto")
/////////////////////////////////////////////////////////////////////////////
/** The version of Pepperminty Wiki currently running. */
$version = "v0.17-dev";
$commit = "67648199d7ebd8a1b2ec400af0192dc0bb94b233";
$commit = "3d3b6c491a0848922e0dc3d8c1c89a5f87673c0e";
/// Environment ///
/** Holds information about the current request environment. */
$env = new stdClass();
@ -3787,7 +3787,7 @@ register_module([
$start = microtime(true);
foreach($results as &$result) {
$result["context"] = search::extract_context(
$invindex, ids::getid($result["pagename"]),
$invindex, $result["pagename"],
$_GET["query"],
file_get_contents($env->storage_prefix . $result["pagename"] . ".md")
);
@ -4454,13 +4454,14 @@ class search
reset($pageindex); // Reset array/object pointer
foreach ($pageindex as $pagename => $pagedata)
{
// Get the current page's id
$pageid = ids::getid($pagename);
// Seteup a variable to hold the current page's id
$pageid = false; // Only fill this out if we find a match
// Consider matches in the page title
$title_matches = mb_stripos_all($literator->transliterate($pagename), $qterm);
$title_matches_count = $title_matches !== false ? count($title_matches) : 0;
if($title_matches_count > 0)
{
$pageid = ids::getid($pagename); // Fill out the page id
// We found the qterm in the title
if(!isset($matching_pages[$pageid]))
$matching_pages[$pageid] = [ "nterms" => [] ];
@ -4478,6 +4479,9 @@ class search
if($tag_matches_count > 0) // And we found the qterm in the tags
{
if($pageid == false) // Fill out the page id if it hasn't been already
$pageid = ids::getid($pagename);
if(!isset($matching_pages[$pageid]))
$matching_pages[$pageid] = [ "nterms" => [] ];
@ -4554,14 +4558,17 @@ class search
/**
* Extracts a context string (in HTML) given a search query that could be displayed
* in a list of search results.
* @param string $query The search queary to generate the context for.
* @param string $source The page source to extract the context from.
* @return string The generated context string.
* @param string $invindex The inverted index to consult.
* @param string $pagename The name of the paget that this source belongs to. Used when consulting the inverted index.
* @param string $query The search queary to generate the context for.
* @param string $source The page source to extract the context from.
* @return string The generated context string.
*/
public static function extract_context($invindex, $pageid, $query, $source)
public static function extract_context($invindex, $pagename, $query, $source)
{
global $settings;
$pageid = ids::getid($pagename);
$nterms = self::tokenize($query);
$matches = [];

View file

@ -104,7 +104,7 @@
"author": "Starbeamrainbowlabs",
"description": "Adds proper search functionality to Pepperminty Wiki using an inverted index to provide a full text search engine. If pages don't show up, then you might have hit a stop word. If not, try requesting the `invindex-rebuild` action to rebuild the inverted index from scratch.",
"id": "feature-search",
"lastupdate": 1530018727,
"lastupdate": 1530019335,
"optional": false
},
{

View file

@ -141,7 +141,7 @@ register_module([
$start = microtime(true);
foreach($results as &$result) {
$result["context"] = search::extract_context(
$invindex, ids::getid($result["pagename"]),
$invindex, $result["pagename"],
$_GET["query"],
file_get_contents($env->storage_prefix . $result["pagename"] . ".md")
);
@ -808,13 +808,15 @@ class search
reset($pageindex); // Reset array/object pointer
foreach ($pageindex as $pagename => $pagedata)
{
// Get the current page's id
$pageid = ids::getid($pagename);
// Seteup a variable to hold the current page's id
$pageid = false; // Only fill this out if we find a match
// Consider matches in the page title
// FUTURE: We may be able to optimise this further by using preg_match_all + preg_quote instead of mb_stripos_all. Experimentation / benchmarking is required to figure out which one is faster
$title_matches = mb_stripos_all($literator->transliterate($pagename), $qterm);
$title_matches_count = $title_matches !== false ? count($title_matches) : 0;
if($title_matches_count > 0)
{
$pageid = ids::getid($pagename); // Fill out the page id
// We found the qterm in the title
if(!isset($matching_pages[$pageid]))
$matching_pages[$pageid] = [ "nterms" => [] ];
@ -832,6 +834,9 @@ class search
if($tag_matches_count > 0) // And we found the qterm in the tags
{
if($pageid == false) // Fill out the page id if it hasn't been already
$pageid = ids::getid($pagename);
if(!isset($matching_pages[$pageid]))
$matching_pages[$pageid] = [ "nterms" => [] ];
@ -908,14 +913,17 @@ class search
/**
* Extracts a context string (in HTML) given a search query that could be displayed
* in a list of search results.
* @param string $query The search queary to generate the context for.
* @param string $source The page source to extract the context from.
* @return string The generated context string.
* @param string $invindex The inverted index to consult.
* @param string $pagename The name of the paget that this source belongs to. Used when consulting the inverted index.
* @param string $query The search queary to generate the context for.
* @param string $source The page source to extract the context from.
* @return string The generated context string.
*/
public static function extract_context($invindex, $pageid, $query, $source)
public static function extract_context($invindex, $pagename, $query, $source)
{
global $settings;
$pageid = ids::getid($pagename);
$nterms = self::tokenize($query);
$matches = [];