Friendica/friendica-github
1
0
Fork 0
mirror of https://github.com/friendica/friendica synced 2025-04-25 15:50:10 +00:00
Code Issues Projects Releases Packages Wiki Activity

Merge remote-tracking branch 'upstream/develop' into 1602-remove-simplepie

Browse source
This commit is contained in:
Michael Vogel 2016-02-16 07:58:01 +01:00
commit ade5ef7065
11 changed files with 320 additions and 239 deletions
Download patch file Download diff file Expand all files Collapse all files

86
include/Scrape.php
View file

@ -13,6 +13,20 @@ function scrape_dfrn($url, $dont_probe = false) {
logger('scrape_dfrn: url=' . $url);
// Try to fetch the data from noscrape. This is faster than parsing the HTML
$noscrape = str_replace("/hcard/", "/noscrape/", $url);
$noscrapejson = fetch_url($noscrape);
$noscrapedata = array();
if ($noscrapejson) {
$noscrapedata = json_decode($noscrapejson, true);
if (is_array($noscrapedata)) {
if ($noscrapedata["nick"] != "")
return($noscrapedata);
} else
$noscrapedata = array();
}
$s = fetch_url($url);
if(! $s)
@ -92,8 +106,7 @@ function scrape_dfrn($url, $dont_probe = false) {
}
}
}
return $ret;
return array_merge($ret, $noscrapedata);
}}
@ -743,18 +756,18 @@ function probe_url($url, $mode = PROBE_NORMAL, $level = 1) {
$vcard['fn'] = $url;
if (($notify != "") AND ($poll != "")) {
$baseurl = matching(normalise_link($notify), normalise_link($poll));
$baseurl = matching_url(normalise_link($notify), normalise_link($poll));
$baseurl2 = matching($baseurl, normalise_link($profile));
$baseurl2 = matching_url($baseurl, normalise_link($profile));
if ($baseurl2 != "")
$baseurl = $baseurl2;
}
if (($baseurl == "") AND ($notify != ""))
$baseurl = matching(normalise_link($profile), normalise_link($notify));
$baseurl = matching_url(normalise_link($profile), normalise_link($notify));
if (($baseurl == "") AND ($poll != ""))
$baseurl = matching(normalise_link($profile), normalise_link($poll));
$baseurl = matching_url(normalise_link($profile), normalise_link($poll));
$baseurl = rtrim($baseurl, "/");
@ -809,19 +822,56 @@ function probe_url($url, $mode = PROBE_NORMAL, $level = 1) {
return $result;
}
function matching($part1, $part2) {
$len = min(strlen($part1), strlen($part2));
/**
* @brief Find the matching part between two url
*
* @param string $url1
* @param string $url2
* @return string The matching part
*/
function matching_url($url1, $url2) {
if (($url1 == "") OR ($url2 == ""))
return "";
$url1 = normalise_link($url1);
$url2 = normalise_link($url2);
$parts1 = parse_url($url1);
$parts2 = parse_url($url2);
if (!isset($parts1["host"]) OR !isset($parts2["host"]))
return "";
if ($parts1["scheme"] != $parts2["scheme"])
return "";
if ($parts1["host"] != $parts2["host"])
return "";
if ($parts1["port"] != $parts2["port"])
return "";
$match = $parts1["scheme"]."://".$parts1["host"];
if ($parts1["port"])
$match .= ":".$parts1["port"];
$pathparts1 = explode("/", $parts1["path"]);
$pathparts2 = explode("/", $parts2["path"]);
$match = "";
$matching = true;
$i = 0;
while (($i <= $len) AND $matching) {
if (substr($part1, $i, 1) == substr($part2, $i, 1))
$match .= substr($part1, $i, 1);
else
$matching = false;
$path = "";
do {
$path1 = $pathparts1[$i];
$path2 = $pathparts2[$i];
$i++;
}
return($match);
if ($path1 == $path2)
$path .= $path1."/";
} while (($path1 == $path2) AND ($i++ <= count($pathparts1)));
$match .= $path;
return normalise_link($match);
}