From 7b4425b096fa8c18d0db9fd9b1ae96d63ee8af55 Mon Sep 17 00:00:00 2001 From: Christian Weiske Date: Mon, 1 Feb 2016 20:18:59 +0100 Subject: first kinda working version --- bin/crawl.php | 145 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 145 insertions(+) create mode 100755 bin/crawl.php (limited to 'bin/crawl.php') diff --git a/bin/crawl.php b/bin/crawl.php new file mode 100755 index 0000000..26cf994 --- /dev/null +++ b/bin/crawl.php @@ -0,0 +1,145 @@ +#!/usr/bin/env php +send(); +if ($res->getStatus() !== 200) { + echo "Response code is not 200 but " . $res->getStatus() . ", stopping\n"; + exit(3); +} +$mimetype = explode(';', $res->getHeader('content-type'))[0]; +if (!in_array($mimetype, $supportedCrawlTypes)) { + echo "MIME type not supported for crawling: $mimetype\n"; + exit(4); +} + +//FIXME: mime type switch for cdata +$doc = new \DOMDocument(); +//@ to hide parse warning messages in invalid html +@$doc->loadHTMLFile($url); + +//FIXME: extract base url from html +$base = new \Net_URL2($url); + +$xpath = new \DOMXPath($doc); +$links = $xpath->evaluate('//a'); +//FIXME: link rel, img, video + +$alreadySeen = array(); + +foreach ($links as $link) { + $linkTitle = $link->textContent; + $href = ''; + foreach ($link->attributes as $attribute) { + if ($attribute->name == 'href') { + $href = $attribute->textContent; + } + } + if ($href == '' || $href{0} == '#') { + //link on this page + continue; + } + + $linkUrlObj = $base->resolve($href); + $linkUrlObj->setFragment(false); + $linkUrl = (string) $linkUrlObj; + if (isset($alreadySeen[$linkUrl])) { + continue; + } + + switch ($linkUrlObj->getScheme()) { + case 'http': + case 'https': + break; + default: + continue 2; + } + + if ($es->isKnown($linkUrl)) { + continue; + } + + //FIXME: check target type + //FIXME: check nofollow + //var_dump($linkTitle, $linkUrl); + $es->markQueued($linkUrl); + addToIndex($linkUrl, $linkTitle, $url); + if (isUrlAllowed($linkUrl)) { + addToCrawl($linkUrl); + } + $alreadySeen[$linkUrl] = true; +} + +function addToIndex($linkUrl, $linkTitle, $sourceUrl) +{ + echo "Queuing for indexing: $linkUrl\n"; + $gmclient = new \GearmanClient(); + $gmclient->addServer('127.0.0.1'); + $gmclient->doBackground( + 'phinde_index', + serialize( + array( + 'url' => $linkUrl, + 'title' => $linkTitle, + 'source' => $sourceUrl + ) + ) + ); + if ($gmclient->returnCode() != GEARMAN_SUCCESS) { + echo 'Error queueing URL indexing for ' + . $linkUrl . "\n" + . 'Error code: ' . $gmclient->returnCode() . "\n"; + exit(2); + } +} + +function addToCrawl($linkUrl) +{ + echo "Queuing for crawling: $linkUrl\n"; + $gmclient = new \GearmanClient(); + $gmclient->addServer('127.0.0.1'); + $gmclient->doBackground( + 'phinde_crawl', + serialize( + array( + 'url' => $linkUrl + ) + ) + ); + if ($gmclient->returnCode() != GEARMAN_SUCCESS) { + echo 'Error queueing URL crawling for ' + . $linkUrl . "\n" + . 'Error code: ' . $gmclient->returnCode() . "\n"; + exit(2); + } +} +?> \ No newline at end of file -- cgit v1.2.3