pepepper/tissue
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Merge pull request #134 from eai04191/feature/resolver-melonbooks-format

Browse Source 
MelonbooksResolverで余分な説明文を削除・整形する
This commit is contained in:
shibafu 2019-03-16 10:44:13 +09:00 committed by GitHub
commit 285e529aea
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 29 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

29
app/MetadataResolver/MelonbooksResolver.php
View File

@ -30,11 +30,40 @@ class MelonbooksResolver implements Resolver
if ($res->getStatusCode() === 200) { if ($res->getStatusCode() === 200) {
$metadata = $this->ogpResolver->parse($res->getBody()); $metadata = $this->ogpResolver->parse($res->getBody());
$dom = new \DOMDocument();
@$dom->loadHTML(mb_convert_encoding($res->getBody(), 'HTML-ENTITIES', 'UTF-8'));
$xpath = new \DOMXPath($dom);
$descriptionNodelist = $xpath->query('//div[@id="description"]//p');
$specialDescriptionNodelist = $xpath->query('//div[@id="special_description"]//p');
// censoredフラグの除去 // censoredフラグの除去
if (mb_strpos($metadata->image, '&c=1') !== false) { if (mb_strpos($metadata->image, '&c=1') !== false) {
$metadata->image = preg_replace('/&c=1/u', '', $metadata->image); $metadata->image = preg_replace('/&c=1/u', '', $metadata->image);
} }
// 抽出
preg_match('~^(.+)（(.+)）の通販・購入はメロンブックス$~', $metadata->title, $match);
$title = $match[1];
$maker = $match[2];
// 整形
$description = 'サークル: ' . $maker . "\n";
if ($specialDescriptionNodelist->length !== 0) {
$description .= trim(str_replace('<br>', "\n", $specialDescriptionNodelist->item(0)->nodeValue)) . "\n";
if ($specialDescriptionNodelist->length === 2) {
$description .= "\n";
$description .= trim(str_replace('<br>', "\n", $specialDescriptionNodelist->item(1)->nodeValue)) . "\n";
}
}
if ($descriptionNodelist->length !== 0) {
$description .= trim(str_replace('<br>', "\n", $descriptionNodelist->item(0)->nodeValue));
}
$metadata->title = $title;
$metadata->description = trim($description);
return $metadata; return $metadata;
} else { } else {
throw new \RuntimeException("{$res->getStatusCode()}: $url"); throw new \RuntimeException("{$res->getStatusCode()}: $url");