From c1d0fc6242ed01b79351af83535cd24c1e910420 Mon Sep 17 00:00:00 2001 From: j <0x006A@0x2620.org> Date: Tue, 14 Aug 2012 15:30:52 +0200 Subject: [PATCH] parse archive.org --- ox/web/archive.py | 18 +++++++++++++++++- 1 file changed, 17 insertions(+), 1 deletion(-) diff --git a/ox/web/archive.py b/ox/web/archive.py index 54c45a4..4be1e69 100644 --- a/ox/web/archive.py +++ b/ox/web/archive.py @@ -1,6 +1,7 @@ # -*- coding: utf-8 -*- # vi:si:et:sw=4:sts=4:ts=4 - +from .. import cache +from ..utils import json def getId(url): return url.split("/")[-1] @@ -8,3 +9,18 @@ def getId(url): def getUrl(id): return "http://www.archive.org/details/%s" % id +def getData(id): + data = {} + url = getUrl(id) + details = cache.readUrl('%s?output=json' % url) + details = json.loads(details) + for key in ('title', 'description', 'runtime'): + data[key] = details['metadata'][key] + if isinstance(data[key], list): + data[key] = data[key][0] + data['url'] = url + data['image'] = 'http://archive.org/download/%s/format=thumbnail' % id + data['ogg'] = 'http://archive.org/download/%s/format=Ogg+video' % id + data['mp4'] = 'http://archive.org/download/%s/format=512Kb+MPEG4' % id + return data +