[feat] engines: add pixabay for royalty free images/videos

What's changed?
- this PR adds Pixabay, a collection of royalty free images
- additionaly it seems to have some videos, so there's an engine for it too

Author Notes
- when using SearXNG's transport, all our requests will get blocked, probably due to fingerprinting
- we should find an alternative solution because this is just a hacky change to make things work for now, but idk how ...

1 files changed, 98 insertions, 0 deletions

diff --git a/searx/engines/pixabay.py b/searx/engines/pixabay.py
new file mode 100644
index 000000000..3a01d8af0
--- /dev/null
+++ b/searx/engines/pixabay.py
@@ -0,0 +1,98 @@
+# SPDX-License-Identifier: AGPL-3.0-or-later
+"""Pixabay provides royalty-free media (images, videos)"""
+
+from datetime import timedelta
+from urllib.parse import quote_plus, urlencode
+from dateutil import parser
+from searx.utils import gen_useragent
+
+# about
+about = {
+    "website": 'https://pixabay.com',
+    "wikidata_id": 'Q1746538',
+    "official_api_documentation": 'https://pixabay.com/api/docs/',
+    "use_official_api": False,
+    "require_api_key": False,
+    "results": 'JSON',
+}
+
+base_url = 'https://pixabay.com'
+categories = ['images']
+pixabay_type = "images"  # alternative: 'videos'
+
+paging = True
+safesearch = True
+time_range_support = True
+
+safesearch_map = {0: 'off', 1: '1', 2: '1'}
+time_range_map = {'day': '1d', 'week': '1w', 'month': '1m', 'year': '1y'}
+
+# using http2 returns forbidden errors
+enable_http2 = False
+
+
+def request(query, params):
+    args = {
+        'pagi': params['pageno'],
+    }
+    if params['time_range']:
+        args['date'] = time_range_map[params['time_range']]
+
+    params['url'] = f"{base_url}/{pixabay_type}/search/{quote_plus(query)}/?{urlencode(args)}"
+    params['headers'] = {
+        'User-Agent': gen_useragent() + " Pixabay",
+        'Accept': 'application/json',
+        'x-bootstrap-cache-miss': '1',
+        'x-fetch-bootstrap': '1',
+    }
+    params['cookies']['g_rated'] = safesearch_map[params['safesearch']]
+
+    # prevent automatic redirects to first page on pagination
+    params['allow_redirects'] = False
+
+    return params
+
+
+def _image_result(result):
+    return {
+        'template': 'images.html',
+        'url': base_url + result["href"],
+        # images are sorted in ascending quality
+        'thumbnail_src': list(result['sources'].values())[0],
+        'img_src': list(result['sources'].values())[-1],
+        'title': result.get('name'),
+        'content': result.get('description', ''),
+    }
+
+
+def _video_result(result):
+    return {
+        'template': 'videos.html',
+        'url': base_url + result["href"],
+        # images are sorted in ascending quality
+        'thumbnail': result['sources'].get('thumbnail'),
+        'iframe_src': result['sources'].get('embed'),
+        'title': result.get('name'),
+        'content': result.get('description', ''),
+        'length': timedelta(seconds=result['duration']),
+        'publishedDate': parser.parse(result['uploadDate']),
+    }
+
+
+def response(resp):
+    results = []
+
+    # if there are no results on this page, we get a redirect
+    # to the first page
+    if resp.status_code == 302:
+        return results
+
+    json_data = resp.json()
+
+    for result in json_data.get('page', {}).get('results', []):
+        if result['mediaType'] in ('photo', 'illustration', 'vector'):
+            results.append(_image_result(result))
+        elif result['mediaType'] == 'video':
+            results.append(_video_result(result))
+
+    return results