Merge 02ebea58fb into 0f9694c90b

The engine has been revised; there is now the option ``adobe_content_types``
with which it is possible to configure engines for video and audio from the
adobe stock.  BTW this patch adds documentation to the engine.

To test all three engines in one use a search term like::

    !asi !asv !asa sound

Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>

docs/dev/engines/online/adobe_stock.rst

@@ -0,0 +1,13 @@
+.. _adobe stock engine:
+
+===========
+Adobe Stock
+===========
+
+.. contents:: Contents
+   :depth: 2
+   :local:
+   :backlinks: entry
+
+.. automodule:: searx.engines.adobe_stock
+   :members:

docs/index.rst

@@ -4,22 +4,27 @@ Welcome to SearXNG
 
   *Search without being tracked.*
 
-SearXNG is a free internet metasearch engine which aggregates results from more
-than 70 search services.  Users are neither tracked nor profiled.  Additionally,
-SearXNG can be used over Tor for online anonymity.
+.. jinja:: searx
+
+   SearXNG is a free internet metasearch engine which aggregates results from up
+   to {{engines | length}} :ref:`search services <configured engines>`.  Users
+   are neither tracked nor profiled.  Additionally, SearXNG can be used over Tor
+   for online anonymity.
 
 Get started with SearXNG by using one of the instances listed at searx.space_.
 If you don't trust anyone, you can set up your own, see :ref:`installation`.
 
+.. jinja:: searx
+
    .. sidebar::  features
 
       - :ref:`self hosted <installation>`
       - :ref:`no user tracking / no profiling <SearXNG protect privacy>`
       - script & cookies are optional
       - secure, encrypted connections
-   - :ref:`about 200 search engines <configured engines>`
-   - `about 60 translations <https://translate.codeberg.org/projects/searxng/searxng/>`_
-   - about 100 `well maintained <https://uptime.searxng.org/>`__ instances on searx.space_
+      - :ref:`{{engines | length}} search engines <configured engines>`
+      - `58 translations <https://translate.codeberg.org/projects/searxng/searxng/>`_
+      - about 70 `well maintained <https://uptime.searxng.org/>`__ instances on searx.space_
       - :ref:`easy integration of search engines <demo online engine>`
       - professional development: `CI <https://github.com/searxng/searxng/actions>`_,
 	`quality assurance <https://dev.searxng.org/>`_ &

requirements.txt

@@ -19,3 +19,4 @@ tomli==2.0.2; python_version < '3.11'
 msgspec==0.18.6
 eval_type_backport; python_version < '3.9'
 typer-slim==0.13.1
+isodate==0.7.2

searx/engines/adobe_stock.py

@@ -1,67 +1,229 @@
 # SPDX-License-Identifier: AGPL-3.0-or-later
-"""Adobe Stock (images)
-"""
+"""`Adobe Stock`_ is a service that gives access to millions of royalty-free
+assets. Assets types include photos, vectors, illustrations, templates, 3D
+assets, videos, motion graphics templates and audio tracks.
 
+.. Adobe Stock: https://stock.adobe.com/
+
+Configuration
+=============
+
+The engine has the following mandatory setting:
+
+- SearXNG's :ref:`engine categories`
+- Adobe-Stock's :py:obj:`adobe_order`
+- Adobe-Stock's :py:obj:`adobe_content_types`
+
+.. code:: yaml
+
+  - name: adobe stock
+    engine: adobe_stock
+    shortcut: asi
+    categories: [images]
+    adobe_order: relevance
+    adobe_content_types: ["photo", "illustration", "zip_vector", "template", "3d", "image"]
+
+  - name: adobe stock video
+    engine: adobe_stock
+    network: adobe stock
+    shortcut: asi
+    categories: [videos]
+    adobe_order: relevance
+    adobe_content_types: ["video"]
+
+Implementation
+==============
+
+"""
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+from datetime import datetime, timedelta
 from urllib.parse import urlencode
-from searx.utils import gen_useragent
+
+import isodate
+
+if TYPE_CHECKING:
+    import logging
+
+    logger: logging.Logger
 
 about = {
-    "website": 'https://stock.adobe.com/',
-    "wikidata_id": 'Q5977430',
+    "website": "https://stock.adobe.com/",
+    "wikidata_id": "Q5977430",
     "official_api_documentation": None,
     "use_official_api": False,
     "require_api_key": False,
-    "results": 'JSON',
+    "results": "JSON",
 }
 
-categories = ['images']
+categories = []
 paging = True
-
-base_url = 'https://stock.adobe.com'
-
+send_accept_language_header = True
 results_per_page = 10
-adobe_order = "relevance"  # one of 'relevant', 'featured', 'creation' or 'nb_downloads'
+
+base_url = "https://stock.adobe.com"
+
+adobe_order: str = ""
+"""Sort order, can be one of:
+
+- ``relevance`` or
+- ``featured`` or
+- ``creation`` (most recent) or
+- ``nb_downloads`` (number of downloads)
+"""
+
+ADOBE_VALID_TYPES = ["photo", "illustration", "zip_vector", "video", "template", "3d", "audio", "image"]
+adobe_content_types: list = []
+"""A list of of content types.  The following content types are offered:
+
+- Images: ``image``
+- Videos: ``video``
+- Templates: ``template``
+- 3D: ``3d``
+- Audio ``audio``
+
+Additional subcategories:
+
+- Photos: ``photo``
+- Illustrations: ``illustration``
+- Vectors: ``zip_vector`` (Vectors),
+"""
+
+# Do we need support for "free_collection" and "include_stock_enterprise"?
+
+
+def init(_):
+    if not categories:
+        raise ValueError("adobe_stock engine: categories is unset")
+
+    # adobe_order
+    if not adobe_order:
+        raise ValueError("adobe_stock engine: adobe_order is unset")
+    if adobe_order not in ["relevance", "featured", "creation", "nb_downloads"]:
+        raise ValueError(f"unsupported adobe_order: {adobe_order}")
+
+    # adobe_content_types
+    if not adobe_content_types:
+        raise ValueError("adobe_stock engine: adobe_content_types is unset")
+
+    if isinstance(adobe_content_types, list):
+        for t in adobe_content_types:
+            if t not in ADOBE_VALID_TYPES:
+                raise ValueError("adobe_stock engine: adobe_content_types: '%s' is invalid" % t)
+    else:
+        raise ValueError(
+            "adobe_stock engine: adobe_content_types must be a list of strings not %s" % type(adobe_content_types)
+        )
 
 
 def request(query, params):
+
     args = {
-        'k': query,
-        'limit': results_per_page,
-        'order': adobe_order,
-        'search_page': params['pageno'],
-        'search_type': 'pagination',
-        'filters[content_type:video]': 0,
-        'filters[content_type:audio]': 0,
+        "k": query,
+        "limit": results_per_page,
+        "order": adobe_order,
+        "search_page": params["pageno"],
+        "search_type": "pagination",
     }
-    params['url'] = f"{base_url}/de/Ajax/Search?{urlencode(args)}"
+
+    for content_type in ADOBE_VALID_TYPES:
+        args[f"filters[content_type:{content_type}]"] = 1 if content_type in adobe_content_types else 0
+
+    params["url"] = f"{base_url}/de/Ajax/Search?{urlencode(args)}"
 
     # headers required to bypass bot-detection
-    params['headers'] = {
-        "User-Agent": gen_useragent(),
-        "Accept-Language": "en-US,en;q=0.5",
-    }
+    if params["searxng_locale"] == "all":
+        params["headers"]["Accept-Language"] = "en-US,en;q=0.5"
 
     return params
 
 
+def parse_image_item(item):
+    return {
+        "template": "images.html",
+        "url": item["content_url"],
+        "title": item["title"],
+        "content": item["asset_type"],
+        "img_src": item["content_thumb_extra_large_url"],
+        "thumbnail_src": item["thumbnail_url"],
+        "resolution": f"{item['content_original_width']}x{item['content_original_height']}",
+        "img_format": item["format"],
+        "author": item["author"],
+    }
+
+
+def parse_video_item(item):
+
+    # in video items, the title is more or less a "content description", we try
+    # to reduce the lenght of the title ..
+
+    title = item["title"]
+    content = ""
+    if "." in title.strip()[:-1]:
+        content = title
+        title = title.split(".", 1)[0]
+    elif "," in title:
+        content = title
+        title = title.split(",", 1)[0]
+    elif len(title) > 50:
+        content = title
+        title = ""
+        for w in content.split(" "):
+            title += f" {w}"
+            if len(title) > 50:
+                title = title.strip() + "\u2026"
+                break
+
+    return {
+        "template": "videos.html",
+        "url": item["content_url"],
+        "title": title,
+        "content": content,
+        # https://en.wikipedia.org/wiki/ISO_8601#Durations
+        "length": isodate.parse_duration(item["time_duration"]),
+        "publishedDate": datetime.strptime(item["creation_date"], "%Y-%m-%d"),
+        "thumbnail": item["thumbnail_url"],
+        "iframe_src": item["video_small_preview_url"],
+        "metadata": item["asset_type"],
+    }
+
+
+def parse_audio_item(item):
+    audio_data = item["audio_data"]
+    content = audio_data.get("description") or ""
+    if audio_data.get("album"):
+        content = audio_data["album"] + " - " + content
+
+    return {
+        "url": item["content_url"],
+        "title": item["title"],
+        "content": content,
+        # "thumbnail": base_url + item["thumbnail_url"],
+        "iframe_src": audio_data["preview"]["url"],
+        "publishedDate": datetime.fromisoformat(audio_data["release_date"]) if audio_data["release_date"] else None,
+        "length": timedelta(seconds=round(audio_data["duration"] / 1000)) if audio_data["duration"] else None,
+        "author": item.get("artist_name"),
+    }
+
+
 def response(resp):
     results = []
 
     json_resp = resp.json()
 
-    for item in json_resp['items'].values():
-        results.append(
-            {
-                'template': 'images.html',
-                'url': item['content_url'],
-                'title': item['title'],
-                'content': '',
-                'img_src': item['content_thumb_extra_large_url'],
-                'thumbnail_src': item['thumbnail_url'],
-                'resolution': f"{item['content_original_width']}x{item['content_original_height']}",
-                'img_format': item['format'],
-                'author': item['author'],
-            }
-        )
+    if isinstance(json_resp["items"], list):
+        return None
+    for item in json_resp["items"].values():
+        if item["asset_type"].lower() in ["image", "premium-image", "illustration", "vector"]:
+            result = parse_image_item(item)
+        elif item["asset_type"].lower() == "video":
+            result = parse_video_item(item)
+        elif item["asset_type"].lower() == "audio":
+            result = parse_audio_item(item)
+        else:
+            logger.error("no handle for %s --> %s", item["asset_type"], item)
+            continue
+        results.append(result)
 
     return results

searx/engines/internet_archive_scholar.py

@@ -1,71 +0,0 @@
-# SPDX-License-Identifier: AGPL-3.0-or-later
-"""Internet Archive scholar(science)
-"""
-
-from datetime import datetime
-from urllib.parse import urlencode
-from searx.utils import html_to_text
-
-about = {
-    "website": "https://scholar.archive.org/",
-    "wikidata_id": "Q115667709",
-    "official_api_documentation": "https://scholar.archive.org/api/redoc",
-    "use_official_api": True,
-    "require_api_key": False,
-    "results": "JSON",
-}
-categories = ['science', 'scientific publications']
-paging = True
-
-base_url = "https://scholar.archive.org"
-results_per_page = 15
-
-
-def request(query, params):
-    args = {
-        "q": query,
-        "limit": results_per_page,
-        "offset": (params["pageno"] - 1) * results_per_page,
-    }
-    params["url"] = f"{base_url}/search?{urlencode(args)}"
-    params["headers"]["Accept"] = "application/json"
-    return params
-
-
-def response(resp):
-    results = []
-
-    json = resp.json()
-
-    for result in json["results"]:
-        publishedDate, content, doi = None, '', None
-
-        if result['biblio'].get('release_date'):
-            publishedDate = datetime.strptime(result['biblio']['release_date'], "%Y-%m-%d")
-
-        if len(result['abstracts']) > 0:
-            content = result['abstracts'][0].get('body')
-        elif len(result['_highlights']) > 0:
-            content = result['_highlights'][0]
-
-        if len(result['releases']) > 0:
-            doi = result['releases'][0].get('doi')
-
-        results.append(
-            {
-                'template': 'paper.html',
-                'url': result['fulltext']['access_url'],
-                'title': result['biblio'].get('title') or result['biblio'].get('container_name'),
-                'content': html_to_text(content),
-                'publisher': result['biblio'].get('publisher'),
-                'doi': doi,
-                'journal': result['biblio'].get('container_name'),
-                'authors': result['biblio'].get('contrib_names'),
-                'tags': result['tags'],
-                'publishedDate': publishedDate,
-                'issns': result['biblio'].get('issns'),
-                'pdf_url': result['fulltext'].get('access_url'),
-            }
-        )
-
-    return results

searx/engines/loc.py

@@ -27,7 +27,7 @@ categories = ['images']
 paging = True
 
 endpoint = 'photos'
-base_url = 'https://loc.gov'
+base_url = 'https://www.loc.gov'
 search_string = "/{endpoint}/?sp={page}&{query}&fo=json"
 
 

searx/network/network.py

@@ -233,8 +233,7 @@ class Network:
             del kwargs['raise_for_httperror']
         return do_raise_for_httperror
 
-    @staticmethod
-    def patch_response(response, do_raise_for_httperror):
+    def patch_response(self, response, do_raise_for_httperror):
         if isinstance(response, httpx.Response):
             # requests compatibility (response is not streamed)
             # see also https://www.python-httpx.org/compatibility/#checking-for-4xx5xx-responses
@@ -242,8 +241,11 @@ class Network:
 
             # raise an exception
             if do_raise_for_httperror:
+                try:
                     raise_for_httperror(response)
-
+                except:
+                    self._logger.warning(f"HTTP Request failed: {response.request.method} {response.request.url}")
+                    raise
         return response
 
     def is_valid_response(self, response):
@@ -269,7 +271,7 @@ class Network:
                 else:
                     response = await client.request(method, url, **kwargs)
                 if self.is_valid_response(response) or retries <= 0:
-                    return Network.patch_response(response, do_raise_for_httperror)
+                    return self.patch_response(response, do_raise_for_httperror)
             except httpx.RemoteProtocolError as e:
                 if not was_disconnected:
                     # the server has closed the connection:

searx/search/processors/online.py

@@ -137,9 +137,6 @@ class OnlineProcessor(EngineProcessor):
         self.engine.request(query, params)
 
         # ignoring empty urls
-        if params['url'] is None:
-            return None
-
         if not params['url']:
             return None
 

searx/settings.yml

@@ -327,9 +327,32 @@ engines:
 
   - name: adobe stock
     engine: adobe_stock
-    # available search orders: 'relevant', 'featured', 'creation', 'nb_downloads'
-    # adobe_order: relevance
-    shortcut: as
+    shortcut: asi
+    categories: ["images"]
+    # https://docs.searxng.org/dev/engines/online/adobe_stock.html
+    adobe_order: relevance
+    adobe_content_types: ["photo", "illustration", "zip_vector", "template", "3d", "image"]
+    timeout: 6
+    disabled: true
+
+  - name: adobe stock video
+    engine: adobe_stock
+    shortcut: asv
+    network: adobe stock
+    categories: ["videos"]
+    adobe_order: relevance
+    adobe_content_types: ["video"]
+    timeout: 6
+    disabled: true
+
+  - name: adobe stock audio
+    engine: adobe_stock
+    shortcut: asa
+    network: adobe stock
+    categories: ["music"]
+    adobe_order: relevance
+    adobe_content_types: ["audio"]
+    timeout: 6
     disabled: true
 
   - name: alpine linux packages
@@ -1629,11 +1652,6 @@ engines:
     api_site: 'askubuntu'
     categories: [it, q&a]
 
-  - name: internetarchivescholar
-    engine: internet_archive_scholar
-    shortcut: ias
-    timeout: 15.0
-
   - name: superuser
     engine: stackexchange
     shortcut: su