searxng/searx/engines/wolframalpha_api.py

135 lines
4.2 KiB
Python
Raw Normal View History

# SPDX-License-Identifier: AGPL-3.0-or-later
"""
Wolfram|Alpha (Science)
"""
2015-12-23 06:01:00 +00:00
from lxml import etree
from urllib.parse import urlencode
2015-12-23 06:01:00 +00:00
# about
about = {
"website": 'https://www.wolframalpha.com',
"wikidata_id": 'Q207006',
"official_api_documentation": 'https://products.wolframalpha.com/api/',
"use_official_api": True,
"require_api_key": False,
"results": 'XML',
}
2015-12-23 06:01:00 +00:00
# search-url
search_url = 'https://api.wolframalpha.com/v2/query?appid={api_key}&{query}'
site_url = 'https://www.wolframalpha.com/input/?{query}'
2016-01-02 06:41:14 +00:00
api_key = '' # defined in settings.yml
2015-12-28 07:17:42 +00:00
2016-01-02 04:02:10 +00:00
# xpath variables
failure_xpath = '/queryresult[attribute::success="false"]'
2016-02-28 06:47:36 +00:00
input_xpath = '//pod[starts-with(attribute::id, "Input")]/subpod/plaintext'
pods_xpath = '//pod'
subpods_xpath = './subpod'
pod_primary_xpath = './@primary'
2016-02-28 06:47:36 +00:00
pod_id_xpath = './@id'
pod_title_xpath = './@title'
plaintext_xpath = './plaintext'
image_xpath = './img'
img_src_xpath = './@src'
img_alt_xpath = './@alt'
# pods to display as image in infobox
# this pods do return a plaintext, but they look better and are more useful as images
2016-02-28 06:47:36 +00:00
image_pods = {'VisualRepresentation',
'Illustration'}
2016-01-02 04:02:10 +00:00
2015-12-23 06:01:00 +00:00
# do search-request
def request(query, params):
2016-11-30 17:43:03 +00:00
params['url'] = search_url.format(query=urlencode({'input': query}), api_key=api_key)
params['headers']['Referer'] = site_url.format(query=urlencode({'i': query}))
2015-12-23 06:01:00 +00:00
return params
2015-12-28 07:17:42 +00:00
2015-12-23 06:01:00 +00:00
# replace private user area characters to make text legible
def replace_pua_chars(text):
pua_chars = {'\uf522': '\u2192', # rigth arrow
'\uf7b1': '\u2115', # set of natural numbers
'\uf7b4': '\u211a', # set of rational numbers
'\uf7b5': '\u211d', # set of real numbers
'\uf7bd': '\u2124', # set of integer numbers
'\uf74c': 'd', # differential
'\uf74d': '\u212f', # euler's number
'\uf74e': 'i', # imaginary number
'\uf7d9': '='} # equals sign
2015-12-23 06:01:00 +00:00
2016-11-30 17:43:03 +00:00
for k, v in pua_chars.items():
2015-12-23 06:01:00 +00:00
text = text.replace(k, v)
return text
2015-12-28 07:17:42 +00:00
2015-12-23 06:01:00 +00:00
# get response from search-request
def response(resp):
results = []
search_results = etree.XML(resp.content)
2015-12-23 06:01:00 +00:00
# return empty array if there are no results
2016-01-02 04:02:10 +00:00
if search_results.xpath(failure_xpath):
2015-12-23 06:01:00 +00:00
return []
2016-02-28 06:47:36 +00:00
try:
infobox_title = search_results.xpath(input_xpath)[0].text
except:
infobox_title = ""
pods = search_results.xpath(pods_xpath)
result_chunks = []
result_content = ""
for pod in pods:
2016-02-28 06:47:36 +00:00
pod_id = pod.xpath(pod_id_xpath)[0]
pod_title = pod.xpath(pod_title_xpath)[0]
pod_is_result = pod.xpath(pod_primary_xpath)
subpods = pod.xpath(subpods_xpath)
if not subpods:
continue
2016-02-28 06:47:36 +00:00
# Appends either a text or an image, depending on which one is more suitable
for subpod in subpods:
content = subpod.xpath(plaintext_xpath)[0].text
image = subpod.xpath(image_xpath)
2016-01-02 04:02:10 +00:00
2016-02-28 06:47:36 +00:00
if content and pod_id not in image_pods:
if pod_is_result or not result_content:
if pod_id != "Input":
result_content = "%s: %s" % (pod_title, content)
2016-02-28 06:47:36 +00:00
# if no input pod was found, title is first plaintext pod
if not infobox_title:
infobox_title = content
2016-02-28 06:47:36 +00:00
content = replace_pua_chars(content)
result_chunks.append({'label': pod_title, 'value': content})
elif image:
result_chunks.append({'label': pod_title,
'image': {'src': image[0].xpath(img_src_xpath)[0],
'alt': image[0].xpath(img_alt_xpath)[0]}})
if not result_chunks:
return []
title = "Wolfram|Alpha (%s)" % infobox_title
2016-02-28 06:47:36 +00:00
# append infobox
results.append({'infobox': infobox_title,
'attributes': result_chunks,
2016-11-30 17:43:03 +00:00
'urls': [{'title': 'Wolfram|Alpha', 'url': resp.request.headers['Referer']}]})
# append link to site
2016-11-30 17:43:03 +00:00
results.append({'url': resp.request.headers['Referer'],
'title': title,
'content': result_content})
2015-12-23 06:01:00 +00:00
return results