Merge pull request #30 from matejc/smallissues

fix: robot fw, entry points, some flake8, package searx egg
This commit is contained in:
Adam Tauber 2014-01-19 16:29:56 -08:00
commit ffc93ba256
17 changed files with 293 additions and 130 deletions

20
.gitignore vendored
View File

@ -1,20 +1,24 @@
env
engines.cfg
.installed.cfg
.coverage
coverage/
.installed.cfg
engines.cfg
env
robot_log.html
robot_output.xml
robot_report.html
setup.cfg
*.pyc
*/*.pyc
bin/
build/
covearge/
develop-eggs/
dist/
eggs/
include/
lib/
build/
develop-eggs/
eggs/
local/
searx.egg-info/
parts/
searx.egg-info/
var/

View File

@ -21,11 +21,7 @@ $(python):
tests: .installed.cfg
@bin/test
enginescfg:
@test -f ./engines.cfg || echo "Copying engines.cfg ..."
@cp --no-clobber engines.cfg_sample engines.cfg
robot: .installed.cfg enginescfg
robot: .installed.cfg
@bin/robot
flake8: .installed.cfg
@ -37,18 +33,18 @@ coverage: .installed.cfg
@bin/coverage report --show-missing
@bin/coverage html --directory ./coverage
production: bin/buildout production.cfg setup.py enginescfg
production: bin/buildout production.cfg setup.py
bin/buildout -c production.cfg $(options)
@echo "* Please modify `readlink --canonicalize-missing ./searx/settings.py`"
@echo "* Hint 1: on production, disable debug mode and change secret_key"
@echo "* Hint 2: searx will be executed at server startup by crontab"
@echo "* Hint 3: to run immediatley, execute 'bin/supervisord'"
minimal: bin/buildout minimal.cfg setup.py enginescfg
minimal: bin/buildout minimal.cfg setup.py
bin/buildout -c minimal.cfg $(options)
clean:
@rm -rf .installed.cfg .mr.developer.cfg bin parts develop-eggs \
searx.egg-info lib include .coverage coverage
.PHONY: all tests enginescfg robot flake8 coverage production minimal clean
.PHONY: all tests robot flake8 coverage production minimal clean

View File

@ -16,8 +16,6 @@ recipe = zc.recipe.egg:script
eggs = ${buildout:eggs}
interpreter = py
dependent-scripts = true
entry-points =
searx-run=searx.webapp:run
[robot]

View File

@ -13,5 +13,3 @@ parts +=
recipe = zc.recipe.egg:script
eggs = ${buildout:eggs}
interpreter = py
entry-points =
searx-run=searx.webapp:run

View File

@ -15,8 +15,6 @@ parts +=
recipe = zc.recipe.egg:script
eggs = ${buildout:eggs}
interpreter = py
entry-points =
searx-run=searx.webapp:run
[supervisor]

View File

@ -1,5 +1,5 @@
from os import environ
from os.path import realpath, dirname, join
from os.path import realpath, dirname, join, abspath
try:
from yaml import load
except:
@ -7,8 +7,7 @@ except:
stderr.write('[E] install pyyaml\n')
exit(2)
searx_dir = realpath(dirname(realpath(__file__))+'/../')
searx_dir = abspath(dirname(__file__))
engine_dir = dirname(realpath(__file__))
if 'SEARX_SETTINGS_PATH' in environ:
@ -19,4 +18,3 @@ else:
with open(settings_path) as settings_yaml:
settings = load(settings_yaml)

View File

@ -35,6 +35,7 @@ engines = {}
categories = {'general': []}
def load_module(filename):
modname = splitext(filename)[0]
if modname in sys.modules:
@ -50,7 +51,7 @@ if not 'engines' in settings or not settings['engines']:
for engine_data in settings['engines']:
engine_name = engine_data['engine']
engine = load_module(engine_name+'.py')
engine = load_module(engine_name + '.py')
for param_name in engine_data:
if param_name == 'engine':
continue
@ -58,38 +59,50 @@ for engine_data in settings['engines']:
if engine_data['categories'] == 'none':
engine.categories = []
else:
engine.categories = map(str.strip, engine_data['categories'].split(','))
engine.categories = map(
str.strip, engine_data['categories'].split(','))
continue
setattr(engine, param_name, engine_data[param_name])
for engine_attr in dir(engine):
if engine_attr.startswith('_'):
continue
if getattr(engine, engine_attr) == None:
print '[E] Engine config error: Missing attribute "{0}.{1}"'.format(engine.name, engine_attr)
print '[E] Engine config error: Missing attribute "{0}.{1}"'.format(engine.name, engine_attr) # noqa
sys.exit(1)
engines[engine.name] = engine
engine.stats = {'result_count': 0, 'search_count': 0, 'page_load_time': 0, 'score_count': 0, 'errors': 0}
engine.stats = {
'result_count': 0,
'search_count': 0,
'page_load_time': 0,
'score_count': 0,
'errors': 0
}
if hasattr(engine, 'categories'):
for category_name in engine.categories:
categories.setdefault(category_name, []).append(engine)
else:
categories['general'].append(engine)
def default_request_params():
return {'method': 'GET', 'headers': {}, 'data': {}, 'url': '', 'cookies': {}}
return {
'method': 'GET', 'headers': {}, 'data': {}, 'url': '', 'cookies': {}}
def make_callback(engine_name, results, suggestions, callback, params):
# creating a callback wrapper for the search engine results
def process_callback(response, **kwargs):
cb_res = []
response.search_params = params
engines[engine_name].stats['page_load_time'] += (datetime.now() - params['started']).total_seconds()
engines[engine_name].stats['page_load_time'] += \
(datetime.now() - params['started']).total_seconds()
try:
search_results = callback(response)
except Exception, e:
engines[engine_name].stats['errors'] += 1
results[engine_name] = cb_res
print '[E] Error with engine "{0}":\n\t{1}'.format(engine_name, str(e))
print '[E] Error with engine "{0}":\n\t{1}'.format(
engine_name, str(e))
return
for result in search_results:
result['engine'] = engine_name
@ -101,23 +114,25 @@ def make_callback(engine_name, results, suggestions, callback, params):
results[engine_name] = cb_res
return process_callback
def score_results(results):
flat_res = filter(None, chain.from_iterable(izip_longest(*results.values())))
flat_res = filter(
None, chain.from_iterable(izip_longest(*results.values())))
flat_len = len(flat_res)
engines_len = len(results)
results = []
# deduplication + scoring
for i,res in enumerate(flat_res):
for i, res in enumerate(flat_res):
res['parsed_url'] = urlparse(res['url'])
res['engines'] = [res['engine']]
weight = 1.0
if hasattr(engines[res['engine']], 'weight'):
weight = float(engines[res['engine']].weight)
score = int((flat_len - i)/engines_len)*weight+1
score = int((flat_len - i) / engines_len) * weight + 1
duplicated = False
for new_res in results:
p1 = res['parsed_url'].path[:-1] if res['parsed_url'].path.endswith('/') else res['parsed_url'].path
p2 = new_res['parsed_url'].path[:-1] if new_res['parsed_url'].path.endswith('/') else new_res['parsed_url'].path
p1 = res['parsed_url'].path[:-1] if res['parsed_url'].path.endswith('/') else res['parsed_url'].path # noqa
p2 = new_res['parsed_url'].path[:-1] if new_res['parsed_url'].path.endswith('/') else new_res['parsed_url'].path # noqa
if res['parsed_url'].netloc == new_res['parsed_url'].netloc and\
p1 == p2 and\
res['parsed_url'].query == new_res['parsed_url'].query and\
@ -125,7 +140,7 @@ def score_results(results):
duplicated = new_res
break
if duplicated:
if len(res.get('content', '')) > len(duplicated.get('content', '')):
if len(res.get('content', '')) > len(duplicated.get('content', '')): # noqa
duplicated['content'] = res['content']
duplicated['score'] += score
duplicated['engines'].append(res['engine'])
@ -139,6 +154,7 @@ def score_results(results):
results.append(res)
return sorted(results, key=itemgetter('score'), reverse=True)
def search(query, request, selected_engines):
global engines, categories, number_of_searches
requests = []
@ -160,12 +176,19 @@ def search(query, request, selected_engines):
request_params['started'] = datetime.now()
request_params = engine.request(query, request_params)
callback = make_callback(selected_engine['name'], results, suggestions, engine.response, request_params)
callback = make_callback(
selected_engine['name'],
results,
suggestions,
engine.response,
request_params
)
request_args = dict(headers = request_params['headers']
,hooks = dict(response=callback)
,cookies = request_params['cookies']
,timeout = settings['server']['request_timeout']
request_args = dict(
headers=request_params['headers'],
hooks=dict(response=callback),
cookies=request_params['cookies'],
timeout=settings['server']['request_timeout']
)
if request_params['method'] == 'GET':
@ -180,7 +203,7 @@ def search(query, request, selected_engines):
requests.append(req(request_params['url'], **request_args))
grequests.map(requests)
for engine_name,engine_results in results.items():
for engine_name, engine_results in results.items():
engines[engine_name].stats['search_count'] += 1
engines[engine_name].stats['result_count'] += len(engine_results)
@ -192,6 +215,7 @@ def search(query, request, selected_engines):
return results, suggestions
def get_engines_stats():
# TODO refactor
pageloads = []
@ -200,14 +224,15 @@ def get_engines_stats():
errors = []
scores_per_result = []
max_pageload = max_results = max_score = max_errors = max_score_per_result = 0
max_pageload = max_results = max_score = max_errors = max_score_per_result = 0 # noqa
for engine in engines.values():
if engine.stats['search_count'] == 0:
continue
results_num = engine.stats['result_count']/float(engine.stats['search_count'])
load_times = engine.stats['page_load_time']/float(engine.stats['search_count'])
results_num = \
engine.stats['result_count'] / float(engine.stats['search_count'])
load_times = engine.stats['page_load_time'] / float(engine.stats['search_count']) # noqa
if results_num:
score = engine.stats['score_count'] / float(engine.stats['search_count'])
score = engine.stats['score_count'] / float(engine.stats['search_count']) # noqa
score_per_result = score / results_num
else:
score = score_per_result = 0.0
@ -220,30 +245,39 @@ def get_engines_stats():
results.append({'avg': results_num, 'name': engine.name})
scores.append({'avg': score, 'name': engine.name})
errors.append({'avg': engine.stats['errors'], 'name': engine.name})
scores_per_result.append({'avg': score_per_result, 'name': engine.name})
scores_per_result.append({
'avg': score_per_result,
'name': engine.name
})
for engine in pageloads:
engine['percentage'] = int(engine['avg']/max_pageload*100)
engine['percentage'] = int(engine['avg'] / max_pageload * 100)
for engine in results:
engine['percentage'] = int(engine['avg']/max_results*100)
engine['percentage'] = int(engine['avg'] / max_results * 100)
for engine in scores:
engine['percentage'] = int(engine['avg']/max_score*100)
engine['percentage'] = int(engine['avg'] / max_score * 100)
for engine in scores_per_result:
engine['percentage'] = int(engine['avg']/max_score_per_result*100)
engine['percentage'] = int(engine['avg'] / max_score_per_result * 100)
for engine in errors:
if max_errors:
engine['percentage'] = int(float(engine['avg'])/max_errors*100)
engine['percentage'] = int(float(engine['avg']) / max_errors * 100)
else:
engine['percentage'] = 0
return [('Page loads (sec)', sorted(pageloads, key=itemgetter('avg')))
,('Number of results', sorted(results, key=itemgetter('avg'), reverse=True))
,('Scores', sorted(scores, key=itemgetter('avg'), reverse=True))
,('Scores per result', sorted(scores_per_result, key=itemgetter('avg'), reverse=True))
,('Errors', sorted(errors, key=itemgetter('avg'), reverse=True))
return [
('Page loads (sec)', sorted(pageloads, key=itemgetter('avg'))),
(
'Number of results',
sorted(results, key=itemgetter('avg'), reverse=True)
),
('Scores', sorted(scores, key=itemgetter('avg'), reverse=True)),
(
'Scores per result',
sorted(scores_per_result, key=itemgetter('avg'), reverse=True)
),
('Errors', sorted(errors, key=itemgetter('avg'), reverse=True)),
]

View File

@ -8,7 +8,8 @@ locale = 'en-US' # see http://msdn.microsoft.com/en-us/library/dd251064.aspx
def request(query, params):
search_path = search_string.format(query=urlencode({'q': query, 'setmkt': locale}))
search_path = search_string.format(
query=urlencode({'q': query, 'setmkt': locale}))
#if params['category'] == 'images':
# params['url'] = base_url + 'images/' + search_path
params['url'] = base_url + search_path

View File

@ -7,6 +7,7 @@ weight = 100
parser_re = re.compile(r'^\W*(\d+(?:\.\d+)?)\W*([a-z]{3})\W*(?:in)?\W*([a-z]{3})\W*$', re.I)
def request(query, params):
m = parser_re.match(query)
if not m:
@ -19,7 +20,7 @@ def request(query, params):
# wrong params
return params
q = (from_currency+to_currency).upper()
q = (from_currency + to_currency).upper()
params['url'] = url.format(query=q)
params['ammount'] = ammount
@ -33,24 +34,26 @@ def response(resp):
global base_url
results = []
try:
_,conversion_rate,_ = resp.text.split(',', 2)
_, conversion_rate, _ = resp.text.split(',', 2)
conversion_rate = float(conversion_rate)
except:
return results
title = '{0} {1} in {2} is {3}'.format(resp.search_params['ammount']
,resp.search_params['from']
,resp.search_params['to']
,resp.search_params['ammount']*conversion_rate
title = '{0} {1} in {2} is {3}'.format(
resp.search_params['ammount'],
resp.search_params['from'],
resp.search_params['to'],
resp.search_params['ammount'] * conversion_rate
)
content = '1 {0} is {1} {2}'.format(resp.search_params['from'], conversion_rate, resp.search_params['to'])
now_date = datetime.now().strftime('%Y%m%d')
url = 'http://finance.yahoo.com/currency/converter-results/{0}/{1}-{2}-to-{3}.html'
url = url.format(now_date
,resp.search_params['ammount']
,resp.search_params['from'].lower()
,resp.search_params['to'].lower()
url = url.format(
now_date,
resp.search_params['ammount'],
resp.search_params['from'].lower(),
resp.search_params['to'].lower()
)
results.append({'title': title, 'content': content, 'url': url})

View File

@ -8,9 +8,11 @@ locale = 'en_US'
# see http://www.dailymotion.com/doc/api/obj-video.html
search_url = 'https://api.dailymotion.com/videos?fields=title,description,duration,url,thumbnail_360_url&sort=relevance&limit=25&page=1&{query}'
def request(query, params):
global search_url
params['url'] = search_url.format(query=urlencode({'search': query, 'localization': locale }))
params['url'] = search_url.format(
query=urlencode({'search': query, 'localization': locale}))
return params
@ -32,6 +34,7 @@ def response(resp):
results.append({'url': url, 'title': title, 'content': content})
return results
def text_content_from_html(html_string):
desc_html = html.fragment_fromstring(html_string, create_parent=True)
return desc_html.text_content()

View File

@ -1,16 +0,0 @@
port = 11111
secret_key = "ultrasecretkey" # change this!
debug = False
request_timeout = 5.0 # seconds
weights = {} # 'search_engine_name': float(weight) | default is 1.0
blacklist = [] # search engine blacklist
categories = {} # custom search engine categories
base_url = None # "https://your.domain.tld/" or None (to use request parameters)

107
searx/settings_robot.yml Normal file
View File

@ -0,0 +1,107 @@
server:
port : 11111
secret_key : "ultrasecretkey" # change this!
debug : False
request_timeout : 3.0 # seconds
base_url: False
engines:
- name : wikipedia
engine : mediawiki
url : https://en.wikipedia.org/
number_of_results : 1
- name : bing
engine : bing
locale : en-US
- name : currency
engine : currency_convert
categories : general
- name : deviantart
engine : deviantart
categories : images
- name : ddg definitions
engine : duckduckgo_definitions
- name : duckduckgo
engine : duckduckgo
locale : en-us
- name : filecrop
engine : filecrop
categories : files
- name : flickr
engine : flickr
categories : images
- name : github
engine : github
categories : it
- name : google
engine : json_engine
search_url : https://ajax.googleapis.com/ajax/services/search/web?v=2.0&start=0&rsz=large&safe=off&filter=off&q={query}
categories : general
url_query : /responseData/results/unescapedUrl
content_query : /responseData/results/content
title_query : /responseData/results/titleNoFormatting
- name : google images
engine : google_images
categories : images
- name : piratebay
engine : piratebay
categories : videos, music, files
- name : soundcloud
engine : soundcloud
categories : music
- name : stackoverflow
engine : stackoverflow
categories : it
- name : startpage
engine : startpage
- name : twitter
engine : twitter
categories : social media
- name : urbandictionary
engine : xpath
search_url : http://www.urbandictionary.com/define.php?term={query}
url_xpath : //div[@class="word"]//a/@href
title_xpath : //div[@class="word"]//a
content_xpath : //div[@class="definition"]
- name : yahoo
engine : xpath
search_url : http://search.yahoo.com/search?p={query}
results_xpath : //div[@class="res"]
url_xpath : .//h3/a/@href
title_xpath : .//h3/a
content_xpath : .//div[@class="abstr"]
suggestion_xpath : //div[@id="satat"]//a
- name : youtube
engine : youtube
categories : videos
- name : dailymotion
engine : dailymotion
locale : en_US
categories : videos
- name : vimeo
engine : vimeo
categories : videos
results_xpath : //div[@id="browse_content"]/ol/li
url_xpath : ./a/@href
title_xpath : ./a/div[@class="data"]/p[@class="title"]/text()
content_xpath : ./a/img/@src

View File

@ -7,10 +7,10 @@ from unittest2 import TestCase
import os
import subprocess
import sys
class SearxTestLayer:
"""Base layer for non-robot tests."""
__name__ = u'SearxTestLayer'
@ -36,24 +36,37 @@ class SearxRobotLayer(Layer):
def setUp(self):
os.setpgrp() # create new process group, become its leader
# get program paths
webapp = os.path.join(
os.path.abspath(os.path.dirname(os.path.realpath(__file__))),
'webapp.py'
)
exe = os.path.abspath(os.path.dirname(__file__) + '/../bin/py')
# set robot settings path
os.environ['SEARX_SETTINGS_PATH'] = os.path.abspath(
os.path.dirname(__file__) + '/settings_robot.yml')
# run the server
self.server = subprocess.Popen(
[exe, webapp, 'settings_robot'],
[exe, webapp],
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT
)
def tearDown(self):
# TERM all processes in my group
# send TERM signal to all processes in my group, to stop subprocesses
os.killpg(os.getpgid(self.server.pid), 15)
# remove previously set environment variable
del os.environ['SEARX_SETTINGS_PATH']
SEARXROBOTLAYER = SearxRobotLayer()
class SearxTestCase(TestCase):
"""Base test case for non-robot tests."""
layer = SearxTestLayer

View File

@ -5,10 +5,12 @@ import codecs
import cStringIO
import re
def gen_useragent():
# TODO
return "Mozilla/5.0 (X11; Linux x86_64; rv:26.0) Gecko/20100101 Firefox/26.0"
def highlight_content(content, query):
if not content:
@ -34,10 +36,11 @@ def highlight_content(content, query):
return content
class HTMLTextExtractor(HTMLParser):
def __init__(self):
HTMLParser.__init__(self)
self.result = [ ]
self.result = []
def handle_data(self, d):
self.result.append(d)
@ -54,6 +57,7 @@ class HTMLTextExtractor(HTMLParser):
def get_text(self):
return u''.join(self.result)
def html_to_text(html):
s = HTMLTextExtractor()
s.feed(html)

View File

@ -17,13 +17,7 @@ along with searx. If not, see < http://www.gnu.org/licenses/ >.
(C) 2013- by Adam Tauber, <asciimoo@gmail.com>
'''
import os
import sys
if __name__ == "__main__":
sys.path.append(os.path.realpath(os.path.dirname(os.path.realpath(__file__))+'/../'))
from searx import settings
from flask import Flask, request, render_template, url_for, Response, make_response, redirect
from searx.engines import search, categories, engines, get_engines_stats
import json
@ -33,11 +27,17 @@ from flask import send_from_directory
from searx.utils import highlight_content, html_to_text
import os
app = Flask(
__name__,
static_folder=os.path.join(os.path.dirname(__file__), 'static'),
template_folder=os.path.join(os.path.dirname(__file__), 'templates')
)
app = Flask(__name__)
app.secret_key = settings['server']['secret_key']
#TODO configurable via settings.yml
favicons = ['wikipedia', 'youtube', 'vimeo', 'soundcloud',
'twitter', 'stackoverflow', 'github']
@ -81,6 +81,7 @@ def render(template_name, **kwargs):
kwargs['selected_categories'] = ['general']
return render_template(template_name, **kwargs)
def parse_query(query):
query_engines = []
query_parts = query.split()
@ -94,7 +95,7 @@ def parse_query(query):
def index():
global categories
if request.method=='POST':
if request.method == 'POST':
request_data = request.form
else:
request_data = request.args
@ -106,7 +107,7 @@ def index():
query, selected_engines = parse_query(request_data['q'].encode('utf-8'))
if not len(selected_engines):
for pd_name,pd in request_data.items():
for pd_name, pd in request_data.items():
if pd_name.startswith('category_'):
category = pd_name[9:]
if not category in categories:
@ -159,22 +160,23 @@ def index():
response.headers.add('Content-Disposition', 'attachment;Filename=searx_-_{0}.csv'.format('_'.join(query.split())))
return response
elif request_data.get('format') == 'rss':
response_rss = render('opensearch_response_rss.xml'
,results=results
,q=request_data['q']
,number_of_results=len(results)
,base_url=get_base_url()
response_rss = render(
'opensearch_response_rss.xml',
results=results,
q=request_data['q'],
number_of_results=len(results),
base_url=get_base_url()
)
return Response(response_rss, mimetype='text/xml')
return render('results.html'
,results=results
,q=request_data['q']
,selected_categories=selected_categories
,number_of_results=len(results)+len(featured_results)
,featured_results=featured_results
,suggestions=suggestions
return render(
'results.html',
results=results,
q=request_data['q'],
selected_categories=selected_categories,
number_of_results=len(results) + len(featured_results),
featured_results=featured_results,
suggestions=suggestions
)
@ -192,9 +194,9 @@ def list_engines():
@app.route('/preferences', methods=['GET', 'POST'])
def preferences():
if request.method=='POST':
if request.method == 'POST':
selected_categories = []
for pd_name,pd in request.form.items():
for pd_name, pd in request.form.items():
if pd_name.startswith('category_'):
category = pd_name[9:]
if not category in categories:
@ -203,7 +205,10 @@ def preferences():
if selected_categories:
resp = make_response(redirect('/'))
# cookie max age: 4 weeks
resp.set_cookie('categories', ','.join(selected_categories), max_age=60*60*24*7*4)
resp.set_cookie(
'categories', ','.join(selected_categories),
max_age=60 * 60 * 24 * 7 * 4
)
return resp
return render('preferences.html')
@ -238,6 +243,7 @@ def opensearch():
mimetype="application/xml")
return resp
@app.route('/favicon.ico')
def favicon():
return send_from_directory(os.path.join(app.root_path, 'static/img'),
@ -248,9 +254,10 @@ def run():
from gevent import monkey
monkey.patch_all()
app.run(debug = settings['server']['debug']
,use_debugger = settings['server']['debug']
,port = settings['server']['port']
app.run(
debug=settings['server']['debug'],
use_debugger=settings['server']['debug'],
port=settings['server']['port']
)

View File

@ -49,4 +49,19 @@ setup(
'zope.testrunner',
]
},
entry_points={
'console_scripts': [
'searx-run = searx.webapp:run'
]
},
package_data={
'searx': [
'settings.yml',
'../README.md',
'static/*/*',
'templates/*.html',
'templates/result_templates/*.html',
],
},
)