Exposes a basic API to search using keywordtool.io. That search uses autosuggest from various search engines to complete a base phrase.
Navigate to any directory using your root/system conda and run the following commands:
git clone https://github.com/quantum-programmer/keywordtool-scraper.git
conda build keywordtool-scraper -c conda-forge -c r
(OPTIONAL) Remove unnecessary build files with
conda build purge
(OPTIONAL) Activate the environment in which you wish to install keywordtool-scraper
:
source activate my_environment
To actually install the package, run
conda install keywordtool-scraper -c conda-forge -c r --use-local
Note: If I ever add support for a modern alternative to PhantomJS (like Headless Chrome), I'll add this to the Anaconda Cloud. Feel free to contribute/fork as desired.
The interface is a KeywordTool
object with several helper methods/properties.
import keywordtool
kt = keywordtool.KeywordTool(source='google', timeout=5) # kwargs are optional
keywords = kt.get_keywords('florida man arrested')
The usage doesn't get too crazy. What follows is a more-or-less comprehensive demonstration of the capabilities. Feel free to run and experiment with the code to get a feel for how it works.
import keywordtool
kt = keywordtool.KeywordTool(source='google', timeout=5) # kwargs are optional
print 'Current URL: %s\n' % kt.current_url
print 'There are %d search engines to choose from:' % len(kt.sources)
for s in kt.sources:
print ' %s' % s
kt.timeout = 10
kt.source = 'bing'
arrested_keywords = kt.get_keywords('florida man arrested')
print '\nHeadlines from Florida include "%s"' % arrested_keywords[0].upper()
book_keywords = kt.get_keywords('good books', source='amazon')
print '\nPeople on Amazon are searching for "%s"' % book_keywords[0]
(IMPORTANT) The source
kwarg silently defaults to 'google'
if the provided argument is invalid. Available options for the source
kwarg are
'amazon'
'app-store'
'bing'
'ebay'
'google'
'youtube'