Workflow for more granular Web Search?

[Codewyn]

Not sure if this exists or if the question has already been posed in the forum, but I am looking for a workflow that extends the functionality of the Web Search feature to make it more granular. Perhaps it's not a question of a workflow being developed but of the base feature being extended, but I want to pose the idea either way...

 

So the current Web Search gives you one variable: {query}. This only enables you to replace the location of the variable in the URL with the literal string you search. This has limits, though. Many sites don't have a clear structure for this. For instance, if I am searching for a Bible verse, if I use BibleGateway.com I can easily replace the full string...

 

http://www.biblegateway.com/passage/?search={query}&version=NASB

Searching for "John 8:32" yields a valid URL...

http://www.biblegateway.com/passage/?search=John%208%3A32&version=NASB

 

However, if I search another similar site - BibleHub.com - I am not able to replace it the same. Browsing to the same verse yields the following URL structure...

 

http://biblehub.com/john/8-32.htm

 

{query} would not be sufficient since there are separators. So if I were to search for "John 8:32" at the same site, I would need three separate variables to translate into the URL: "John" being {query1}, "8" being {query2} and "32" being {query3}, for the following search URL...

 

http://biblehub.com/{query1}/{query2}-{query3}.htm

For it to yield...

http://biblehub.com/john/8-32.htm

 

What's more, I would need to define variables for the search string as well, so it interprets based on placement and ignores certain characters (i.e.: the colon ":"). So a separate field for defining the search string variables would look something like this...

 

"{searchString1} {searchString2}:{searchString3}"

 

Then the placement of the user input is interpreted by the searchString# and translated to the corresponding query#.

 

I apologize if any of this is confusing. If so, I would be happy to elabote further.

 

Does anyone know if a similar solution exists, or is in the works, or can otherwise be incorporated into Alfred?

[deanishe]

To do that, you'd need to write your own workflow that understands how to parse your input (john 8:32) into the component parts and assemble them into a URL. Should be pretty simple to do.

 

Alternatively, you could create a custom web search using Google to search the site with its "I'm feeling lucky" function:

 

http://www.google.com/search?q=site:biblehub.com+john+8:32&btnI

http://www.google.com/search?q=site:biblehub.com+john+8:32&btnI

which will take you straight to http://biblehub.com/john/8-32.htm

Edited February 21, 2014 by deanishe

[rice.shawn]

If you're just looking for Bible verses and not necessarily comparative verses between websites, then you can use this workflow.