ArXiv Wrapper

This wrapper endpoint is provided for free as is and under no circumstances do I bear any responsibility for the consequences of you using it. Please email any questions or comments to h e d e e r (at) e l s h o w k . c o m.

The purpose of this wrapper is to expose arXiv's API (see documentation here) with JSON return objects instead of the default XML.

Usage is straightforward,

http://arxivwrap.org/wrapper?method=[json|xml]&args

Where args correspond to the arXiv API arguments.

The XML to JSON conversion follows these rules:

A variable "entries_count" has the count of the number of entries,
Every entry is an object inside the "entries" array,
Every tag underneath the entry has a "value" and "attribs" variable,
Nested tags show up as arrays inside the "value" tag (only the author tag has children),
If multiple tags exist at the same level (e.g. link and category objects) they're also listed as an array of objects (each with a "value" and "attribs"). The only difference is that they are flat such as {"category": [ {"value":..., "attribs": ..}, ... ] } instead of the nested objects under author which show up as an array of { "name": { "value": ..., "attribs": ... } } objects. You can test out the search below to see what this looks like.
The link,title, updated, and id tags at the root are each objects with their attribs listed as variables and their strings under "value" (e.g. href is a variable inside link)

Here's a sample call,

http://arxivwrap.org/wrapper?method=json&search_query=au:bengio&max_results=10

If you notice any odd behavior or incorrect results please don't hesitate to contact me.

The input box below will show sample results for the given search query with max_results=2. White spaces are replaced with a '+' in the query, feel free to use boolean operators. Sample search string (copy and paste this into the search box), au:bengio ANDNOT ti:deep learning