fix(http_requests): allow blob: to prepend URL string

[xgui3783]

This is my first PR in this repo. Apologies if I have made any mistakes.

Currently, if the data source originates from the web (thus have the URL of https?://), the parseUrl function correctly parses the protcol, host and path.

In our usecase, we often need to download the entire (nifti) image, preprocess them, and generate a blob URL. According to mdn doc on URL.createObjectURL , it creates a URL with blob: prepended. This fails the check as it is implemented currently.

This PR addresses this issue.

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.

[xgui3783]

I signed it.

(Looks like I have initiated the CLA, but not signed it until a few seconds ago)

[jbms]

Does this actually work, though? It looks like it just discards the "blob:" prefix when parsing.

[xgui3783]

The main issue is, without ^(blob:)? the function parseUrl throw an error, and does not get passed on.

Otherwise a nifti layer source such as nifti://blob:https://localhost:8080/{blob-id} works just fine.

[jbms]

The issue is that some code in Neuroglancer attempts to reassemble the URL from the result of calling parseUrl --- but you have modified parseUrl to just strip blob:, which means the URL would get reassembled as just http://... without the blob: prefix. The nifti datasource probably doesn't do that, but in any case I think it would be better to include blob: in the protocol, i.e. make blob:http the protocol. That should be possible with just a small change to your regular expression. If you can also add a test, e.g. a Python test that uses webdriver apis to construct a blob URL and then loads it, that would be great.