Strings for token parsing

[pwalczysko]

@will-moore suggested following strategy: Allow users to select the strings on which to parse the tokens in the names, e.g. slashes, underscores and more via an input box in the top bar in central pane. This would get away with confusions like German-BioImaging/webtagging#37 (comment)

[dpwrussell]

Part of the reason for the client side rewrite is to add behaviour like this, so I'm definitely going to add it. I guess I might as well do so at this time.

How do you envision this being offered to the user? Options:

• Dropdown multi-select list
• Free text box regex
• Free text box (escaping must be done to convert to regex for splitting)

[will-moore]

I would imagine a simple text box where the user could enter any characters they want to split with. By default it would contain E.g. "/._" and they could then remove these and/or add their own. Each of the single characters they enter would be used to split (don't try to split by 2 or more characters). Not sure how you handle splitting by whitespace, but maybe if they enter a space then just use that? Don't expect users to know regex - even devs find it hard enough.

[dpwrussell]

This is actually a bit more complicated than I thought. I will leave this open and I will get to it, but for now I have to move onto other priorities.