Base publication slug (URL) on title rather BibTeX ID
gcushen opened this issue · comments
Currently, we use the ID provided by Bibtex. An ID based on the title may be more intuitive but this kind of ID is unavailable in Bibtex, so would require implementing the below algorithm to generate a unique ID based on the title.
Requirements:
- consider up to first 5 words of title (after removal of stop words?)
- strips special chars
- converts all characters to lowercaps
- slugify: replace whitespaces, underscores and periods by hyphens/dashes
- reduce multiple consecutive dashes to one
- Either check generated ID is unique and append numeric ID if clashes, or else append hash to name
Are you thinking of the folders being names as such, or simply adding the "slug" parameter to specify this URL? (Mainly asking because most researchers I've been around mentally catalog <first-author-last-name> [et al], <year> – so ensuring folders are named along those lines, or similar, is probably more helpful for them?)
@jmuchovej an extract of the title followed by a unique ID/hash might be more meaningful for SEO and users searching Google than just a surname and a date in the URL?
Otherwise, for the researcher, it might be more convenient if the folders are named with their publish date first (e.g. 2022-01-01-name-hash) so they are ordered within their file browser...And then concatenate something like first author surname or title extract followed by a unique hash, as in that example?
Or if everyone is happy with the current Bibtex ID naming of folders, perhaps we can just close this...
I am pretty happy with the current naming scheme
@jmuchovej an extract of the title followed by a unique ID/hash might be more meaningful for SEO and users searching Google than just a surname and a date in the URL?
Otherwise, for the researcher, it might be more convenient if the folders are named with their publish date first (e.g.
2022-01-01-name-hash) so they are ordered within their file browser...And then concatenate something like first author surname or title extract followed by a unique hash, as in that example?Or if everyone is happy with the current Bibtex ID naming of folders, perhaps we can just close this...
Hi,
I'm getting to know this software now and I'm pretty amazed by it. Is this issue still relevant? I may be able to help with this feature. From my perspective, it might be interesting to allow the user to select which option to use, keeping the current style as the standard one.
Something like this:
academic --name-style ????
What do you think?
When I used acadamic-cli a few years ago (which is a fantastic help by the way), the folders, and in turn the urls, looked like:
Now they look like:
I just want to confirm - is that second screenshot what I should expect the BibtexID naming scheme to look like?