The link checker at present warns about every instance of data URIs in every page, eg
info Line: 1963 data:image/gif;base64,R0lGODlhBAAEAPAAMQAAAIiIiCwAAAAABAAEAAACBkwAhqgZBQA7
Status: (N/A) Access to 'data' URIs has been disabled
Accessing links with this URI scheme has been disabled in link checker.
This adds a lot of clutter to my reports (I have, for various CSS & dark mode reasons, 6 data URIs inlined on every page on my site, and each one generates a useless warning taking up 10 vertical lines each collectively taking up a good third of the report for most pages).
I've looked at the docs & bug reports, and I can't see any discussion of why link-checker might hypothetically check data URIs, such that the user needs to be warned that it did not in fact check them. Is there some way in which data URIs can be "broken links" within the scope of link-checker? A data URI is pretty much defined as not being a hyperlink elsewhere but literally encoding, inline, a file which could've been hyperlinked elsewhere but is not. So it seems like any 'broken' data URIs would either be broken HTML (and so checked by the other HTML validation tool instead), or would simply encode a bad data payload such as invalid SVGs (which, while possibly useful, is way beyond a link checker's scope given how many things are encoded into arbitrary data URIs).
If there is no plausible use-case, perhaps the warnings can be omitted entirely? Or if there are some use-cases, maybe a toggle for reporting them (defaulting to off or on depending on how common said use-cases are). Any of those would make it easier to read the more meaningful warnings & errors.
The link checker at present warns about every instance of data URIs in every page, eg
This adds a lot of clutter to my reports (I have, for various CSS & dark mode reasons, 6 data URIs inlined on every page on my site, and each one generates a useless warning taking up 10 vertical lines each collectively taking up a good third of the report for most pages).
I've looked at the docs & bug reports, and I can't see any discussion of why link-checker might hypothetically check data URIs, such that the user needs to be warned that it did not in fact check them. Is there some way in which data URIs can be "broken links" within the scope of link-checker? A data URI is pretty much defined as not being a hyperlink elsewhere but literally encoding, inline, a file which could've been hyperlinked elsewhere but is not. So it seems like any 'broken' data URIs would either be broken HTML (and so checked by the other HTML validation tool instead), or would simply encode a bad data payload such as invalid SVGs (which, while possibly useful, is way beyond a link checker's scope given how many things are encoded into arbitrary data URIs).
If there is no plausible use-case, perhaps the warnings can be omitted entirely? Or if there are some use-cases, maybe a toggle for reporting them (defaulting to off or on depending on how common said use-cases are). Any of those would make it easier to read the more meaningful warnings & errors.