Olf0
commented
1 week ago Thank you for denoting this, but as handling multi-image comic strips is not an advertised feature, IMO this constitutes a feature request, not a bug report.
Conceptionally, on first sight I do not believe there is a generic way to detect which images on a web-page belong to the same comic strip.
Hence (just brainstorming) this may become addressable when extraction scripts (extract.js) are allowed to return a list of image-URLs. This scheme may be expanded to also allow for returning text, e.g. as first or last element of this list (which would have to be empty if no text is returned, hence requiring to trivially adapt all extant extraction scripts); another approach for this would be to restrict the image-URLs returned to full URLs starting with https:// or http:// (every other element is treated as plain text, but at most a single list element), as IIRC currently also URLs relative to the base-URL used for extraction are supported (but currently no extraction script utilises this, IIRC). Either approach may then be utilised to address your feature suggestion to allow for returning text, which you still may (or may not) file as a proper issue report. :wink:
Edit: Oh, returning multiple values is already filed as issue #53, which suggests yet another approach.
As stated, I am not at all good at writing JavaScript, QML, C++, Python etc. code, hence I will not implement any of this, but gladly review a PR addressing this and ask checks&balances questions (as I feel it is my duty as the currently active maintainer of this piece of software).
ExTechOp
SailfishOS VERSION (Settings → About product → Build):
4.6.0.15HARDWARE (Settings → About product → Manufacturer & Product name): Sony, Xperia 10 III
Daily Comics VERSION (Daily Comics → [Top pulley] About):
1.9.95BUG DESCRIPTION
pbfcomics (The Perry Bible Fellowship Comics) is occasionally published as split into multiple separate images, which would need to be separately recognized and loaded
STEPS TO REPRODUCE