Macro flaws included for recursively sub-renders in other pages

If you open http://localhost:3000/en-US/docs/Web/HTML/Element/img#_flaws you’ll see that it complains about a MacroRedirectedLinkError on line 1212. But that flaw is actually not in this page. This page (en-us/web/html/element/img/index.html) only has 417 lines! So, this flaw shouldn’t even be mentioned when you press the “Show flaws” button.

The actual macro flaw is actually in http://localhost:3000/en-US/docs/Web/Media/Formats/Image_types#_flaws There, it should appear when you press the “Show flaws” button.

[escattone]

@peterbe Here are my initial thoughts. What you’re describing, knowing which macro in the original document’s raw HTML forced the rendering of other documents which then had errors, would be a nice thing to have. The problem is that it gets very complicated depending on how detailed we want to get. For instance, in the example you cited (http://localhost:3000/en-us/docs/web/html/element/input/number), there are 3 places in the raw HTML that call the page macro asking for various sections of http://localhost:3000/en-us/docs/web/html/element/input/text:

line 100 --> {{page("/en-US/docs/Web/HTML/Element/input/text", "list", 0, 1, 2)}}
line 114 --> {{page("/en-US/docs/Web/HTML/Element/input/text", "placeholder", 0, 1, 2)}}
line 116 --> {{page("/en-US/docs/Web/HTML/Element/input/text", "readonly", 0, 1, 2)}}

so it’s impossible to point to a single line in the raw HTML of /en-us/docs/web/html/element/input/number that introduces the rendering of /en-us/docs/web/html/element/input/text and its associated MacroBrokenLinkError’s. Any one of those 3 lines, those page macro calls, would introduce the prerequisite. Also, since the page macro is typically used to extract only a specific section of a prerequisite document, it’s difficult to know if the extracted section actually includes any of the non-fatal errors (like MacroBrokenLinkError) that may have occurred while rendering that prerequisite.

Because of this complexity and all of the other work we have to do, I suspect it’s not worth getting bogged down in these details to determine an elegant approach at the moment, and maybe for now we should go with some kind of quickfix?

[peterbe]

Pretty sure it’s no longer a problem. We don’t get crazy recursion any more and we have a solution to the “Show flaws” UI to alert when you’re seeing a macro flaw coming from a prerequisite.