Cover multiple import maps

[yoavweiss]

Description

This PR aims to cover the "multiple import maps" HTML change

Motivation

Current import map language is misleading, at least in ToT Chromium/WebKit.

Additional details

Related issues and pull requests

Fixes mdn/mdn#622

[github-actions]

Preview URLs

• /en-US/docs/Web/HTML/Element/script/type/importmap

Flaws (23)

URL: /en-US/docs/Web/HTML/Element/script/type/importmap
Title: <script type="importmap">
Flaw count: 23

• macros:
  • Wrong xref macro used (consider changing which macro you use). Error processing path /en-US/docs/Learn/HTML
  • Wrong xref macro used (consider changing which macro you use). Error processing path /en-US/docs/Learn/Getting_started_with_the_web/HTML_basics
  • Wrong xref macro used (consider changing which macro you use). Error processing path /en-US/docs/Learn/HTML/Introduction_to_HTML
  • Wrong xref macro used (consider changing which macro you use). Error processing path /en-US/docs/Learn/HTML/Introduction_to_HTML/Getting_started
  • Wrong xref macro used (consider changing which macro you use). Error processing path /en-US/docs/Learn/HTML/Introduction_to_HTML/The_head_metadata_in_HTML
  • and 18 more flaws omitted

External URLs (1)

URL: /en-US/docs/Web/HTML/Element/script/type/importmap
Title: <script type="importmap">

• https://github.com/guybedford/es-module-shims (1 time) (Note! This may be a new URL 👀)

(comment last updated: 2025-01-22 10:30:40)

[chrisdavidmills]

Hi @yoavweiss!

Thanks for the awesome work here in getting this documented. I have a few horrible nitpicky comments for you to consider (too much coffee, maybe), but I think this is going in the right direction.

[chrisdavidmills]

There seems to be no explanation here of how the browser handles multiple import maps. I can guess from the heading that they are merged in some way into the single global import map representation when you include them in a document, but it is not explicitly covered.

[yoavweiss]

I merged (pun not intended) this sentence and the next to try and make that clearer. Let me know if more details are needed.

[chrisdavidmills]

The last two sentences make sense, but I do have a few comments for you to consider:

• "Future resolution of these specifiers will provide the same results as their previous resolution." seems to contradict the "are dropped" assertion — it sounds like "are dropped" might be saying that that they are ignored, whereas the first quoted sentence sounds like they are evaluated but provide the same result. It would be good to be clearer about what's going on.
• "prevails" makes sense, but seems like a slightly weird word to use in a tech context. I had to read it a few times to figure out what you were saying, and I wonder how understandable it is to the majority of readers who have English as an additional language. You could say something like "previous mapping continues to take effect", or maybe even something simpler like "previous mapping wins"?
• I think this stuff would be easier to understand if you described a scenario where you have a first import map that contains mappings A and B, then a second map containing mappings A and C. When the first one is evaluated, A and B are added to the global map. When the second one is evaluated, C is added, but A is ignored because it was already previously added, etc. At the moment, you can guess that you are talking about specifiers that were already included in previous import maps, but it would be helpful to make this explicit.

[yoavweiss]

There's a series of examples after the processing model in https://html.spec.whatwg.org/#merge-module-specifier-maps. Do you think it would make sense to bring in some of those here to make the behavior clearer?