chore(deps): update d3* and @types/d3*, and override d3-sankey and @types/d3-sankey

[Yokozuna59]

📑 Summary

• update d3 version to 7.8.5 exclude the 7.4.0 version.
• override d3-sankey and @types/d3-sankey dependencies versions.
• move @types/d3-scale and @types/d3-scale-chromatic to devDependencies.
• update all d3 related dependencies.

Resolves #4872

📏 Design Decisions

We were waiting for d3/d3-sankey#117 to be merged to resolve the d3-sankey issues, but it appears it won't be anytime soon. So we're taking the suggested workaround here for now to exclude older versions.

As far as I can tell, taking such an approach shouldn't affect the layout or types of d3-sankey, and renovate should also be able to update overriden versions in package.json.

📋 Tasks

Make sure you

• 📖 have read the contribution guidelines
• 🔖 targeted develop branch

[netlify]

✅ Deploy Preview for mermaid-js ready!

Name	Link
🔨 Latest commit	b63fdd9
🔍 Latest deploy log	https://app.netlify.com/sites/mermaid-js/deploys/65c4a979e729c80007e9d3a5
😎 Deploy Preview	https://deploy-preview-4880--mermaid-js.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify site configuration.

[Yokozuna59]

I don't know how to see the size difference after the commit, but according to this comment, it should be less by ~50Kb.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 85.09%. Comparing base (9c6ad54) to head (b63fdd9).
Report is 2189 commits behind head on develop.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #4880      +/-   ##
===========================================
+ Coverage    79.51%   85.09%   +5.57%     
===========================================
  Files          175      174       -1     
  Lines        14396    11666    -2730     
  Branches       855      844      -11     
===========================================
- Hits         11447     9927    -1520     
+ Misses        2755     1543    -1212     
- Partials       194      196       +2     

Flag	Coverage Δ
e2e	85.09% <ø> (ø)
unit	?

Flags with carried forward coverage won't be shown. Click here to find out more.

see 23 files with indirect coverage changes

[aloisklink]

Hi @Yokozuna59,

Two questions:

• Why have we made this change: * update `d3` version to `7.8.5` exclude the `7.4.0` version. Is it just to keep things updated? We might want to make a separate PR for this, just so there's a separate entry in the release notes for it.
• Unit tests are failing due to:

    packages/mermaid/src/mermaid.spec.ts > when using mermaid and  > when using #registerExternalDiagrams > should throw error (but still render) if registerExternalDiagrams fails
AssertionError: expected [Function] to throw error including 'Failed to load 1 external diagrams' but got 'Failed to load 2 external diagrams'

- Expected
+ Received

- Failed to load 1 external diagrams
+ Failed to load 2 external diagrams

 ❯ packages/mermaid/src/mermaid.spec.ts:72:7
     70|       node.appendChild(document.createTextNode('graph TD;\na;'));
     71| 
     72|       await expect(
       |       ^
     73|         mermaid.registerExternalDiagrams(
     74|           [

The sankey e2e tests seem to be fine, though, so I'm not sure if that's the issue.

By the way, this is super minor and it's probably too much effort to do this in this PR, but in the future, making a separate commit for each change would be more ideal, e.g. something like:

• chore(deps): update `d3` version to `^7.8.5`
• chore(deps): override `d3-sankey` and `@types/d3-sankey` version
• chore(deps): move `@types/d3-scale` and `@types/d3-scale-chromatic` to devDependencies
• chore(deps): update all `d3` related dependencies

[Yokozuna59]

Hi @aloisklink,

• Why have we made this change: * update `d3` version to `7.8.5` exclude the `7.4.0` version. Is it just to keep things updated? We might want to make a separate PR for this, just so there's a separate entry in the release notes for it.

I'd say it's not, in package.json of the previous commit, it uses [email protected], but in the pnpm-lock.yaml it uses kinda uses both [email protected] and [email protected]:

mermaid/pnpm-lock.yaml

Lines 212 to 214 in 2049915

	d3:
	specifier: ^7.4.0
	version: 7.8.2

So I updated both versions in packages/mermaid/package.json and packages/mermaid-example-diagram/package.json to resolve the issue.

• Unit tests are failing due to:

The sankey e2e tests seem to be fine, though, so I'm not sure if that's the issue.

Same here. I wasn't sure what's the issue so I created a PR until I figure it out.

By the way, this is super minor and it's probably too much effort to do this in this PR, but in the future, making a separate commit for each change would be more ideal, e.g. something like:

I planed to do so, but then I thought that it would has unnecessary commits. And the pnpm-lock.yaml is already updated and it would need more effort to partially git add lines for each commit.

[aloisklink]

By the way, this is super minor and it's probably too much effort to do this in this PR, but in the future, making a separate commit for each change would be more ideal, e.g. something like:

I planed to do so, but then I thought that it would has unnecessary commits. And the pnpm-lock.yaml is already updated and it would need more effort to partially git add lines for each commit.

Sounds good! Having each separate change in a separate commit would make it easier to find which change caused the unit tests to fail (e.g. using something like git bisect, but that's up to you.

I'll set this PR as a draft until we figure out what's happening with the unit tests. Other than that, this PR looks good!

[Yokozuna59]

Sounds good! Having each separate change in a separate commit would make it easier to find which change caused the unit tests to fail (e.g. using something like git bisect, but that's up to you.

It's because of sankey dependencies overriding. I've removed them locally and tested it out, and it worked fine.
The weird part is that e2e tests are passing.