Lints to ensure link text for EIPs should match the EIP's number

[SamWilsn]

Let's create two lints:

markdown-link-eip

If the link destination matches the pattern eip-([^.]*)\.md$, trigger if the link text does not match ^(EIP|ERC)-\1$, where \1 is the proposal's number from the file name (aka capture group 1.)

For example:

• [EIP-1](./eip-2.md)
• [EIP-1: Foo](./eip-1.md)
• [Another Proposal](./eip-1.md)

If the link destination matches the pattern eip-([^.]*)\.md#.+$, trigger if the link text does not match ^(EIP|ERC)-\1\s*\S+, where \1 is the proposal's number from the file name (aka capture group 1.) Note that this pattern allows trailing characters.

For example:

• [EIP-1](./eip-1.md#motivation)
• [EIP-2: Hello](./eip-1.md#abstract)
• [Another Proposal](./eip-1.md#rationale)

markdown-link-other

If the link text begins with EIP or ERC, the link target must be an EIP. For example:

• [EIP-2](../assets/eip-2/foo.txt)

[Pandapip1]

Note: if it links to a section, then this lint should be ignored, such as [EIP-721's Motivation](./eip-721.md#motivation).

[SamWilsn]

That's a good point!

[swetshaw]

Hi, I would like take this up with Katya

[KatyaRyazantseva]

Hi @SamWilsn, yep, we both would like to contribute into it. Please, assign us to this issue

[KatyaRyazantseva]

Hi @SamWilsn, taking into account @Pandapip1 's comment about sections, should all these links matching the pattern eip-([^.]*)\.md#.+$ be ignored? Those ones having the section in the link, the second part of markdown-link-eip part:

• [EIP-1](./eip-1.md#motivation)
• [EIP-2: Hello](./eip-1.md#abstract)
• [Another Proposal](./eip-1.md#rationale)

[SamWilsn]

This paragraph covers those examples:

If the link destination matches the pattern eip-([^.]*)\.md#.+$, trigger if the link text does not match ^(EIP|ERC)-\1\s*\S+, where \1 is the proposal's number from the file name (aka capture group 1.) Note that this pattern allows trailing characters.

So:

• [EIP-1](./eip-1.md#motivation) would not be allowed because it doesn't have anything in the \S+ position of the regex.
• [EIP-2: Hello](./eip-1.md#abstract) would not be allowed because it doesn't match EIP number.
• [Another Proposal](./eip-1.md#rationale) would not be allowed because it doesn't begin with either ERC or EIP.

However, [EIP-1: Motivation](./eip-1.md#motivation) would be allowed.

[KatyaRyazantseva]

@SamWilsn could you please explain why [EIP-1: Foo](./eip-1.md) is incorrect?

[SamWilsn]

Because it doesn't have a subsection (after the # in the link.)

[KatyaRyazantseva]

if [EIP-N] is extended with extra text like ': Foo' it means the link should have a section?

[SamWilsn]

Pretty much. So URLs with a # should have : ...

[KatyaRyazantseva]

@SamWilsn, for markdown-link-other should I check eip number match in the text and the link?

[SamWilsn]

markdown-link-other should prevent links that look like EIPs/ERCs that don't point to EIPs.