Is the SUMMARY.md parser too flexible?

[szabgab]

Question

I was trying to create a test for the Summary parsing failed error in book.rs, but so far it parsed everything.
Sometimes not as I expected, but I could not create a SUMMARY.md file that would fail to parse.

Here is one thing that seems to be odd. These example all generate the same thing (two prefix chapters)

Both starting on the 1st column

[A](./a.md)
[B](./b.md)

B Starting indented (the

[A](./a.md)
  [B](./b.md)

[A](./a.md) [B](./b.md)

Even when there are 1-3 leading spaces

  [A](./a.md) [B](./b.md)

However, when there are 4 leading spaces, then these entries are disregarded

    [A](./a.md) [B](./b.md)

This is also accepted:

[A](./a.md)
- [B](./b.md)
[C](./c.md)
- [D](./d.md)
[E](./e.md)

My understanding of the docs is that you cannot have a prefix (like) chapter after the first numbered chapter.

The result is

A
1. B
2. D

without error.

IMHO mdbook should have stricter rules and it should report errors instead of just disregarding things.

Version

0.4.46