> I had to add a literal? field to the literal blocks to distinguish them
from regular blocks and prevent the normalizer from recursing infinitely.
You can always traverse manually. The number of AST nodes are small and it
will give you more control. Alternatively, give postwalk a try, since the
Hi there! In the projects I'm working on some dependencies are fetched
directly from private git. First time latest master is fetched and then its
sha is locked in mix.lock. No tags, no branches are used (sure, might be
not the best practice). Lately I was moving "private" libraries from GitHub
I'm currently experimenting on the second point you outlined
I've uploaded what I've been trying so far to this
repository: https://github.com/doorgan/formatter. The relevant file
is https://github.com/doorgan/formatter/blob/main/lib/normalizer.ex as well
as the tests.
The approach I'm taking
I think there are two fronts:
1. Today, all of the API the formatter uses, except the extraction of code
comments, is public. So providing a string_to_quoted_with_comments would be
a logical next step. We only need to convert the comments from a tuple to a
map before.
2. The second step is more
Hi folks,
Nice to see someone got the ball rolling, I'm also interested (and would
like to contribute). For me it's a bit unclear what the next steps would
be?
If someone could clarify, it'd be happy to spend some time on PR
Best,
Tonći
On Friday, May 7, 2021 at 6:15:48 AM UTC+2 José Valim
