suggest declaring modules when file found but module not defined

[bend-n]

suggests declaring modules when a module is found but not defined, i.e

├── main.rs: `use thing::thang;`
└── thing.rs: `struct thang`

or

├── main.rs: `use thing::thang;`
└── thing
    └── mod.rs: `struct thang`

which currently is just

error[E0432]: unresolved import `yeah`
 --> src/main.rs:1:1
  |
1 | use thing::thang;
  |     ^^^^^ use of unresolved module or unlinked crate `thing`
  |

but now would have this nice help:

= help: you may have forgotten to declare the module `thing`. use `mod thing` in this file to declare this module.

[bend-n]

@rustbot label A-diagnostics

[rustbot]

r? @petrochenkov

rustbot has assigned @petrochenkov.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[petrochenkov]

Could you add a comment explaining what are you trying to do here?
(Also format this on a separate line to improve readability let sp = map.span_extend_to_prev_char(ident.span, '\n', true);)

[petrochenkov]

Ah, wait, I think this can actually use current_crate_outer_attr_insert_span instead of UsePlacementFinder.
(The newly added comment tells that the current logic is clearly incorrect.)

[bend-n]

oh yeah that seems like what i wanted i was wondering what the extern crate thing was spawned from

[petrochenkov]

Could you also add a new test exercising the various naming schemes (like …/x/y/mod.rs and others) for the module file?

[bend-n]

Could you also add a new test exercising the various naming schemes (like …/x/y/mod.rs and others) for the module file?

im not sure how to make a test with a normal directory structure? everything seems to use #[path = auxiliary/]...

[petrochenkov]

Well, tests/ui/modules_and_files_visibility/mod_file_disambig.rs caught some file to report, the new test can probably do something similar to that case.