Use correct line offsets for doctests #47274
Conversation
13ed204
to
82e30be
Compare
|
this solution is currently incorrect, ignore it for now |
c90c134
to
fe33a42
Compare
fe33a42
to
0443b7f
Compare
|
Ready for review. r? @QuietMisdreavus |
|
Need to write a test, unsure where it should go. maketest perhaps? |
src/librustdoc/test.rs
Outdated
| } | ||
|
|
||
| // Now push any outer attributes from the example, assuming they | ||
| // are intended to be crate attributes. | ||
| prog.push_str(&crate_attrs); | ||
| line_offset += crate_attrs.lines().count(); |
There was a problem hiding this comment.
Is line_offset supposed to be the number of lines inserted by rustdoc? IIRC, crate_attrs is actually the #![] attributes from the example itself.
There was a problem hiding this comment.
Yeah, you're right. Given that the attrs can only be up top we should just not count this.
|
Nice! If i'm looking at this correctly, this will help map between the line that rustc reports an error at, and the line in the actual doctest that was responsible for it? Do we already map these lines back to the file it's in? I'm thinking about this in terms of getting the lines to work with Re: where/how to test this, i'd take a look at whether/where the regular line number reporting is tested, and seeing if you can piggyback onto that by making it call |
yeah.
We say "foobar.rs - itemname (line N)". But then the rustc error shows |
|
Added test. The existing stuff isn't easy to rejigger; it's built around assuming rustc. I just wrote a simple run-make test. |
0443b7f
to
44b659a
Compare
|
Seems good to me. Let's get this in! @bors r+ |
|
📌 Commit 44b659a has been approved by |
Use correct line offsets for doctests Not yet tested. This doesn't handle char positions. It could if I collected a map of char offsets and lines, but this is a bit more work and requires hooking into the parser much more (unsure if it's possible). r? @QuietMisdreavus (fixes #45868)
|
☀️ Test successful - status-appveyor, status-travis |
Not yet tested.
This doesn't handle char positions. It could if I collected a map of char offsets and lines, but this is a bit more work and requires hooking into the parser much more (unsure if it's possible).
r? @QuietMisdreavus
(fixes #45868)