Split up files with // ignore-tidy-filelength

[varkor]

These files are over 3,000 lines, which is not ideal for navigating or comprehending. If we can, it would be good to split these up. See also #60015, which is a specific issue of this problem.

T-compiler:

• rustc_parse/src/parser/expr.rs
• rustc_borrowck/src/diagnostics/conflict_errors.rs
• rustc_trait_selection/src/error_reporting/traits/suggestions.rs
• rustc_hir_typeck/src/method/suggest.rs
• rustc_resolve/src/late.rs
• rustc_resolve/src/late/diagnostics.rs

T-rustdoc:

• src/librustdoc/html/static/js/search.js

[petrochenkov]

#61097 is going to help with some of the files.

[olamiko]

Which files @varkor? The files in #60015 are done

[varkor]

I count 17 different files that still have this comment.

[alex-griffiths]

Hey, I'm not sure if anyone is working on this, but, this looks like it might be a good place to get started contributing.

I am quite new to the language, but I'd like to get involved. Is this something as simple as pulling out structs and other related blocks of code into their own files and then including them in the original?

[varkor]

@alex-griffiths: yes, it should be as simple as figuring out some sensible way to split up large files into related types and implementations!

[alex-griffiths]

@varkor Alright, mind if I take a crack at this?

[varkor]

@alex-griffiths: yes, go ahead! You can claim the issue with:

@rustbot claim

[alex-griffiths]

@rustbot claim

[alex-griffiths]

@varkor I'm finding that a lot of the files that include the // ignore-tidy-filelength comment are files that include lots of the documentation.

I'm still kind of new to the language, but for example in src/libstd/fs.rs there are 6 structs where the struct definition itself is a single line, but with varying amounts of docs. Is this the kind of thing that is able to be extracted into it's own file which is then included as a module in the fs.rs file?

[varkor]

@alex-griffiths: that's an interesting observation. I think in these cases, it's not so important to split up the files, because the length doesn't pose an issue for comprehension. Anyone editing the file is going to be interested in one or two particular functions, which are well-isolated within the file, and wouldn't benefit from being in their own file. In some sense, files like these are "false positives", though I don't think we could automatically exclude them. The lint is really more for files with large functions, and interdependency that we want to make easier to navigate. Thanks for pointing this out!

[alex-griffiths]

I'll start working through all the files that have the comment in them, and I'll compile a list of files that appear to mostly documentation, and then for the rest, I'll start trying to improve the navigability of them.

[danii]

Would like to help with this, too. Is there a possibility to share an issue?
Specifically, I have my eyes on core::num.

[alex-griffiths]

Yeah I'm happy with this @varkor.

Perhaps we could do something similar to how #75080 is managing the multiple people doing for that issue?

[jyn514]

Seems like a good idea. You need

• a list of the files to be split up
• instructions for how to split them up (more detailed is better, but simple is fine as long as you update it along the way)
• someone to keep the tracking issue up to date as people claim files. Maybe in this case that would be @varkor?

[danii]

Oh sweet, thanks! This'll be my first contributions to the rust standard library. 😄
I personally don't think we'll need instructions on splitting the files up as long as everyone's respectful. Although, it is @alex-griffiths decision.
If no one minds I'm gonna list out the files that need truncating.

[danii]

This is the list of all the files that need some splitage. ✅ #00000 means a split is done, 🔲 #00000 means it's in progress, and just 🔲 means it's free for the taking.

• refactor: removing library/alloc/src/vec/mod.rs ignore-tidy-filelength #78934 alloc::vec
• refactor: removing alloc::collections::vec_deque ignore-tidy-filelength #78590 alloc::collections::vec_deque
• Split core/num/mod.rs to smaller mods #76327 core::num
• Split up files with // ignore-tidy-filelength #75701 core::slice
• Split up core/test/iter.rs into multiple files #76391 core::tests::iter
• core::iter::traits::iterator
• compiletest::runtest tracked separatedly by Splitting up compiletest::runtest #89475
• Shorten rustc_middle::ty::mod #82964 rustc_middle::ty
• Split rustc_typeck::check into separate files #76906, Moved the main impl for FnCtxt to its own file. #77808 rustc_typeck::check::fn_ctxt
• Shorten html::render #82088 rustdoc::html::render

Out Of Scope

These files are very long, but they should be handled in a different manner.

• rustc_resolve per Split up files with // ignore-tidy-filelength #75701 (comment)
• src/test/ui/issues/issue-29466.rs per Refactor: reduce length of issue-72933 and issue-74564 test #82075 (comment)
• src/test/ui/issues/issue-72933-match-stack-overflow.rs per Refactor: reduce length of issue-72933 and issue-74564 test #82075 (comment)
• src/test/ui/issues/issue-74564-if-expr-stack-overflow.rs per Refactor: reduce length of issue-72933 and issue-74564 test #82075 (comment)
• rustc_apfloat::tests::ieee is code that should not really be altered currently per eddy's remarks in Split tests/ieee.rs into smaller files #93849

[alex-griffiths]

I believe changes to rustc_resolve need to be handled in a different manner.
#75701 (comment)

[tesuji]

I am doing core::slice splitting (#76311 )

[tesuji]

Hi @alex-griffiths, Would you mind if I take care of core::str part ?

[jyn514]

@lzutao maybe do core::num first since alex is in the middle of splitting core::str

[tesuji]

Yeah, he also did alloc/vec.rs in #75701 . But he seems having trouble in splitting core/str.rs.
I would like to help him a hand.

Edit: It's up as a draft PR #76325

[pierwill]

Looks like there are three left in compiler/:

• rustc_apfloat/tests/ieee.rs
• rustc_resolve/src/late/lifetimes.rs
• rustc_resolve/src/lib.rs

[dtolnay]

@rustbot release-assignment

[workingjubilee]

Further cleanup should focus on the files in rustc_resolve/src as the motivation for improving navigation in does not apply to the tests in question in rustc_apfloat, in effect.

[jmj0502]

Hey! I would like to start contributing to the project and this seems like a good starting point. Is the compiletest::runtest file still available?

[workingjubilee]

Yes, go ahead.

[Sam-Buckley]

I'd like to start contributing, and was directed here as a good starting point. What's still outstanding in this issue?
@rustbot claim

[workingjubilee]

@Sam-Buckley I updated the top of this issue with the current list: #60302 (comment)

[jieyouxu]

Updated issue description to not include compiletest::runtest, which is tracked separately in #89475.

[Sam-Buckley]

Sorry to waste everyone's time, family matters came up and I'll have to stop for a while

[Sam-Buckley]

@rustbot release-assignment

[ranger-ross]

Hey there, I'll take a crack at cleaning some of these large files up.

@rustbot claim

[NoobProgrammer31]

@rustbot claim

assign this to me i will work upon it

[NoobProgrammer31]

can i start working on mod.rs first in first file ( i am new to it )

[jieyouxu]

You don't need to finish all of them in one PR and please do not break all of them up in one PR. You can tackle each of the large files in a standalone PR.