Skip to content

Latest commit

 

History

History
187 lines (148 loc) · 8.26 KB

CONTRIBUTING.md

File metadata and controls

187 lines (148 loc) · 8.26 KB

Contributing to Neovim

Getting started

If you want to help but don't know where to start, here are some low-risk/isolated tasks:

Developer guidelines

  • Nvim contributors should read :help dev (especially :help dev-api).
  • External UI developers should read :help dev-ui.
  • API client developers should read :help dev-api-client.
  • Nvim developers are strongly encouraged to install ninja for faster builds.
    sudo apt-get install ninja-build
    make distclean
    make  # Nvim build system uses ninja automatically, if available.
    

Reporting problems

  • Check the FAQ.
  • Search existing issues (including closed!)
  • Update Neovim to the latest version to see if your problem persists.
  • Disable plugins incrementally, to narrow down the cause of the issue.
  • When reporting a crash, include a stacktrace.
  • Bisect to the cause of a regression, if you are able. This is extremely helpful.
  • Check $NVIM_LOG_FILE, if it exists.
  • Include cmake --system-information for build issues.

Pull requests ("PRs")

  • To avoid duplicate work, create a [WIP] pull request as soon as possible.
  • Avoid cosmetic changes to unrelated files in the same commit.
  • Use a feature branch instead of the master branch.
  • Use a rebase workflow for small PRs.
    • After addressing review comments, it's fine to rebase and force-push.
  • Use a merge workflow for big, high-risk PRs.
    • Merge master into your PR when there are conflicts or when master introduces breaking changes.
    • Use the ri git alias:
      [alias]
      ri = "!sh -c 't=\"${1:-master}\"; s=\"${2:-HEAD}\"; mb=\"$(git merge-base \"$t\" \"$s\")\"; if test \"x$mb\" = x ; then o=\"$t\"; else lm=\"$(git log -n1 --merges \"$t..$s\" --pretty=%H)\"; if test \"x$lm\" = x ; then o=\"$mb\"; else o=\"$lm\"; fi; fi; test $# -gt 0 && shift; test $# -gt 0 && shift; git rebase --interactive \"$o\" \"$@\"'"
      
      This avoids unnecessary rebases yet still allows you to combine related commits, separate monolithic commits, etc.
    • Do not edit commits that come before the merge commit.
  • During a squash/fixup, use exec make -C build unittest between each pick/edit/reword.

Stages: WIP, RFC, RDY

Pull requests have three stages: [WIP] (Work In Progress), [RFC] (Request For Comment) and [RDY] (Ready).

  • Untagged PRs are assumed to be [RFC], i.e. you are requesting a review.
  • Prepend [WIP] to the PR title if you are not requesting feedback and the work is still in flux.
  • Prepend [RDY] to the PR title if you are done with the PR and are only waiting on it to be merged.

For example, a typical workflow is:

  1. You open a [WIP] PR where the work is not ready for feedback, you just want to let others know what you are doing.
  2. Once the PR is ready for review, you replace [WIP] in the title with [RFC]. You may add fix up commits to address issues that come up during review.
  3. Once the PR is ready for merging, you rebase/squash your work appropriately and then replace [RFC] in the title with [RDY].

Commit messages

Follow commit message hygiene to make reviews easier and to make the VCS/git logs more valuable.

  • Try to keep the first line under 72 characters.
  • Prefix the commit subject with a scope: doc:, test:, foo.c:, runtime:, ...
    • For commits that contain only style/lint changes, a single-word subject line is preferred: style or lint.
  • A blank line must separate the subject from the description.
  • Use the imperative voice: "Fix bug" rather than "Fixed bug" or "Fixes bug."

Automated builds (CI)

Each pull request must pass the automated builds on travis CI, quickbuild and AppVeyor.

  • CI builds are compiled with -Werror, so compiler warnings will fail the build.
  • If any tests fail, the build will fail. See test/README.md#running-tests to run tests locally. Passing locally doesn't guarantee passing the CI build, because of the different compilers and platforms tested against.
  • CI runs ASan and other analyzers.
    • To run valgrind locally: VALGRIND=1 make test
    • To run Clang ASan/UBSan locally: CC=clang make CMAKE_FLAGS="-DCLANG_ASAN_UBSAN=ON"
  • The lint build (#3174) checks modified lines and their immediate neighbors. This is to encourage incrementally updating the legacy style to meet our style guidelines.
    • A single word (lint or style) is sufficient as the subject line of a commit that contains only style changes.
  • How to investigate QuickBuild failures

QuickBuild uses this invocation:

mkdir -p build/${params.get("buildType")} \
&& cd build/${params.get("buildType")} \
&& cmake -G "Unix Makefiles" -DBUSTED_OUTPUT_TYPE=TAP -DCMAKE_BUILD_TYPE=${params.get("buildType")}
-DTRAVIS_CI_BUILD=ON ../.. && ${node.getAttribute("make", "make")}
VERBOSE=1 nvim unittest-prereqs functionaltest-prereqs

Clang scan-build

The auto-generated clang-scan report presents walk-throughs of bugs found by Clang's scan-build static analyzer. To verify a fix locally, run scan-build like this:

rm -rf build/
scan-build --use-analyzer=/usr/bin/clang make

Coverity

Coverity runs against the master build. To view the defects, just request access; you will be approved.

Use this commit-message format for coverity fixes:

coverity/<id>: <description of what fixed the defect>

where <id> is the Coverity ID (CID). For example see #804.

PVS-Studio

View the PVS analysis report to see bugs found by PVS Studio. You can run scripts/pvscheck.sh locally to run PVS on your machine.

Reviewing

To help review pull requests, start with this checklist.

Reviewing can be done on GitHub, but you may find it easier to do locally. Using hub, you can create a new branch with the contents of a pull request, e.g. #1820:

hub checkout https://github.com/neovim/neovim/pull/1820

Use git log -p master..FETCH_HEAD to list all commits in the feature branch which aren't in the master branch; -p shows each commit's diff. To show the whole surrounding function of a change as context, use the -W argument as well.