add support for BUN_CONFIG_ELIDE_LINES (& fixes #16286) by martinamps · Pull Request #18111 · oven-sh/bun

martinamps · 2025-03-12T06:15:48Z

Summary

Adds an BUN_CONFIG_ELIDE_LINES env var per this comment and fixes #16286

Repro

Script:

mkdir -p /tmp/elide && echo '{"name":"a","private":true,"workspaces":["."],"scripts":{"s":"yes l|head -5"}}' > /tmp/elide/package.json
run() { local r; r=$( (cd /tmp/elide && FORCE_COLOR=1 BUN_CONFIG_ELIDE_LINES=2 "$1" run --filter='*' s) 2>&1 | grep -oE '\[[0-9]+ lines elided\]' | head -1); echo "${r:-(no elision)}"; }
echo "bun-debug: $(run "$PWD/build/debug/bun-debug")"
echo "bun:       $(run bun)"

Output before/after:

bun-debug: [3 lines elided]
bun:       (no elision)

Cause

Not implemented

Fix

Implement support for the flag

Verification

Added regression tests
All existing tests still pass

 bun bd test test/cli/run/filter-workspace.test.ts -t elide
$ BUN_DEBUG_QUIET_LOGS=1 bun scripts/build.ts --profile=debug --quiet test test/cli/run/filter-workspace.test.ts -t elide
bun test v1.3.11 (60619bdc)

test/cli/run/filter-workspace.test.ts:
✓ bun > elides output by default when using --filter [233.32ms]
✓ bun > respects --elide-lines argument [233.29ms]
✓ bun > --elide-lines=0 shows all output [236.42ms]

jakeboone02 · 2025-03-12T21:45:28Z

You should update the description to say "fixes #16286" so it gets closed as soon as this is merged.

And thanks for this!

martinamps · 2025-03-12T22:23:02Z

aha good point, done. Hopefully gets unblocked soon cc @RiskyMH

martinamps · 2025-03-14T21:34:03Z

sorry @Jarred-Sumner those windows ones are a pain 😁 i should've made a way to mock that behavior probably

martinamps · 2025-03-16T07:16:14Z

All passing @RiskyMH @Jarred-Sumner :) anything else I need to do on this one, especially around docs? first time changing those

martinamps · 2025-04-22T22:25:01Z

@RiskyMH @Jarred-Sumner 👀 ?

martinamps · 2025-06-01T19:25:42Z

bump @Jarred-Sumner @RiskyMH 🙇

just-toby · 2025-08-06T16:04:27Z

@martinamps thanks for this change, I'd love to see it merged! Could you please resolve the conflicts and maybe add the maintainers as reviewers explicitly? Hopefully that'll help :)

Trosterud · 2025-08-14T12:41:13Z

So looking forward to this 🔥

spalger · 2026-03-19T18:08:31Z

@martinamps would be very grateful if you could get this over the line. The fact that using bun run in CI hides error messages is a trap I see people fall in over and over

…non-terminal environments (oven-sh#16286) - add support for BUN_CONFIG_ELIDE_LINES (oven-sh#11465)

coderabbitai · 2026-04-03T03:56:34Z

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Pro

Run ID: 4d72811a-1476-4c5e-b83e-7dc3f53913aa

📥 Commits

Reviewing files that changed from the base of the PR and between 60619bd and 6a0dc0c.

📒 Files selected for processing (1)

scripts/build/flags.ts

Walkthrough

Adds a new exported env var BUN_CONFIG_ELIDE_LINES (documented) and uses it as a fallback for --elide-lines. Adjusts runtime behavior to warn (not exit) when --elide-lines is used in non-terminal environments. Tests updated to cover env-var behavior and CLI precedence.

Changes

Cohort / File(s)	Summary
Environment Variable & Docs `src/env_var.zig`, `docs/runtime/environment-variables.mdx`	Added exported unsigned env-var `BUN_CONFIG_ELIDE_LINES` and documented its semantics in the runtime environment variables reference.
CLI Argument Parsing `src/cli/Arguments.zig`	When `--elide-lines` is not provided, read `BUN_CONFIG_ELIDE_LINES` as a fallback for `ctx.bundler_options.elide_lines` in two CLI parsing sites.
Filter Run Behavior `src/cli/filter_run.zig`	Replaced hard exit on `--elide-lines` in non-terminal environments with detection of stdout TTY (`windowsIsTerminal()` on Windows, `Output.isStdoutTTY()` elsewhere) and emit a non-fatal `Output.warn(...)` so execution continues when stdout is not a terminal.
Tests `test/cli/run/filter-workspace.test.ts`	Refactored test helper to accept an `env` object and optional `elideLines`; removed win32-specific failure expectations; merged spawned env handling; added tests verifying `BUN_CONFIG_ELIDE_LINES` controls elision and that explicit `--elide-lines` takes precedence.
Build flags (unrelated) `scripts/build/flags.ts`	Added imports and a conditional macOS linker flag that prepends `-L<...>/lib/c++` derived from the configured `c.cxx` path when that directory exists.

🚥 Pre-merge checks | ✅ 3 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Out of Scope Changes check	⚠️ Warning	The only out-of-scope change is in scripts/build/flags.ts (adding macOS linker flags), which appears unrelated to the elide-lines feature or issue `#16286`.	Remove the macOS linker flag changes from scripts/build/flags.ts or clarify their necessity and relation to the primary objectives.

✅ Passed checks (3 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and concisely describes the main change: adding support for BUN_CONFIG_ELIDE_LINES environment variable and fixing issue `#16286`.
Description check	✅ Passed	The PR description provides a clear summary, includes a reproducer script, explains the cause and fix, and documents verification with test results.
Linked Issues check	✅ Passed	The PR addresses issue `#16286` by changing --elide-lines to warn instead of error in non-terminal environments, and adds BUN_CONFIG_ELIDE_LINES support as proposed.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

coderabbitai

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@src/cli/filter_run.zig`:
- Around line 553-555: The warning currently checks state.pretty_output (derived
from Output.enable_ansi_colors_stdout) which can be true when colors are forced;
instead perform a real TTY check for stdout before warning. Modify the branch
that inspects ctx.bundler_options.elide_lines and state.pretty_output to call
the platform TTY detection (e.g., use the std.os isatty/getStdOut equivalent)
and only warn when stdout is not a TTY; update the condition around
ctx.bundler_options.elide_lines to use that TTY check rather than
state.pretty_output so forced-color scenarios don't bypass the non-terminal
warning.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

Push a commit to this branch (recommended)
Create a new PR with the fixes

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Pro

Run ID: f950bcff-8330-4665-b2f8-32b422547f88

📥 Commits

Reviewing files that changed from the base of the PR and between 17bd6cb and 1f77577.

📒 Files selected for processing (5)

docs/runtime/environment-variables.mdx
src/cli/Arguments.zig
src/cli/filter_run.zig
src/env_var.zig
test/cli/run/filter-workspace.test.ts

@martinamps

Split out from #18111 by @martinamps. `--elide-lines` currently exits with an error when stdout is not a terminal, which breaks scripts that pass the flag and run in both interactive and CI/hook contexts. The flag is already a no-op in this case (the elision code only runs in the TTY redraw path), so the error serves no purpose. This removes it. Fixes #16286 --------- Co-authored-by: Martin Amps <m@rtin.so> Co-authored-by: Martin Amps <mamps@anthropic.com> Co-authored-by: robobun <robobun@bun.sh>

alii · 2026-04-08T23:33:16Z

Split into #28978 and #28977

@martinamps

Split out from #18111 by @martinamps. Adds `BUN_CONFIG_ELIDE_LINES` as an environment-variable equivalent of `--elide-lines`, so the limit can be set once globally instead of per command. The CLI flag takes precedence when both are set. Closes #11465. Rebased after #28977 landed on main (which simplified the non-terminal error path into a silent no-op), so this branch no longer needs its own CLI-source tracking. Fixes on top of a straight env-var fallback: - Consolidate the two duplicated `--elide-lines` parse blocks in `Arguments.zig` into a single `parseElideLinesOption` helper, called after `loadConfigWithCmdArgs` so CLI + env var beat bunfig.toml. - Empty `--elide-lines ""` (e.g. `--elide-lines "$MAYBE_UNSET"`) falls through to the env var instead of shadowing it. - Env-var value is clamped via `std.math.cast` rather than `@intCast`, so 32-bit targets don't panic on values above `maxInt(usize)`. - `bunfig.zig` rejects negative, non-finite, fractional, and out-of-range `run.elide-lines` floats with an `Expected a non-negative integer` error. The `>=` upper bound accounts for f64 rounding `maxInt(u64)` up to `2^64` — a strict `>` would let the boundary value slip into a panicking `@intFromFloat`. Tests cover CLI vs env-var precedence, the empty-string fallback, the bunfig fractional rejection, and the non-terminal no-op contract.

@martinamps

Split out from #18111 by @martinamps. Adds `BUN_CONFIG_ELIDE_LINES` as an environment-variable equivalent of `--elide-lines`, so the limit can be set once globally instead of per command. The CLI flag takes precedence when both are set. Closes #11465. Rebased after #28977 landed on main (which simplified the non-terminal error path into a silent no-op), so this branch no longer needs its own CLI-source tracking. Fixes on top of a straight env-var fallback: - Consolidate the two duplicated `--elide-lines` parse blocks in `Arguments.zig` into a single `parseElideLinesOption` helper, called after `loadConfigWithCmdArgs` so CLI + env var beat bunfig.toml. - Empty `--elide-lines ""` (e.g. `--elide-lines "$MAYBE_UNSET"`) falls through to the env var instead of shadowing it. - Env-var value is clamped via `std.math.cast` rather than `@intCast`, so 32-bit targets don't panic on values above `maxInt(usize)`. - `bunfig.zig` rejects negative, non-finite, fractional, and out-of-range `run.elide-lines` floats with an `Expected a non-negative integer` error. The `>=` upper bound accounts for f64 rounding `maxInt(u64)` up to `2^64` — a strict `>` would let the boundary value slip into a panicking `@intFromFloat`. Tests cover CLI vs env-var precedence, the empty-string fallback, the bunfig fractional rejection, and the non-terminal no-op contract.

@martinamps

…#28977) Split out from oven-sh#18111 by @martinamps. `--elide-lines` currently exits with an error when stdout is not a terminal, which breaks scripts that pass the flag and run in both interactive and CI/hook contexts. The flag is already a no-op in this case (the elision code only runs in the TTY redraw path), so the error serves no purpose. This removes it. Fixes oven-sh#16286 --------- Co-authored-by: Martin Amps <m@rtin.so> Co-authored-by: Martin Amps <mamps@anthropic.com> Co-authored-by: robobun <robobun@bun.sh>

@martinamps

Split out from #18111 by @martinamps. Adds `BUN_CONFIG_ELIDE_LINES` as an environment-variable equivalent of `--elide-lines`, so the limit can be set once globally instead of per command. The CLI flag takes precedence when both are set. Closes #11465. Rebased after #28977 landed on main (which simplified the non-terminal error path into a silent no-op), so this branch no longer needs its own CLI-source tracking. Fixes on top of a straight env-var fallback: - Consolidate the two duplicated `--elide-lines` parse blocks in `Arguments.zig` into a single `parseElideLinesOption` helper, called after `loadConfigWithCmdArgs` so CLI + env var beat bunfig.toml. - Empty `--elide-lines ""` (e.g. `--elide-lines "$MAYBE_UNSET"`) falls through to the env var instead of shadowing it. - Env-var value is clamped via `std.math.cast` rather than `@intCast`, so 32-bit targets don't panic on values above `maxInt(usize)`. - `bunfig.zig` rejects negative, non-finite, fractional, and out-of-range `run.elide-lines` floats with an `Expected a non-negative integer` error. The `>=` upper bound accounts for f64 rounding `maxInt(u64)` up to `2^64` — a strict `>` would let the boundary value slip into a panicking `@intFromFloat`. Tests cover CLI vs env-var precedence, the empty-string fallback, the bunfig fractional rejection, and the non-terminal no-op contract.

@martinamps

…#28977) Split out from oven-sh#18111 by @martinamps. `--elide-lines` currently exits with an error when stdout is not a terminal, which breaks scripts that pass the flag and run in both interactive and CI/hook contexts. The flag is already a no-op in this case (the elision code only runs in the TTY redraw path), so the error serves no purpose. This removes it. Fixes oven-sh#16286 --------- Co-authored-by: Martin Amps <m@rtin.so> Co-authored-by: Martin Amps <mamps@anthropic.com> Co-authored-by: robobun <robobun@bun.sh>

@martinamps

Split out from #18111 by @martinamps. Adds `BUN_CONFIG_ELIDE_LINES` as an environment-variable equivalent of `--elide-lines`, so the limit can be set once globally instead of per command. The CLI flag takes precedence when both are set. Closes #11465. Implemented against the Rust CLI (`src/runtime/cli/Arguments.rs`, `src/bun_core/env_var.rs`, `src/bunfig/bunfig.rs`) after the src port: - Consolidate the two duplicated `--elide-lines` parse blocks into a single `parse_elide_lines_option` helper, called after `load_config_with_cmd_args` so CLI + env var beat `run.elide-lines` in bunfig.toml. - Empty `--elide-lines ""` (e.g. `--elide-lines "$MAYBE_UNSET"`) falls through to the env var instead of shadowing it. - Env var value is clamped via `usize::try_from(...).unwrap_or(usize::MAX)` so 32-bit targets don't lose high bits above `usize::MAX`. - `bunfig.rs` rejects negative, non-finite, fractional, and out-of-range `run.elide-lines` numbers with `Expected a non-negative integer` instead of silently saturating/truncating via `as usize`. Tests cover CLI vs env-var precedence, the empty-string fallback, the bunfig fractional rejection, and the non-terminal no-op contract. The bunfig test uses `--config=./bunfig.toml` (`=` form) because clap's `.one_optional` does not consume a following argv element.

martinamps mentioned this pull request Mar 12, 2025

Allow configuration for max lines when eliding filter output #11465

Closed

martinamps force-pushed the elide-lines-env-var branch from 6b289fc to 9d5e375 Compare March 12, 2025 06:17

martinamps changed the title ~~add support for BUN_CONFIG_ELIDE_LINES~~ add support for BUN_CONFIG_ELIDE_LINES (& fixes #16286) Mar 12, 2025

martinamps force-pushed the elide-lines-env-var branch 2 times, most recently from 3182af1 to 82331bc Compare March 14, 2025 21:39

just-toby mentioned this pull request Aug 27, 2025

--elide-lines should warn or be ignored, not error, in non-terminal environments #16286

Closed

- change from error to warning: --elide-lines will have no effect in …

1f77577

…non-terminal environments (oven-sh#16286) - add support for BUN_CONFIG_ELIDE_LINES (oven-sh#11465)

martinamps force-pushed the elide-lines-env-var branch from 2488bad to 1f77577 Compare April 3, 2026 03:40

coderabbitai Bot reviewed Apr 3, 2026

View reviewed changes

Comment thread src/cli/filter_run.zig Outdated

martinamps and others added 3 commits April 3, 2026 09:30

real tty check for elide lines warning

60619bd

Merge branch 'main' into elide-lines-env-var

6a0dc0c

Merge branch 'main' into elide-lines-env-var

36a53df

This was referenced Apr 8, 2026

cli: make --elide-lines a no-op in non-terminal environments #28977

Merged

cli: add BUN_CONFIG_ELIDE_LINES env var #28978

Open

alii closed this Apr 8, 2026

Navigation Menu

add support for BUN_CONFIG_ELIDE_LINES (& fixes #16286) - #18111

add support for BUN_CONFIG_ELIDE_LINES (& fixes #16286)#18111
martinamps wants to merge 4 commits into
oven-sh:mainfrom
martinamps:elide-lines-env-var

martinamps commented Mar 12, 2025 •

edited

Loading

Uh oh!

jakeboone02 commented Mar 12, 2025

Uh oh!

martinamps commented Mar 12, 2025

Uh oh!

martinamps commented Mar 14, 2025

Uh oh!

martinamps commented Mar 16, 2025

Uh oh!

martinamps commented Apr 22, 2025

Uh oh!

martinamps commented Jun 1, 2025

Uh oh!

just-toby commented Aug 6, 2025

Uh oh!

Trosterud commented Aug 14, 2025

Uh oh!

spalger commented Mar 19, 2026

Uh oh!

coderabbitai Bot commented Apr 3, 2026 •

edited

Loading

❌ Failed checks (1 warning)

Uh oh!

coderabbitai Bot left a comment

Uh oh!

Uh oh!

alii commented Apr 8, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

Conversation

martinamps commented Mar 12, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Repro

Cause

Fix

Verification

Uh oh!

jakeboone02 commented Mar 12, 2025

Uh oh!

martinamps commented Mar 12, 2025

Uh oh!

martinamps commented Mar 14, 2025

Uh oh!

martinamps commented Mar 16, 2025

Uh oh!

martinamps commented Apr 22, 2025

Uh oh!

martinamps commented Jun 1, 2025

Uh oh!

just-toby commented Aug 6, 2025

Uh oh!

Trosterud commented Aug 14, 2025

Uh oh!

spalger commented Mar 19, 2026

Uh oh!

coderabbitai Bot commented Apr 3, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

❌ Failed checks (1 warning)

Uh oh!

coderabbitai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

alii commented Apr 8, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

martinamps commented Mar 12, 2025 •

edited

Loading

coderabbitai Bot commented Apr 3, 2026 •

edited

Loading