Repository navigation
Conversation
One paragraph in the vignette after use_brand() and a seealso entry: branded PDF and Word reports use the openwashdata Quarto extension, installed with quarto add in the package root, reading the _brand.yml that use_brand() installed. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01YZGMkJRUtmXRukGv5PLhEx
Since openwashdata/brand 1.0.0 a logo image entry is a list with path and alt. brand_logo_paths() flattened the entries with unlist() and treated the alt text as a second path, so use_brand(ref = "v1.0.0") failed after writing _brand.yml. The helper now takes the path of a list entry, for named images and size entries alike. A test covers object-style entries. Assisted-by: Claude claude-fable-5-1 Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013ouAd1vCAj75yYpUNZJw3F
washr 1.1.0 reached CRAN on 2026-09-02 and is tagged v1.1.0 at 3a3eeba. Dev moves to the next development version, with a fresh heading in NEWS.md for the changes that follow. Refs #78 Assisted-by: Claude claude-fable-5-1 Claude-Session: https://claude.ai/code/session_011SQ621ToQFmGWRnbGaLMkA
4 tasks
setup_ci() wrote the template through usethis::use_template(), which
renders with whisker. whisker reads {{ }} as its own delimiters, so it
consumed the inner braces of every ${{ ... }} GitHub Actions expression
and left a bare $ behind. The written workflow read "runs-on: $", which
matches no runner, so each matrix job queued until the 24 hour limit and
then failed. A package scaffolded with 1.1.0 shipped a workflow that
never ran, which made the review floor claim in #86 false.
The template carries no variables, so it is copied instead of rendered.
The regression test asserts the expressions survive and that no key is
left with a bare $ as its value. A corrupted file still parses as YAML,
which is why the existing test did not catch this.
Version 1.1.1, with the use_brand() alt text fix already on dev (#123).
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01KniirkTyC9pRpXrCA1JeF3
larnsce
changed the title
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Patch release. Two fixes, both of which stopped a scaffolded package from building or checking.
Closes #123. Closes #130.
setup_ci() wrote a workflow that never ran (#130)
setup_ci()wrote the template throughusethis::use_template(), which renders with whisker. whisker reads{{ }}as its own delimiters, so it consumed the inner braces of every${{ ... }}GitHub Actions expression and left a bare$:runs-on: $matches no runner, so each matrix job queued until the 24 hour limit and then failed. Seen on openwashdata/owdata#17, where all five jobs failed at 24h0m1s. This also made the review floor claim in #86 false, since the workflow a scaffolded package shipped never ran.The template carries no variables, so it is now copied instead of rendered. A regression test asserts the expressions survive and that no key is left with a bare
$as its value; the existing test did not catch this because a corrupted file still parses as YAML.Scope: only packages scaffolded with 1.1.0, when
setup_ci()shipped. A scan of the R-CMD-check workflow on the default branch of every non-archived repository in the organisation found no affected published package.use_brand() failed on brand 1.0.0 logo entries (#123)
brand_logo_paths()reads thepathof a logo entry written as a list of path and alt text, the form openwashdata/brand uses since 1.0.0. Before,unlist()turned the alt text into a second path and the call failed after it had already written_brand.yml.Test plan
devtools::test(): 190 pass, 0 faildevtools::check(): 0 errors, 0 warnings, 0 notes${{ }}expressionswashr::use_brand(ref = "v1.0.0")writes the four new logos (Refresh the brand to openwashdata/brand 1.0.0 fslogisticskampala#17)