washr 1.2.0: readiness report, dictionary updates, Zenodo metadata, values from DESCRIPTION

.Rbuildignore

@@ -11,3 +11,4 @@
 ^CRAN-SUBMISSION$
 ^dev$
 
+^PKGREVIEW_FLOOR$

.github/workflows/scaffold-check.yaml

@@ -0,0 +1,65 @@
+# Scaffold check (#129). Builds a fixture data package with washr in the order
+# the vignette documents, runs every re-runnable step a second time, and runs
+# the pkgreview check script on the result. A change in washr that drops a
+# required item of the review standard turns this workflow red. The pkgreview
+# version comes from the PKGREVIEW_FLOOR file; raise it when pkgreview
+# releases.
+on:
+  push:
+    branches: [main, dev]
+  pull_request:
+    branches: [main, dev]
+  schedule:
+    - cron: '0 6 * * 1'
+  workflow_dispatch:
+
+name: scaffold-check
+
+permissions: read-all
+
+jobs:
+  scaffold-check:
+    runs-on: ubuntu-latest
+    env:
+      GITHUB_PAT: ${{ secrets.GITHUB_TOKEN }}
+      R_KEEP_PKG_SOURCE: yes
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: r-lib/actions/setup-pandoc@v2
+
+      - uses: r-lib/actions/setup-r@v2
+        with:
+          use-public-rspm: true
+
+      - uses: r-lib/actions/setup-r-dependencies@v2
+        with:
+          # fs to openxlsx: the processing script template. dplyr to
+          # kableExtra: the README template.
+          extra-packages: >
+            any::fs, any::here, any::readr, any::readxl, any::openxlsx,
+            any::dplyr, any::stringr, any::gt, any::kableExtra,
+            local::.
+
+      - name: Fetch the pkgreview check script at the recorded floor
+        run: |
+          ref="$(cat PKGREVIEW_FLOOR)"
+          mkdir -p "$RUNNER_TEMP/pkgreview"
+          curl -fsSL "https://github.com/openwashdata/pkgreview/archive/refs/tags/${ref}.tar.gz" \
+            | tar -xz --strip-components=1 -C "$RUNNER_TEMP/pkgreview"
+          echo "pkgreview ${ref}"
+
+      - name: Scaffold the fixture and run every step twice
+        run: Rscript dev/scaffold-check/scaffold.R "$RUNNER_TEMP/scaffoldcheck"
+
+      - name: Run the pkgreview check script
+        run: |
+          set -o pipefail
+          Rscript "$RUNNER_TEMP/pkgreview/skills/pkgreview-core/check/pkgreview-check.R" \
+            "$RUNNER_TEMP/scaffoldcheck" --org=openwashdata | tee "$RUNNER_TEMP/report.md"
+
+      - name: Add the report to the job summary
+        if: always()
+        run: |
+          if [ -f "$RUNNER_TEMP/report.md" ]; then cat "$RUNNER_TEMP/report.md" >> "$GITHUB_STEP_SUMMARY"; fi

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: washr
 Title: Publication Toolkit for Water, Sanitation and Hygiene (WASH) Data
-Version: 1.1.1
-Date: 2026-09-09
+Version: 1.2.0
+Date: 2026-10-01
 Authors@R: c(
     person("Mian", "Zhong", , "mzhong@ethz.ch", role = "aut",
            comment = c(ORCID = "0009-0009-4546-7214")),
@@ -21,6 +21,7 @@ URL: https://openwashdata.github.io/washr/
 BugReports: https://github.com/openwashdata/washr/issues
 Imports:
     cffr (>= 1.0.1),
+    cli (>= 3.6.0),
     desc (>= 1.4.3),
     jsonlite,
     pkgdown (>= 2.0.7),

NAMESPACE

@@ -1,5 +1,9 @@
 # Generated by roxygen2: do not edit by hand
 
+S3method("[",washr_readiness)
+S3method(format,washr_readiness)
+S3method(print,washr_readiness)
+export(check_publication_readiness)
 export(setup_ci)
 export(setup_dictionary)
 export(setup_rawdata)
@@ -8,7 +12,9 @@ export(setup_roxygen)
 export(setup_website)
 export(update_citation)
 export(update_description)
+export(update_dictionary)
 export(update_metadata)
+export(update_zenodo_json)
 export(use_brand)
 importFrom(lifecycle,deprecated)
 importFrom(utils,head)

NEWS.md

@@ -1,3 +1,156 @@
+# washr 1.2.0
+
+The release that makes washr the engine of a guided workflow. Every
+function now reports through cli and returns what it wrote, a readiness
+report says what a package still lacks before publication, and the values
+that were fixed to openwashdata live in DESCRIPTION, so another group can
+publish with its own. Three exports are new: `check_publication_readiness()`,
+`update_dictionary()` and `update_zenodo_json()`. This release also brings
+the changes of 1.1.1 to CRAN, which was released on GitHub only.
+
+## New
+
+- `check_publication_readiness()` reads the package and reports, item by
+  item, whether it is ready for publication: metadata, data dictionary,
+  documentation and the check workflow. Each gap names the step and the
+  washr function that closes it. The result is a data frame with one row
+  per item (`id`, `area`, `check`, `status`, `detail`, `fix`), so review
+  tools such as pkgreview can consume it. Nothing is written to the package
+  (#82).
+
+- `update_dictionary()` brings `data-raw/dictionary.csv` in line with the
+  data after it changed. It adds a row for each new variable, removes the
+  row of a variable that no longer exists, refreshes the types, and keeps
+  every description and every column you added yourself, such as `unit` or
+  `allowed_values`. A file whose content is current is left untouched. The
+  "already exists" error of `setup_dictionary()` now points to it (#13).
+
+- `update_zenodo_json()` writes `.zenodo.json` from DESCRIPTION, and
+  `update_citation()` calls it. Zenodo reads the file when it archives a
+  GitHub release, so a data package is filed as a Dataset in the
+  openwashdata community, with its creators, ORCID iDs, license and
+  keywords, without edits by hand. Set `Config/washr/zenodo-community` in
+  DESCRIPTION for another community (#56).
+
+- `update_citation()` cites the source article of a package that republishes
+  data from a publication. List the article DOI in DESCRIPTION as
+  `X-schema.org-isBasedOn`, and separate several DOIs with commas. Each DOI
+  is looked up at doi.org and written as a `references` entry in
+  CITATION.cff, with a message that asks users to cite both the data package
+  and the article. `inst/CITATION` holds both entries under the same header,
+  so `citation()` prints both. The package stays the work that GitHub's
+  "Cite this repository" shows. A DOI that cannot be looked up keeps its
+  entry from the existing CITATION.cff, so a run without a network
+  connection changes nothing. References that do not come from the field
+  are dropped, because DESCRIPTION is their canonical source. Before, the
+  function had no way to keep such a reference, and a hand-written second
+  entry in `inst/CITATION` was lost on the second run (#134).
+
+- `update_metadata()` writes the same DOIs as schema.org `isBasedOn` in the
+  JSON-LD (#134).
+
+- `options(washr.quiet = TRUE)` silences the messages of washr and of the
+  usethis helpers it calls, for use in scripts. Warnings and errors are
+  never silenced (#84).
+
+- The README of `setup_readme()` says in plain text which version to cite
+  with which DOI, links the authors page of the site and `CITATION.cff`,
+  and names the source article of republished data. Its download links
+  point at the release tag of the version in DESCRIPTION, so a citation of
+  a version leads to the same files later; a development version links the
+  main branch. pkgdown 2.2.0 builds the `llms.txt` of a site from the
+  README, so AI tools now read the same. `check_publication_readiness()`
+  gains two items: `readme_citation` (README.md names the DOI of
+  CITATION.cff) and `pkgdown_llm_docs` (`_pkgdown.yml` keeps the
+  `llms.txt` of the site) (#137).
+
+## Changed
+
+- washr reads the facts of a package from DESCRIPTION, so a group other
+  than openwashdata publishes with its own values by editing that file.
+  The organisation and the repository come from `URL`, the license from
+  `License`. The values with no standard field are `Config/washr/` fields
+  that `update_description()` writes once and never overwrites: `funding`,
+  `analytics-domain`, `doi-provider`, `zenodo-community` and
+  `brand-source`, plus `pages-domain` for a site that is not served from
+  `<organisation>.github.io`. A field set to `none` switches its feature
+  off. The README template, the `_pkgdown.yml` template, the DOI badge,
+  `.zenodo.json`, `use_brand()` and `check_publication_readiness()` read
+  them. For an openwashdata package the output is unchanged (#81).
+
+- `update_description()` takes the GitHub organisation from the repository
+  already listed in `URL` when `github_user` is not given, and no longer
+  adds a second repository under openwashdata. It records the washr
+  version that ran in `Config/washr/version`. For a package under another
+  organisation it writes the funding, analytics, community and brand
+  fields as `none` and lists them (#81).
+
+- `use_brand()` installs the brand at the latest release tag of the brand
+  repository and records the tag in DESCRIPTION as `Config/washr/brand`.
+  Before, it copied from the main branch, which can carry changes no
+  release describes. A second run with no new tag changes no file, and a
+  run after a brand release reports the old and the new tag. Pass
+  `ref = "main"` to try unreleased brand changes (#128).
+
+- Messages now go through cli in every function, with three kinds of lines:
+  a success line for a file that was written, an info line for something
+  that was kept or skipped, and an arrow line for the next step that is left
+  to the user. Errors name the function that was called and carry a hint on
+  how to fix the cause. washr no longer calls the `usethis::ui_*()`
+  functions, which usethis has superseded (#84).
+
+- Every function returns what it wrote, invisibly, and the documentation
+  says so. `setup_rawdata()`, `setup_dictionary()`, `setup_readme()`,
+  `setup_website()` and `update_description()` return the path of their
+  file, `setup_roxygen()` and `update_citation()` return the paths of
+  theirs. Before, most of them returned `NULL` or an internal object (#84).
+
+## Fixed
+
+- Every function acts on the package in the working directory. Before,
+  the usethis helpers wrote into the package of an earlier call when only
+  the working directory had changed since, so files of one package could
+  land in another within the same R session.
+
+- The repository in `URL` is recognised in every common form: with a
+  trailing slash, a `.git` suffix, a fragment or a `www.` prefix. Before,
+  such a URL gave a wrong bug report link or a second repository entry.
+
+- `setup_readme(force = TRUE)` removes the existing `README.Rmd` only once
+  it can write the new one. Before, it deleted the file and then stopped
+  when `data/` held no data object.
+
+- `use_brand()` adds `_brand.yml` and the logo directory to `.Rbuildignore`.
+  Before, `R CMD check` reported both as non-standard top-level files after
+  the brand was installed (#133).
+
+- `use_brand()` wires `_pkgdown.yml` to the brand by adding its lines to
+  the file. Comments and long values stay as they were written. Before, the
+  file was rewritten through the yaml package, which dropped the comments
+  and folded the funding text over two lines, so the review standard no
+  longer found it. The rewrite remains as the fallback for a `template`
+  block that already carries bslib settings (#129).
+
+## Dependencies
+
+- cli is a new import. washr uses it for every message and no longer calls
+  the `usethis::ui_*()` functions, which usethis has superseded (#84).
+
+## Documentation and tests
+
+- The vignette explains packages with more than one dataset, and how a
+  group outside openwashdata sets its own values (#81, #103).
+
+- The README and the vignette name the openwashdata R-universe as a third
+  way to install washr, next to CRAN and GitHub, and say what each source
+  gives you (#131).
+
+- A new workflow, `scaffold-check`, builds a fixture data package with every
+  washr step on each push, runs the steps a second time to confirm that no
+  file changes, and runs the pkgreview check script on the result. The
+  pkgreview version it checks against is recorded in `PKGREVIEW_FLOOR`
+  (#129).
+
 # washr 1.1.1
 
 A patch release with three fixes. The first two stopped a scaffolded package

PKGREVIEW_FLOOR

@@ -0,0 +1 @@
+v1.6.1