Skip to contents

Changelog

Source: NEWS.md

easypaper 0.1.0

First release.

A project that holds only the paper

  • create_paper() writes a Quarto project for one manuscript: the text in _sections/, one file per section, the references and the journal’s citation style in references/, the setup of the analysis in R/setup.R and the data in data/. The build logic lives in the package, so a fix reaches every project with update.packages(). Also in RStudio, under File > New Project.
  • create_example_paper() writes the same project with a small study in it – the trees of Barro Colorado Island – to see how each part is written.
  • Authors and affiliations are written in the YAML of manuscript.qmd, the way Quarto documents them; every render numbers the affiliations and prints the authors on one line, as a paper does (affiliations()).
  • A short title in the YAML is printed under the title of every document but the preprint, and counted.
  • Settings go in the easypaper: block of _quarto.yml, so they travel with the paper.
  • Every function works on the project you name: once a session with set_paper("."), or in a call with path. One that writes never takes the working directory for the project: with none named, it stops before writing. One that only reads takes the working directory’s project when none is named.
  • The project’s README.md, what a repository shows first, carries the title, the authors with their affiliations and ORCID, the keywords and the licence notice from the first commit, and the DOI of the data once it is reserved; every render keeps them in step with manuscript.qmd.

Data

  • convert_data() brings originals into data/ in open formats: workbooks and statistical files become .csv, files already open are copied, and the originals are only ever read. Outside a project, to says where the results go.
  • sync_metadata() describes the data deposit from what the project already knows – the title, the keywords, the authors with their affiliations, the email of the corresponding author and their ORCID, the licence of the data from the project’s LICENSE, every file of data/, and the variables of each table, read with its own separator (,, ; or a tab) – and edit_metadata() opens an editor for the rest. make_preprint() and make_submission() compile it into dataspice.json and its page before they build the compendium, so the deposit describes the data as they are. The page of the metadata draws the area the data cover as a map inside it, with dataspice 1.2.1 and sf, maps and base64enc; without them it has no map. It never loads one from the internet: it opens offline and fetches nothing.

Rendering

  • render_html(), render_docx(), render_pdf() and render_supplementary() write documents to read into output/, in the style of the journal the manuscript names or of any other; render_all() runs render_docx(), render_pdf() and export_code() in one go.
  • Quarto is found where RStudio, Positron or Quarto’s installer put it, also from Rscript or a terminal, where it is not on the PATH. It has to be 1.7 or later. preview() reloads as you save. Every document is named after the project, manuscript_<project>.docx, with the journal added when it is not the manuscript’s own.
  • The main text has its lines numbered, in the .docx and in the .pdf. The affiliations and the keywords line start without the first-line indent of the text.
  • The supplement is a document of its own, with its own reference list and its own numbering: Figure S1, Table S1; -@sfig-x cites the number alone, S1. It opens with “Supporting Information” and, under it, the reference of the paper, flush left; a preprint’s names the server instead of a journal. One that cites nothing has no reference list.
  • Figures are exported as PNG, JPEG and TIFF on every render, at the dpi of _quarto.yml (600 in a new project).
  • The .pdf sets the title block flush left, with an ORCID icon beside each author who has one, justifies the captions, and prints the Greek letters of the text, which LaTeX’s font lacks.
  • export_code() writes the code of every chunk into one script, analysis_code.R, that opens with a table of contents: the line each section starts on and the chunks it holds. Each analysis lives in the results section that reports it, so each is found under its own heading.
  • Captions follow the crossref: block of _quarto.yml, or a caption style for one render: "abbrev", "colon", "compact", "nature", or one of your own. The text may cite with another word than the caption’s: Figure 1. under the figure, Fig. 1 in the text.
  • list_journals() and add_journal() manage the citation styles, fetched from the official CSL repository.
  • A new project cites easypaper and renv in the last paragraph of its Methods, ready to keep.
  • write_packages_bib() writes the references of the R packages the manuscript uses, as knitr::write_bib() does, with the names in their titles kept as written: vegan: Community ecology package, not Vegan, in a journal that sets titles in sentence case.

Checks

  • Every render checks citations, cross-references, packages, data files and the title first (?checks), and the documents for others and the submissions record the environment they came out of in renv.lock; check_renv() compares that record with the library, package by package.
  • Every render sets the scientific names of the reference list in italics, checked against GBIF; italicize_species() does it for any bibliography, into the CSL-JSON file you name, and check_species() lists the names left in doubt.
  • check_species_text() reads the text the way an editor does: italics, first mention, abbreviation, authority.
  • manuscript_stats() gives what a submission form asks for: characters of the title, words of the abstract and the main text, references, figures and tables. Every render prints it. It finds the abstract and the reference list under the names journals give them – Summary, Literature cited – or under the ones abstract-heading: and references-heading: name.

Submitting

  • make_submission() builds the folder for a journal into submission/: double-blind by default, with a title page and a main text that names nobody, standalone figures renumbered and named with their resolution and printed size (Figure_1_600dpi_180x120mm.tiff; figure_dpi changes the resolution for one call), a cover letter already dated, titled and signed by the corresponding author – written from the last one, with those brought up to date, when the paper goes to a second journal – a checklist, and a data and code compendium – with a copy for the reviewers that names nobody either. blind = FALSE for a signed manuscript.
  • A double-blind submission is read back for the authors’ names, emails, ORCID, affiliations and accounts, and what the text itself still says is reported, file by file; a reference that would name them – their own software, their own data paper – is left out of it when blinded-drop-keys: lists it.
  • The compendium carries a renv.lock of its analysis: the packages it loads and what they need, without easypaper, which built the documents and is recorded in the project’s own.
  • make_preprint() builds the preprint: one signed PDF and its supplement, with the same compendium.
  • A submission is never overwritten: a second make_submission() or make_preprint() to the same journal or server builds the next version beside the first, JournalofEcology_v2, with its files named the same way.
  • deposit_zenodo() reserves the DOI of the data on Zenodo and writes it into the text and the cover letters, without publishing. The deposit carries the licence of the project’s LICENSE, as the data’s metadata do; an ORCID iD Zenodo could not read is left out of it, with a warning.

Co-authors

  • td_upload(), td_download() and td_update() share a section with co-authors in Google Docs, through trackdown. Experimental: they need the development version of trackdown, which is not on CRAN.