Cleaned sourcepos records when they come from metadata (thanks, @dmurdoch, #111).
The markdown package is in the maintenance-only mode now. It is feature-complete, and will receive no updates except for fixing CRAN problems. New development will continue only in litedown: https://github.com/yihui/litedown.
Provided three internal functions html_document, html_vignette, and pdf_document as compatibility layers to functions of the same names in the rmarkdown package (thanks, @jangorecki, #108).
The default HTML template no longer wraps meta variables include-before and include-after inside <div></div>, because their values may contain incomplete HTML tags, e.g., include-before = '<div>' and include-after = '</div>'.
```{lang attr1 attr2 ...} were not correctly rendered.```{=lang}) were broken in the previous version when the support for code block attributes was added.Added support for attributes on fenced code blocks, e.g., ```{.lang .class2 #id attr="value"} (thanks, @thothal, #106).
Fixed the bug that the option number_sections: true doesn't work for HTML output when then input contains certain Unicode characters (thanks, @fyuniv, #104).
Added support for rendering HTML Widgets such as ggplotly (thanks, @fyuniv, #105).
mark_html() (thanks, @kenjisato, #103).The file argument of mark() will be treated as a file path only if the file exists and the value is not wrapped in I(). Previously, it would be treated as a file path when it has a file extension, which could lead to confusing errors like #100 (thanks, @LukasWallrich).
When there are emojis in the text, mark() may fail to identify and embed web resources (thanks, @tdhock, yihui/knitr#2254).
Added support for footnotes, fenced Divs, section numbers, {} attributes for images/headings/fenced Divs, and appendices. See vignette('intro', package = 'markdown') for details.
A lot of enhancements to the HTML slides format. See vignette('slides', package = 'markdown') for details.
Added vignette('article', package = 'markdown') to demonstrate how to write an HTML article.
If the input to mark() is a file, the output will also be a file by default. Previously the output would be text. If you want mark() to return text output when the input is a file, you may specify the argument output = NULL.
The Markdown option base64_images has been renamed to embed_resources. This option can take two possible values, "local" and "https", meaning whether to embed local and/or web (https) resources. You can specify none, either, or both of them. See vignette('intro', package = 'markdown') for details.
Removed the option standalone from the list of Markdown options. Please use the argument template = TRUE/FALSE of mark() instead. The option standalone = TRUE was equivalent to template = TRUE.
Added the option auto_identifiers (enabled by default) to automatically add IDs to headings, e.g., # Hello world! will be converted to <h1 id="hello-world">Hello world!</h1>. You can certainly override the automatic ID by providing an ID manually via the {#id} attribute, e.g., # Hello world! {#hello}.
Renamed the mathjax option to js_math to allow for other JS math libraries. The default library was changed from MathJax to KaTeX. To continue using MathJax, you may set js_math: mathjax.
Removed the option mathjax_embed from the list of Markdown options. To embed the MathJax library, enable "https" in the embed_resources option instead. Note that only MathJax v3 can be partially embedded, and lower versions cannot.
Renamed the option highlight_code to js_highlight, and added support for an alternative syntax highlighting JS library Prism.js, which became the default. To continue using the old default highlight.js, you may set the js_highlight option to highlight.
The default version of MathJax has been changed from v2 to v3.
The default version of highlight.js has been changed from 11.6.0 to 11.7.0, and the default style has been switched from github to xcode.
Values of meta variables title, author, and date (if provided) will be transformed to the target output format before they are passed into templates.
Fixed the bug that the default CSS was not added to HTML output.
Removed dependency on the mime package.
Added experimental support for HTML slides: markdown::mark_html(..., meta = list(css = c('default', 'slides'), js = 'slides')). If you prefer knitting Rmd documents in RStudio, you may use the output format:
output:
markdown::html_format:
meta:
css: [default, slides]
js: [slides]
See https://yihui.org/en/2023/01/minimal-r-markdown/ for a demo.
Empty \title{} in LaTeX output will be removed (along with \maketitle).
highlight.js is loaded from https://www.jsdelivr.com/package/gh/highlightjs/cdn-release by default now. This means more languages are supported (not only R), but also means syntax-highlighting will not work offline at the moment (it will be improved in future).
MathJax failed to load in the previous version. The bug has been fixed now.
Removed the function markdownExtensions().
Switched the underlying Markdown rendering engine from the C library sundown (which has been deprecated for a decade) to the R package commonmark (thanks, @jeroen, yihui/knitr#1329).
The functions renderMarkdown() and markdownToHTML() have been renamed to mark() and mark_html(), respectively. The old names are still kept in this package for backward-compatibility.
Removed the arguments stylesheet and fragment.only in mark_html(). For stylesheet, please use the argument meta = list(css = ...) to provide the CSS stylesheet. For fragment.only, please use mark_html(template = FALSE) or mark_html(options = '-standalone') instead of fragment.only = TRUE. Currently these old arguments are still accepted internally, but may be deprecated and dropped in the long run.
The file argument of mark() and mark_html() can also take a character vector of Markdown text now.
Removed functions rendererExists(), rendererOutputType(), and registeredRenderer(). They were primarily for internal use.
Deprecated the function markdownExtensions(). All extensions should be specified via the options argument of functions like mark(), e.g., mark(options = '+table+tasklist'). See all options on the help page ?markdown::markdown_options.
Renamed markdownHTMLOptions() to markdown_options().
The default value of the encoding argument of renderMarkdown() and markdownToHTML() has been changed from getOption("encoding") to "UTF-8". The encoding of the input file will always be assumed to be UTF-8.
markdownToHTML() will return a character vector encoded in UTF-8 (instead of the system's native encoding) when not writing to an output file.
renderMarkdown() works now even if text = character(0) or ""
added an encoding argument to renderMarkdown() since multi-byte characters in renderMarkdown() did not work on Windows (thanks, Kohske Takahashi, #63)
fixed #64: invalid 'n' argument in rpubsUpload() (thanks, Wouter van Atteveldt)
images that have already been base64 encoded will not be encoded again (#61)
the URL of the MathJax CDN was updated to cdn.mathjax.org
fixed #60: MathJax may be included even if it is unnecessary when syntax highlighting is enabled (thanks, @aoles)
fixed a bug which may hang R when building R Markdown vignettes in a wrong locale (thanks, Dan Tenenbaum, yihui/knitr#782)
if both the 'file' and 'text' arguments are provided but file = NULL, e.g. markdownToHTML(file = NULL, text = ?), markdownToHTML() can throw an error indicating the file is invalid (thanks, Tyler Rinker, hadley/staticdocs#66)
markdownToHTML(text = ?, output = ?) was broken (#54)
added 'mathjax_embed' to HTML options for embedding the MathJax JavaScript in the HTML document rather than linking to it online. Note the JavaScript code is read from the http instead of https MathJax URL. Contributed by Henrik Bengtsson.
added another vignette to show the HTML output of the original vignette (see browseVignettes('markdown'))
the default CSS style was tweaked (major changes include: page width is at most 800px, more line height, slightly larger fonts, and a different syntax highlighting theme)
added a new argument 'template' to markdownToHTML() so that we can customize the HTML template (by default, it uses the template 'resources/markdown.html' in this package); thanks, Nacho Caballero
the options markdown.HTML.stylesheet and markdown.HTML.header used in markdownToHTML() can be character vectors (they will be processed by paste(x, collapse = '\n')
fixed #26: the HTML output file will be written relative to the current working directory now when it contains images that need to be base64 encoded
fixed #28: the image URL should be decoded before the image is based64 encoded
the markdown extension 'ignore_math' was replaced with 'latex_math'.
users can now use the markdown.HTML.stylesheet option to override the package default stylesheet.
setting the fragment_only rendering option or the fragment.only parameter to markdownToHTML will base64 encode images if applicable. version 0.5.1 did not.
fixed a GUIDgenerator bug; for escaping math equations before markdown parsing begins.
image encoding was fixed for the case when there are more than one included in a markdown document.
added fragment.only parameter to markdownToHTML
added new html rendering options base64_images, fragment_only, mathjax, and highlight_code
added new markdown extension ignore_math
removed safelink from default html rendering options
the default html rendering options are now hard_wrap, use_xhtml, smartypants, and base64_images.
added support for post-processing HTML using smartypants filter
added optional support for rendering a table of contents
changed exported C functions to use an rmd_ prefix (eliminating potential symbol conflicts with other packages)
changed default html rendering options to use_xhtml, hard_wrap, safelink, and smartypants