Updated md4c to version 0.6.0.
Added support for the new md4c extensions (footnotes, admonitions, preserved blank lines, highlight, insert, spoilers, superscripts, and subscripts), each with a parser flag, node class, constructor, and to_md() / to_html() methods.
MD_DIALECT_GITHUB (and so parse_gfm()) now also enables footnotes and admonitions, following upstream.
to_html() now renders surrogate and out of range numeric character references as U+FFFD instead of erroring.
to_md() round trip fixes: strikethrough is written as ~~text~~, indented code following a list stays out of the list, [ and ] are escaped in wiki link targets, and emphasis starting with * uses _.
Updated md4c to version 0.5.3.
to_html() no longer doubles the language- prefix on a fenced code block
whose info string already starts with language-, matching the upstream
md4c-html behavior.
Added $ and $<- methods for md_node objects to read and modify node
content and attributes by name.
Renamed md_text_br() and md_text_softbr() to md_text_break() and
md_text_softbreak() so constructor classes match parser output.
Added md_text_*, md_span_*, and md_block_* functions to create markdown elements.
Switched spec-based tests to use code generation (thanks to @krlmlr).
Clearer errors in to_md() for empty md_block_p and for md_block_li
missing required arguments.
C++ parser now emits an md_span_other fallback for unrecognized spans.
Added rlang to Suggests; test regeneration skipped cleanly when
rlang/styler are missing.
Fixed wrong dependency vector in helper-tests.R md4c test regeneration.
Initial CRAN submission.
Based on md4c version 0.5.2
Implemented functionality for parsing markdown, displaying the AST, and generating html or md