@@ -1,3 +1,271 @@
1+pandoc (2.6) PROVISIONAL
2+
3+ * Implement task lists (#3051, Mauro Bieg). Added `task_list`
4+ extension. Task lists are supported form markdown and gfm input.
5+ They should work, to some degree, in all output formats, though in
6+ most formats you'll get a bullet list with a following task list
7+ box. In HTML and LaTeX/PDF output, the bullet will be suppressed
8+ for nicer output. API changes:
9+
10+ + Added constructor `Ext_task_lists` to `Extension`.
11+ + Added `taskListItemFromAscii` and `taskListItemToAscii` to
12+ Text.Pandoc.Shared.
13+
14+ * Add DokuWiki reader (#1792, Alexander Krotov). This adds
15+ Text.Pandoc.Readers.DokuWiki [API change], and adds `dokuwiki`
16+ as an input format.
17+
18+ * HTML reader: handle empty `start` attribute (see #5162).
19+
20+ * RST reader:
21+
22+ + Change treatment of `number-lines` directive (Brian Leung,
23+ #5207). Directives of this type without numeric inputs should
24+ not have a `startFrom` attribute; with a blank value, the
25+ writers can produce extra whitespace.
26+ + Removed superfluous `sourceCode` class on code blocks (#5047).
27+ + Handle `sourcecode` directive as synonynm for `code` (#5204).
28+
29+ * Markdown reader:
30+
31+ + Remove `sourceCode` class for literate Haskell code blocks
32+ (#5047). Reverse order of `literate` and `haskell` classes on
33+ code blocks when parsing literate Haskell, so `haskell` is
34+ first.
35+
36+ * Org reader:
37+
38+ + Handle `minlevel` option differently (#5190, Brian Leung).
39+ When `minlevel` exceeds the original minimum level observed in
40+ the file to be included, every heading should be shifted
41+ rightward.
42+ + Allow for case of `:minlevel == 0` (#5190).
43+ + Fix treatment of links to images (#5191, Albert Krewinkel).
44+ Links with descriptions which are pointing to images are no
45+ longer parsed as inline images, but as links.
46+
47+ * TWiki reader: Fix performance issue with underscores (#3921).
48+
49+ * LaTeX reader:
50+
51+ + Remove `sourceCode` class for literate Haskell code blocks
52+ (#5047). Reverse order of `literate` and `haskell` classes on
53+ code blocks when parsing literate Haskell, so `haskell` is
54+ first.
55+ + Support `\DeclareMathOperator` (#5149).
56+ + Support `\inputminted` (#5103).
57+
58+ * Man reader:
59+
60+ + Use `mapLeft` from Shared instead of defining own.
61+
62+ * Docx reader (Jesse Rosenthal):
63+
64+ + Handle level overrides (#5134).
65+
66+ * Docx writer (Jesse Rosenthal):
67+
68+ + Make Level into a real type, instead of an alias for a tuple.
69+
70+ * ICML writer (Mauro Bieg):
71+
72+ + Support custom-styles (#5137, see #2106).
73+ + Support unnumbered headers.
74+
75+ closes #5140
76+
77+ * Org writer: preserve line-numbering for example and code blocks
78+ (Brian Leung).
79+
80+ * Man/Ms writers: don't escape `-` as `\-`. The `\-` gets rendered
81+ in HTML and PDF as a unicode minus sign.
82+
83+ * RST writer: don't wrap simple table header lines (#5128).
84+
85+ * Asciidoc writer: Shorter delimiters for tables, blockquotes
86+ (#4364). This matches asciidoctor reference docs.
87+
88+ * Dokuwiki writer: Remove automatic `:` prefix before internal image
89+ links (#5183, Damien Clochard). This prevented users from making
90+ relative image links.
91+
92+ * Zimwiki writer: remove automatic colon prefix before internal
93+ images (#5183, Damien Clochard).
94+
95+ * MediaWiki writer: fix caption, use 'thumb' instead of 'frame'
96+ (#5105). Captions used to have the word 'caption' prepended; this
97+ has been removed. Also, 'thumb' is used instead of 'frame' to
98+ allow images to be resized.
99+
100+ * reveal.js writer:
101+
102+ + Ensure that we don't get > 2 levels of section nesting,
103+ even with slide level > 2 (#5168).
104+ + If slide level == N but there is no N-level header, make
105+ sure the next header with level > N gets treated as a slide
106+ and put in a section, rather than remaining loose (#5168).
107+
108+ * Markdown writer:
109+
110+ + Make `plain` RawBlocks pass through in `plain` output.
111+ + Include needed whitespace after HTML figure (#5121).
112+ We use HTML for a figure in markdown dialects that can't
113+ represent it natively.
114+
115+ * Commonmark writer:
116+
117+ + Fix handling of SoftBreak with `hard_line_breaks` (#5195).
118+ + Implement `--toc` (`writerTableOfContents`)
119+ in commonmark/gfm writers (#5172).
120+
121+ * EPUB writer:
122+
123+ + Ensure that picture transforms are done on metadata too.
124+ + Small fixes to `nav.xhtml`:
125+ Add 'landmarks' id attribute to the landmarks nav.
126+ Replace old default CSS removing numbers from ol.toc li
127+ with new rules that match `nav#toc ol, nav#landmarks ol`.
128+ We keep the `toc` class on `ol` for backwards compatibility.
129+
130+ * LaTeX writer:
131+
132+ + Make raw content marked `beamer` pass through in `beamer`
133+ output (pandoc/lua-filters#40).
134+ + Beamer: avoid duplicated `fragile` property in some cases
135+ (#5208).
136+
137+ * RTF writer: use `toTableOfContents` from Shared to replace
138+ old duplicated code.
139+
140+ * Pptx writer: use `toTableOfContents` from Shared to replace
141+ old duplicated code.
142+
143+ * Custom writers:
144+
145+ + Allow '-' in filenames for custom lua writers (#5187).
146+ + sample.lua: add `SingleQuoted`, `DoubleQuoted` (#5104).
147+
148+ * reveal.js template: Add `zoomKey` config (#4249).
149+
150+ * HTML5 template: Remove unnecessary type="text/css" on style and
151+ link for HTML5 (#5146).
152+
153+ * LaTeX template (Andrew Dunning)
154+
155+ + Prevent fontspec from scaling `mainfont` to match the default
156+ font, Latin Modern. For example, a main font set to 12pt could
157+ previously appear between 11pt to 13pt depending on its design.
158+ To return to the earlier rendering, use `-V
159+ mainfontoptions="Scale=MatchLowercase"` (#5212, #5218).
160+ + Display monospaced fonts without TeX ligatures when using
161+ `--pdf-engine=lualatex`. It now matches the behaviour of other
162+ engines (#5212, #5218).
163+ + Remove the deprecated `romanfont` variable. The functionality of
164+ `mainfont` is identical (#5218).
165+ + Render `\subtitle` with the standard document classes.
166+ Previously, `subtitle` only appeared when using the KOMA-Script
167+ classes or Beamer (#5213).
168+ + Use Babel instead of Polyglossia for LuaLaTeX. This avoids
169+ several language selection problems, notably with retaining
170+ French spacing conventions when switching to a verbatim
171+ environment or another language; and in printing Greek text
172+ without hyphenation (#5193).
173+ + Use the [`xurl`](https://ctan.org/pkg/xurl) package if
174+ available, improving the appearance of URLs by allowing them to
175+ break at additional points. Requires TeX Live 2018 or later; run
176+ `tlmgr update --all` if you experience problems with URLs
177+ appearing after link text (#5193).
178+ + Use [`bookmark`](https://ctan.org/pkg/bookmark) if available,
179+ which can correct the levels of headings where `hyperref`
180+ cannot: see the [KOMA-Script 3.26 release
181+ notes](https://komascript.de/release3.26) (#5193).
182+ + Require the [`xcolor`](https://ctan.org/pkg/xcolor) package to
183+ avoid a possible error when using additional packages alongside
184+ footnotes in tables (#5193, closes #4861).
185+ + Remove obsolete `fixltx2e` package, which has no functionality
186+ with TeX Live 2015 or later (#5193).
187+ + Allow multiple `fontfamilies.options` (#5193, closes #5194).
188+ + Reindent file (#5193).
189+
190+ * LaTeX template: let KOMA document class handle parskip when
191+ applicable (#5143, Enno).
192+
193+ * Text.Pandoc.Readers: Changed types of `readJSON`; it now runs
194+ in an instance of PandocMonad, like the other readers and
195+ writers. [API change]
196+
197+ * Text.Pandoc.Writers: Changed types of `writeJSON`; it now runs
198+ in an instance of PandocMonad, like the other readers and
199+ writers. [API change]
200+
201+ * Text.Pandoc.Error: Added `PandocUTF8DecodingError` constructor
202+ for `PandocError`. [API change]
203+
204+ * Text.Pandoc.Writers.Shared - add `toTableOfContents`. [API change]
205+ This is refactored out from the Markdown writer.
206+ It can be used in other writers to create a table of contents.
207+
208+ * Improve error messages for UTF-8 decoding errors. Now we give
209+ the filename and byte offset (#4765).
210+
211+ * Text.Pandoc.XML: Strip out illegal XML characters in
212+ `escapeXMLString` (#5119).
213+
214+ * Text.Pandoc.Process: update `pipeProcess` (Albert Krewinkel).
215+ The implementation of `pipeProcess` was rewritten to fix sporadic
216+ failures caused by prematurely closed pipes.
217+
218+ * Use `safeRead` instead of `read` everywhere in the code
219+ (John MacFarlane, Mauro Bieg, #5162, #5180).
220+
221+ * Text.Pandoc.SelfContained: Decompress `.svgz` when
222+ converting to `data:` URI (#5163, Alexander Krotov).
223+
224+ * Text.Pandoc.Parsing: Remove unused `HasHeaderMap`
225+ (#5175, Alexander Krotov). [API change]
226+
227+ * Simplify/fix reading of `--metadata` values on command line
228+ to avoid problems relating to hvr/HsYAML#7 (#5177).
229+
230+ * Removed custom Setup.hs, use build-type: simple. The only thing we
231+ gained from the custom build was automatic installation of the man
232+ page when using 'cabal install'. But custom builds cause problems,
233+ e.g., with cross-compilation. Installation of the man page is
234+ better handled by packagers. Note to packagers (e.g. Debian): it
235+ may be necessary to add a step installing the man page with the
236+ next release.
237+
238+ * Allow latest http-client, tasty, zip-archive, Glob.
239+
240+ * Require skylighting >= 0.7.5, adding support for sml, J,
241+ typescript.
242+
243+ * Tests: Cleaned up `findPandoc` in `Tests.Helpers`, so it
244+ works well with cabal v2.
245+
246+ * INSTALL.md:
247+
248+ + Use button for installer links (John MacFarlane, Mauro Bieg,
249+ #5167).
250+ + Fix links and bump required stack version (max).
251+ + Improve installation notes on associated software (Andrew
252+ Dunning). Includes explanation of how to install related tools
253+ with package managers (since the method of installing
254+ `rsvg-convert` is not obvious).
255+
256+ * doc/org.md: improve documentation of org features (Albert
257+ Krewinkel).
258+
259+ * MANUAL:
260+
261+ + Clarify automatic identifiers (#5201). We remove
262+ non-alphanumerics. This includes, e.g., emojis.
263+ + Fix example for Div with id (Geoffrey Ely).
264+ + Update list of LaTeX packages used.
265+ + Make it clear that `hard_line_breaks` works in `gfm` (see #5195).
266+
267+ * trypandoc: Fix CSS and viewport.
268+
1269pandoc (2.5)
2270
3271 * Text.Pandoc.App: split into several unexported submodules (Albert