@@ -1,3 +1,141 @@
1+pandoc (2.6.1) PROVISIONAL
2+
3+ * `asciidoctor` is now an output format separate from `asciidoc`,
4+ to accommodate some minor implementation-specific differences
5+ (currently just in the treatment of display math).
6+
7+ * Markdown reader:
8+
9+ + Improve tight/loose list handling (#5285). Previously the algorithm
10+ allowed list items with a mix of Para and Plain, which is never wanted.
11+ This patch modifies `compactify` in Text.Pandoc.Shared so that, if a
12+ list's items contain (at the top level) Para elements (aside from perhaps
13+ at the very end), *all* Plains are converted to Paras.
14+ + Add newline when parsing blocks in YAML (#5271).
15+ Otherwise last block gets parsed as a Plain rather than
16+ a Para. This is a regression in pandoc 2.x. This patch restores
17+ pandoc 1.19 behavior.
18+
19+ * LaTeX reader: don't let `\egroup` match `{`. `braced` now actually
20+ requires nested braces. Otherwise some legitimate command and environment
21+ definitions can break.
22+
23+ * Docx reader (Jesse Rosenthal):
24+
25+ + Rename `getDocumentPath` as `getDocumentXmlPath`.
26+ + Use field notation for setting `ReaderEnv`.
27+ + Figure out `document.xml` path once at the beginning of parsing, and
28+ add it to the environment, so we can avoid repeated lookups.
29+ + Dynamically determine main document xml path (#5277).
30+ The desktop Word program places the main document file in
31+ `word/document.xml`, but the online word places it in
32+ `word/document2.xml`. This file path is actually stated in the root
33+ `_rels/.rels` file, in the `Relationship` element with an
34+ `http://../officedocument` type.
35+ + Fix paths in archive to prevent Windows failure (#5277).
36+ Some paths in archives are absolute (have an opening slash) which, for
37+ reasons unknown, produces a failure in the test suite on MS
38+ Windows. This fixes that by removing the leading slash if it exists.
39+
40+ * Muse reader: test that block level markup does not break `<verbatim>`
41+ (Alexander Krotov).
42+
43+ * ipynb reader: handle images referring to attachments.
44+ Previously we didn't strip off the attachment: prefix,
45+ so even though the attachment is available in the mediabag,
46+ pandoc couldn't find it.
47+
48+ * AsciiDoc Writer: Add `writeAsciiDoctor` [API change, Tarik Graba].
49+ Handle display math appropriately for Asciidoctor.
50+
51+ * JATS writer: wrap figure caption in `<p>` to fix validation (#5290,
52+ Mauro Bieg).
53+
54+ * HTML writer: implement WAI-ARIA roles for (end)notes (#4213).
55+
56+ * ipynb writer - ensure final newline.
57+
58+ * Markdown writer: use `markdown="1"` when appropriate for Divs:
59+ when `native_divs` and `markdown_in_html_blocks` are disabled
60+ but `raw_html` and `markdown_attribute` are enabled.
61+
62+ * LaTeX writer:
63+
64+ + Use right fold for `escapeString`. This is more elegant than
65+ the explicit recursive code we were using.
66+ + Avoid `{}` after control sequences when escaping.
67+ `\ldots{}.` doesn't behave as well as `\ldots.` with the latex
68+ ellipsis package. This patch causes pandoc to avoid emitting
69+ the `{}` when it is not necessary. Now `\ldots` and other
70+ control sequences used in escaping will be followed by either
71+ a `{}`, a space, or nothing, depending on context.
72+
73+ * Lua filters (Albert Krewinkel):
74+
75+ + load module `pandoc` before calling `init.lua` (#5287).
76+ The file `init.lua` in pandoc's data directory is run as part of pandoc's
77+ Lua initialization process. Previously, the `pandoc` module was loaded in
78+ `init.lua`, and the structure for marshaling was set-up after. This
79+ allowed simple patching of element marshaling, but made using `init.lua`
80+ more difficult. Now, all required modules are now loaded before calling
81+ `init.lua`. The file can be used entirely for user customization.
82+ Patching marshaling functions, while discouraged, is still possible via
83+ the `debug` module.
84+
85+ + Re-export all bundled modules (Albert Krewinkel).
86+ All Lua modules bundled with pandoc, i.e., `pandoc.List`,
87+ `pandoc.mediabag`, `pandoc.utils`, and `text` are re-exported from the
88+ `pandoc` module. They are assigned to the fields `List`, `mediabag`,
89+ `utils`, and `text`, respectively.
90+
91+ * latex template: support the `subject` metadata variable (#5289,
92+ Pascal Wagler).
93+
94+ * epub3 template: Add titlepage class to section (#5269).
95+
96+ * HTML5 template: Add role with ARIA doc-toc for table of contents (#4213).
97+
98+ * Make --metadata-file use pandoc-markdown (#5279, #5272, Mauro Bieg).
99+
100+ * Text.Pandoc.Extensions:
101+
102+ + More carefully groom ipynb default extensions.
103+ + Add `all_symbols_escapable` to `githubMarkdownExtensions`.
104+
105+ * Text.Pandoc.PDF: Use system temp directory when possible (#1192).
106+ Previously we created temp dirs in the working directory,
107+ partly (a) because there were problems using the system temp
108+ directory on Windows, when their pathnames included tildes,
109+ and partly (b) because programs like `epstopdf.pl` would not
110+ be allowed to write to directories outside the working
111+ directory in restricted mode. We now (a) use the system
112+ temp dir except when the path includes tildes, and
113+ (b) set TEXMFOUTPUT when creating the PDF, so that
114+ subsidiary programs can use the system temp directory.
115+ This addresses problems that occurred when pandoc
116+ was used in a synced directory.
117+
118+ * Text.Pandoc.MIME: add WebP (#5267, Mauro Bieg).
119+
120+ * Tests: avoid calling `findPandoc` multiple times.
121+
122+ * Old tests: remove need for temp files by using `pipeProcess`.
123+
124+ * Added simple ipynb reader/writer tests (#5274).
125+
126+ * Use latest skylighting (adds support for `protobuf`,
127+ updates definitions for other languages).
128+
129+ * trypandoc: Add JATS to output option (Arfon Smith, #5291).
130+
131+ * Use latest basement/foundation on 32bit windows.
132+
133+ * Add missing copyright notices and remove license boilerplate (#5112,
134+ Albert Krewinkel). Quite a few modules were missing copyright notices.
135+ This commit adds copyright notices everywhere via haddock module
136+ headers. The old license boilerplate comment is redundant with this and has
137+ been removed. Update copyright years to 2019. Closes #4592.
138+
1139pandoc (2.6)
2140
3141 * Support ipynb (Jupyter notebook) as input and output format.