| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Previously, when multiple file arguments were provided, pandoc
simply concatenated them and passed the contents to the readers,
which took a Text argument.
As a result, the readers had no way of knowing which file
was the source of any particular bit of text. This meant that
we couldn't report accurate source positions on errors or
include accurate source positions as attributes in the AST.
More seriously, it meant that we couldn't resolve resource
paths relative to the files containing them
(see e.g. #5501, #6632, #6384, #3752).
Add Text.Pandoc.Sources (exported module), with a `Sources` type
and a `ToSources` class. A `Sources` wraps a list of `(SourcePos,
Text)` pairs. [API change] A parsec `Stream` instance is provided for
`Sources`. The module also exports versions of parsec's `satisfy` and
other Char parsers that track source positions accurately from a
`Sources` stream (or any instance of the new `UpdateSourcePos` class).
Text.Pandoc.Parsing now exports these modified Char parsers instead of
the ones parsec provides. Modified parsers to use a `Sources` as stream
[API change].
The readers that previously took a `Text` argument have been
modified to take any instance of `ToSources`. So, they may still
be used with a `Text`, but they can also be used with a `Sources`
object.
In Text.Pandoc.Error, modified the constructor PandocParsecError
to take a `Sources` rather than a `Text` as first argument,
so parse error locations can be accurately reported.
T.P.Error: showPos, do not print "-" as source name.
|
|
|
|
This should speed-up recompilation after changes in `Text.Pandoc.Class`,
as the number of modules affected by a change will be smaller in
general. It also offers faster insights into the parts of `T.P.Class`
used within a module.
|
|
* Use implicit Prelude
The previous behavior was introduced as a fix for #4464. It seems that
this change alone did not fix the issue, and `stack ghci` and `cabal
repl` only work with GHC 8.4.1 or newer, as no custom Prelude is loaded
for these versions. Given this, it seems cleaner to revert to the
implicit Prelude.
* PandocMonad: remove outdated check for base version
Only base versions 4.9 and later are supported, the check for
`MIN_VERSION_base(4,8,0)` is therefore unnecessary.
* Always use custom prelude
Previously, the custom prelude was used only with older GHC versions, as
a workaround for problems with ghci. The ghci problems are resolved by
replacing package `base` with `base-noprelude`, allowing for consistent
use of the custom prelude across all GHC versions.
|
|
* Update copyright year
* Copyright: add notes for Lua and Jira modules
|
|
Errors during parsing are now returned in full and no longer replaced by
a custom message.
|
|
PR #5884.
+ Use pandoc-types 1.20 and texmath 0.12.
+ Text is now used instead of String, with a few exceptions.
+ In the MediaBag module, some of the types using Strings
were switched to use FilePath instead (not Text).
+ In the Parsing module, new parsers `manyChar`, `many1Char`,
`manyTillChar`, `many1TillChar`, `many1Till`, `manyUntil`,
`mantyUntilChar` have been added: these are like their
unsuffixed counterparts but pack some or all of their output.
+ `glob` in Text.Pandoc.Class still takes String since it seems
to be intended as an interface to Glob, which uses strings.
It seems to be used only once in the package, in the EPUB writer,
so that is not hard to change.
|
|
The haddock module header contains essentially the
same information, so the boilerplate is redundant and
just one more thing to get out of sync.
|
|
Quite a few modules were missing copyright notices.
This commit adds copyright notices everywhere via haddock module
headers. The old license boilerplate comment is redundant with this and has
been removed.
Update copyright years to 2019.
Closes #4592.
|
|
This seems to be necessary if we are to use our custom Prelude
with ghci.
Closes #4464.
|
|
|
|
The readers previously assumed that CRs had been filtered
from the input. Now we strip the CRs in the readers themselves,
before parsing. (The point of this is just to simplify the
parsers.)
Shared now exports a new function `crFilter`. [API change]
And `tabFilter` no longer filters CRs.
|
|
Readers: Renamed StringReader -> TextReader.
Updated tests.
API change.
|
|
This follows the suggestions given by the FSF for GPL licensed software.
<https://www.gnu.org/prep/maintain/html_node/Copyright-Notices.html>
|
|
As noted in the previous commit, an autogenerated identifier
may still coincide with an explicit identifier that is given
for a header later in the document, or with an identifier on
a div, span, link, or image. This commit adds a warning
in this case, so users can supply an explicit identifier.
* Added `DuplicateIdentifier` to LogMessage.
* Modified HTML, Org, MediaWiki readers so their custom
state type is an instance of HasLogMessages. This is necessary
for `registerHeader` to issue warnings.
See #1745.
|
|
|
|
|
|
|
|
Comment trees were handled after parsing, as pattern matching on lists
is easier than matching on sequences. The new method of reading
documents as trees allows for more elegant subtree removal.
|
|
Reading of smart quotes can be toggled using the `'` option.
|
|
Block parsing code is moved to a separate module.
This is part of the Org-mode reader cleanup effort.
|
|
Inline parsing code is moved to a separate module. Parsers for block
starts are extracted as well, as those are used in the `endline` parser.
This is part of the Org-mode reader cleanup effort.
|
|
The Org-mode reader uses many functions defined in the
`Text.Pandoc.Parsing` utility module. Some of the functions are
overwritten with versions adapted to Org-mode idiosyncrasies. These
special functions, as well as the normal Pandoc versions, are combined
in a single module to increase the ease of use.
This leads to decoupling of Org-mode and Pandoc and hence to slightly
cleaner code. The downside is code-bloat due to repeated import/export
statements.
|
|
The `d` export option can be used to control which drawers are exported
and which are discarded. Basic support for this option is added here.
|
|
The `ID` property is reserved for internal use by Org-mode and should
not be used. The `CUSTOM_ID` property is to be used instead, it is
converted to the `ID` property for certain export format.
The reader and writer erroneously used `ID`. This is corrected by using
`CUSTOM_ID` where appropriate.
|
|
Headers can have optional `:PROPERTIES:` drawers associated with them.
These drawers contain key/value pairs like the header's `id`. The
reader adds all listed pairs to the header's attributes; `id` and
`class` attributes are handled specially to match the way `Attr` are
defined.
This also changes behavior of how drawers of unknown type are handled.
Instead of including all unknown drawers, those are not read/exported,
thereby matching current Emacs behavior.
This closes #1877.
|
|
Arbitrary key-value pairs can be added to some block types using a
`#+ATTR_HTML` line before the block. Emacs Org-mode only includes these
when exporting to HTML, but since we cannot make this distinction here,
the attributes are always added.
The functionality is now supported for figures.
This closes #1906.
|
|
Additional state changes need to be made after a newline is parsed,
otherwise markup may not be recognized correctly.
This fixes a bug where markup after certain block-types would not be
recognized. E.g. `/emph/` in the following snippet was not parsed as
emphasized.
foo
# comment
/emph/
|
|
A parser state attribute was used to keep track of block attributes
defined in meta-lines. Global state is undesirable, so block attributes
are no longer saved as part of the parser state. Old functions and the
respective part of the parser state are removed.
|
|
All known export options are parsed but ignored.
|
|
Org-mode allows to specify export settings via `#+OPTIONS` lines.
Disabling simple sub- and superscripts is one of these export options,
this options is now supported.
|
|
The org reader code has become large and confusing. Extracting smaller
parts into submodules should help to clean things up.
|
|
The last fix for whitespace handling of inline LaTeX commands was
incorrect, preventing correct recognition of inline LaTeX commands which
contain spaces. This fix ensures that only trailing whitespace is cut
off.
|
|
Org reader: table parsing code refactoring and fixes
|
|
The org-reader was droping space after unescaped LaTeX-style symbol
commands: `\ForAll \Auml` resulted in `∀Ä` but should give `∀ Ä`
instead. This seems to be because the LaTeX-reader treats the
command-terminating space as part of the command. Dropping the trailing
space from the symbol-command fixes this issue.
|
|
This fixes Org mode parsing of some corner cases regarding empty cells
and rows. Empty cells weren't parsed correctly, e.g. `|||` should be
two empty cells, but would be parsed as a single cell containing a pipe
character. Empty rows where parsed as alignment rows and dropped from
the output.
This fixes #2616.
|
|
This refactores the codes conversing a list table lines to an org table
ADT. The old code was simplified and is now slightly less ugly.
|
|
Emacs Org-mode doesn't add any padding to table rows. The first
row (header or first body row) is used to determine the column count, no
other magic is performed.
The org reader was padding rows to the length of the longest table row.
This was done due to a misunderstanding of how Org handles tables. This
feature reflected how Org-mode handles tables when pressing <TAB>. The
Org exporter however, which is what the reader should implement, doesn't
do any of this. So this was a mis-feature that made the reader more
complex and reduced comparability. It was hence removed.
|
|
Fixes #2862
Also fix up tab handling for leading whitespace in code blocks.
|
|
Prefix even empty figure names with "fig:"
|
|
Cleanup of the code for link target handling. Most notably, the
canonicalization of a link is handled by a separate function.
This fixes #2684.
|
|
This avoids performance problems in documents with many identically
named headers.
Closes #2671.
|
|
The convention used by pandoc for figures is to mark them by prefixing
the name with "fig:". The org reader failed to do this if a figure had
no name. The test for this was broken as well.
This fixes #2643.
|
|
Continue scanning for comment subtrees beyond only the first block.
Note to self: when writing an recursive function, don't forget to, you
know, actually recurse.
Shout to @mrvdb for noticing this.
This fixes #2628.
|
|
|
|
Org reader: Require whitespace around def list markers
|
|
Definition list markers (i.e. double colons `::`) must be surrounded by
whitespace to start a definition item. This rule was not checked
before, resulting in bugs with footnotes and some link types.
Thanks to @conklech for noticing and reporting this issue.
This fixes #2518.
|
|
Smart quotes, ellipses, and dashes should behave like normal quotes,
single dashes, and dots with respect to text markup parsing. The parser
state was not updated properly in all cases, which has been fixed.
Thanks to @conklech for reporting this issue.
This fixes #2513.
|
|
Don't use custom prelude for latest ghc.
This is a better approach to making 'stack ghci' and 'cabal repl'
work. Instead of using NoImplicitPrelude, we only use the custom
prelude for older ghc versions. The custom prelude presents a
uniform API that matches the current base version's prelude.
So, when developing (presumably with latest ghc), we don't
use a custom prelude at all and hence have no trouble with ghci.
The custom prelude no longer exports (<>): we now want to
match the base 4.8 prelude behavior.
|
|
This reverts commit c423dbb5a34c2d1195020e0f0ca3aae883d0749b.
|
