pandoc - Conversion between markup formats

Age	Commit message (Collapse)	Author	Files	Lines
2018-01-15	Powerpoint writer: Import reference-doc images properly.	Jesse Rosenthal	1	-2/+2
	There was a glob error that was leading to images from the reference-doc pptx not being imported. We don't need a glob here -- just replace it with `isPrefixOf`.
2018-01-15	Powerpoint writer: Move Presentation.hs out of PandocMonad	Jesse Rosenthal	1	-33/+36
	We don't need it for anything but the log messages, and we can just keep track of that in state and pass it along to the `writePowerpoint` function. This will simplify the code.
2018-01-15	Powerpoint writer: Ignore anchor links to nowehere.	Jesse Rosenthal	1	-3/+6
	We don't convert a '#target' ExternalTarget to an InternalTarget if `target` is not in the AnchorMap. We just remove the link. This prevents broken links in the Powerpoint output.
2018-01-14	Powerpoint writer: Fix anchor links.	Jesse Rosenthal	2	-60/+81
	They were broken when I refactored (the Output module wanted to use state left over from the construction of the Presentation type). This change introduces a new type `LinkTarget = InternalTarget \| ExternalTarget`. Internal target points to a slide number, and these will all be resolved before the Presentation is passed along to the Output module.
2018-01-14	Powerpoint writer: Refactor into separate modules.	Jesse Rosenthal	2	-0/+2132
	There are two steps in the conversion: a conversion from pandoc to a Presentation datatype modeling pptx, and a conversion from Presentation to a pptx archive. The two steps were sharing the same state and environment, and the code was getting a bit spaghetti-ish. This separates the conversion into separate modules (T.P.W.Powerpoint.Presentation, which defineds the Presentation datatype and goes Pandoc->Presentation) and (T.P.W.Pandoc.Output, which goes Presentation->Archive). Text.Pandoc.Writers.Powerpoint a thin wrapper around the two modules.