Fixes https://github.com/kaushalmodi/ox-hugo/issues/30.
- Locate correct subtree to be exported
- Made CUSTOM_ID references unique
- Updated .md files with unique CUSTOM_ID references
- Workaround to prevent exporting of empty special blocks
- Mark preprocessed buffer as unmodified
- Support crosspost custom-id, id and fuzzy link types
- Update test site with cross-post links
- Require 'org-id in setup-ox-hugo.el to handle ID links
- Add instruction to handle ID links to test site
- Kill the preprocessed buffer after exporting
- Fix check for non-nil destination-filename
- Support preprocessing of cross-post "fuzzy" and "id" type links
- Support section paths in the preprocessing buffer
- Retrieve #+HUGO_SECTION keyword from in-buffer settings
Seems like Pandoc Markdown writer has some changes in how the
citations are written. The reference output in this commit is
generated using Pandoc 2.5.
Pandoc introduced its citations related meta-data like bibliography
and nocite which I did not prefer for few reasons:
1. It leaks full paths of the user's bibliography files in public.
2. Those meta-data fields are not needed by Hugo.
3. I like my style of YAML list fields better, it's more concise, and
consistent with the YAML front-matter of non-Pandoc-Citation posts.
Kaushal Modi
Folkert van der Beek