summaryrefslogtreecommitdiffstats
path: root/notes
diff options
context:
space:
mode:
authorGustav Eek <gustav.eek@fripost.org>2019-12-14 19:48:24 +0100
committerGustav Eek <gustav.eek@fripost.org>2020-01-06 16:25:10 +0100
commit9b611caf3208ea05988f4d0bdeacac9eb0c0e59d (patch)
treea52d2e293262d968125b175d4d1bf7f8d7b1e5c9 /notes
parent5dcfb79dd6b229411d4ef35ff88cb898bf412c3f (diff)
Feature. Plugin. Provide sets of meta keys for inclusion
Add further lists of meta keys to include (*hash_meta_keys* and *list_hash_meta_keys*), add them to the *list_meta* map in subroutine *htmlize* and provide them to *pagestate* *meta*. This makes available, e.g. the *reference* attribute in page templates. The commit require some explanation. Until now, meta attributes were available as * Bolean meta (*bool_meta*) * Scalar meta (*scalar_meta*) * List meta (*list_meta*) All meta attributes in documents are provided as page template variables with a 'pandoc_'-prefix. On top of that, attributes listed in *scalar_meta_keys* are made available as regular template variables, without prefix. Lists are problematic. Pure lists are pushed with 'pandoc_' prefix, but they can not be used. The alternative would be via loops (see <https://metacpan.org/pod/HTML::Template#TMPL_LOOP>), but inside loops one need to call for a variable and the pushed attribute does not provide one. This commit properly makes available the *reference* attribute, which is of type list of associated list. This construct actually fully natural fulfil the limitations of lists. Template loops generate the items, whose values are available with keys as variable names: --- references: - id: refone title: Ref-MetaInlines-in-MetaList - id: reftwo title: Ref-MetaInlines-in-MetaList ... Which is called like this: <TMPL_IF REFERENCES> <TMPL_LOOP REFERENCES> <TMPL_VAR ID> <TMPL_VAR TITLE> </TMPL_LOOP> </TMPL_IF> How is hacky. The items in *hash_meta_keys* and *list_hash_meta_keys* are added to *list_meta_keys* and undergo the list processing. The list processing contradictory adopted better for *list_hash_meta_keys* than for *list_meta_keys*.
Diffstat (limited to 'notes')
0 files changed, 0 insertions, 0 deletions