Add support for multiple abstract-like sections

Makefile

@@ -3,7 +3,7 @@ DIFF ?= diff
 PANDOC ?= pandoc
 
 .PHONY: test
-test: test-hrule test-in-div
+test: test-hrule test-in-div test-two-sections
 
 test-%: $(FILTER_FILE) test/input-%.md test/test-%.yaml
 	@$(PANDOC) --defaults test/test-$*.yaml | \

README.md

@@ -113,6 +113,34 @@ output:
 ---
 ```
 
+Multiple abstracts
+------------------------------------------------------------------
+
+It is common for certain works to include two abstracts, one in
+English and one in the local language of an academic institution.
+E.g., theses published at many German universities must have a
+German "Zusammenfassung" in addition to the English "Abstract".
+
+The filter can be configured to support those additional
+abstract-like sections as well. The identifiers of the sections
+must be listed in the `section-identifiers` field below the
+`abstract-section` metadata entry. E.g.:
+
+``` yaml
+---
+abstract-section:
+  section-identifiers:
+    - abstract
+    - sammanfattning
+---
+```
+
+This will place the *Abstract* in the `abstract` variable as
+before, but will also collect the contents of a *Sammanfattning*
+section and place it in the `sammanfattning` field, where it can
+be used for further processing, e.g., with a custom template.
+
+
 License
 ------------------------------------------------------------------
 

_extensions/abstract-section/abstract-section.lua

@@ -1,29 +1,35 @@
 --[[
-abstract-to-meta – move an "abstract" section into document metadata
+abstract-section – move an "abstract" section into document metadata
 
-Copyright: © 2017–2021 Albert Krewinkel
+Copyright: © 2017–2023 Albert Krewinkel
 License:   MIT – see LICENSE file for details
 ]]
-local abstract = {}
+local stringify = (require 'pandoc.utils').stringify
+local section_identifiers = {
+  abstract = true,
+}
+local collected = {}
 
 --- Extract abstract from a list of blocks.
-function abstract_from_blocklist (blocks)
+local function abstract_from_blocklist (blocks)
   local body_blocks = {}
-  local looking_at_abstract = false
+  local looking_at_section = false
 
   for _, block in ipairs(blocks) do
     if block.t == 'Header' and block.level == 1 then
-      if block.identifier == 'abstract' then
-        looking_at_abstract = true
+      if section_identifiers[block.identifier] then
+        looking_at_section = block.identifier
+        collected[looking_at_section] = {}
       else
-        looking_at_abstract = false
+        looking_at_section = false
         body_blocks[#body_blocks + 1] = block
       end
-    elseif looking_at_abstract then
+    elseif looking_at_section then
       if block.t == 'HorizontalRule' then
-        looking_at_abstract = false
+        looking_at_section = false
       else
-        abstract[#abstract + 1] = block
+        local collect = collected[looking_at_section]
+        collect[#collect + 1] = block
       end
     else
       body_blocks[#body_blocks + 1] = block
@@ -33,27 +39,42 @@ function abstract_from_blocklist (blocks)
   return body_blocks
 end
 
-if PANDOC_VERSION >= {2,9,2} then
-  -- Check all block lists with pandoc 2.9.2 or later
-  return {{
-      Blocks = abstract_from_blocklist,
-      Meta = function (meta)
-        if not meta.abstract and #abstract > 0 then
-          meta.abstract = pandoc.MetaBlocks(abstract)
-        end
-        return meta
-      end
-  }}
-else
-  -- otherwise, just check the top-level block-list
-  return {{
-      Pandoc = function (doc)
-        local meta = doc.meta
-        local other_blocks = abstract_from_blocklist(doc.blocks)
-        if not meta.abstract and #abstract > 0 then
-          meta.abstract = pandoc.MetaBlocks(abstract)
-        end
-        return pandoc.Pandoc(other_blocks, meta)
-      end,
-  }}
+Pandoc = function (doc)
+  local meta = doc.meta
+
+  -- configure
+  section_identifiers_list =
+    (doc.meta['abstract-section'] or {})['section-identifiers']
+  if section_identifiers_list and #section_identifiers_list > 0 then
+    section_identifiers = {}
+    for i, ident in ipairs(section_identifiers_list) do
+      section_identifiers[stringify(ident)] = true
+    end
+  end
+  -- unset config in meta
+  doc.meta['abstract-section'] = nil
+
+  local blocks = {}
+  if PANDOC_VERSION >= {2,17} then
+    -- Walk all block lists by default
+    blocks = doc.blocks:walk{Blocks = abstract_from_blocklist}
+  elseif PANDOC_VERSION >= {2,9,2} then
+    -- Do the same with pandoc versions that don't have walk methods but the
+    -- `walk_block` function.
+    blocks = pandoc.utils.walk_block(
+      pandoc.Div(doc.blocks),
+      {Blocks = abstract_from_blocklist}
+    ).content
+  else
+    -- otherwise, just check the top-level block-list
+    blocks = abstract_from_blocklist(doc.blocks)
+  end
+  for metakey in pairs(section_identifiers) do
+    metakey = stringify(metakey)
+    local abstract = collected[metakey]
+    if not meta[metakey] and abstract and #abstract > 0 then
+      meta[metakey] = pandoc.MetaBlocks(abstract)
+    end
+  end
+  return pandoc.Pandoc(blocks, meta)
 end

test/expected-two-sections.md

@@ -0,0 +1,29 @@
+---
+abstract: |
+  Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do
+  eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad
+  minim veniam, quis nostrud exercitation ullamco laboris nisi ut
+  aliquip ex ea commodo consequat. Duis aute irure dolor in
+  reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla
+  pariatur.
+
+  -   one
+  -   two
+  -   three
+
+  Excepteur sint occaecat cupidatat non proident, sunt in culpa qui
+  officia deserunt mollit anim id est laborum.
+sammanfattning: |
+  -   Phasellus purus.
+  -   Praesent fermentum tempor tellus.
+  -   Proin quam nisl, tincidunt et, mattis eget, convallis nec, purus.
+  -   Lorem ipsum dolor sit amet, consectetuer adipiscing elit.
+  -   Curabitur lacinia pulvinar nibh.
+---
+
+# Lorem Ipsum
+
+Quo dolore molestiae et laboriosam occaecati explicabo corrupti. Earum
+expedita ducimus quaerat est quam ut molestiae. Illum deleniti vel
+labore facilis et cum est. Est nemo est vel ad. Assumenda consequatur
+rerum officiis atque officia. Est nihil iste cumque ad qui.

test/input-two-sections.md

@@ -0,0 +1,37 @@
+---
+abstract-section:
+  section-identifiers:
+    - abstract
+    - sammanfattning
+---
+
+# Abstract
+
+Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor
+incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis
+nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat.
+Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu
+fugiat nulla pariatur.
+
+- one
+- two
+- three
+
+Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia
+deserunt mollit anim id est laborum.
+
+# Sammanfattning {#sammanfattning}
+
+* Phasellus purus.
+* Praesent fermentum tempor tellus.
+* Proin quam nisl, tincidunt et, mattis eget, convallis nec, purus.
+* Lorem ipsum dolor sit amet, consectetuer adipiscing elit.
+* Curabitur lacinia pulvinar nibh.
+
+# Lorem Ipsum
+
+Quo dolore molestiae et laboriosam occaecati explicabo corrupti.
+Earum expedita ducimus quaerat est quam ut molestiae. Illum
+deleniti vel labore facilis et cum est. Est nemo est vel ad.
+Assumenda consequatur rerum officiis atque officia. Est nihil iste
+cumque ad qui.

test/test-two-sections.yaml

@@ -0,0 +1,4 @@
+input-files: [test/input-two-sections.md]
+standalone: true
+filters: [{type: lua, path: abstract-section.lua}]
+to: markdown