From 1128c8fb69d4a8557baeee13053e7792339554f9 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Na=C3=AFm=20Favier?= <n@monade.li>
Date: Mon, 3 Jul 2023 22:03:00 +0200
Subject: [PATCH] Bundle pandoc filters (#670)

Continuation of https://github.com/NixOS/nixos-search/pull/669
---
 .../citerefentry-to-rst-role.lua              | 23 ++++++++++++
 flake-info/assets/data/myst-reader/roles.lua  | 36 +++++++++++++++++++
 flake-info/default.nix                        |  1 -
 flake-info/src/data/pandoc.rs                 |  9 ++---
 4 files changed, 62 insertions(+), 7 deletions(-)
 create mode 100644 flake-info/assets/data/docbook-reader/citerefentry-to-rst-role.lua
 create mode 100644 flake-info/assets/data/myst-reader/roles.lua

diff --git a/flake-info/assets/data/docbook-reader/citerefentry-to-rst-role.lua b/flake-info/assets/data/docbook-reader/citerefentry-to-rst-role.lua
new file mode 100644
index 0000000..281e85a
--- /dev/null
+++ b/flake-info/assets/data/docbook-reader/citerefentry-to-rst-role.lua
@@ -0,0 +1,23 @@
+--[[
+Converts Code AST nodes produced by pandoc’s DocBook reader
+from citerefentry elements into AST for corresponding role
+for reStructuredText.
+
+We use subset of MyST syntax (CommonMark with features from rST)
+so let’s use the rST AST for rST features.
+
+Reference: https://www.sphinx-doc.org/en/master/usage/restructuredtext/roles.html#role-manpage
+]]
+
+function Code(elem)
+  elem.classes = elem.classes:map(function (x)
+    if x == 'citerefentry' then
+      elem.attributes['role'] = 'manpage'
+      return 'interpreted-text'
+    else
+      return x
+    end
+  end)
+
+  return elem
+end
diff --git a/flake-info/assets/data/myst-reader/roles.lua b/flake-info/assets/data/myst-reader/roles.lua
new file mode 100644
index 0000000..f4ef6d3
--- /dev/null
+++ b/flake-info/assets/data/myst-reader/roles.lua
@@ -0,0 +1,36 @@
+--[[
+Replaces Str AST nodes containing {role}, followed by a Code node
+by a Code node with attrs that would be produced by rST reader
+from the role syntax.
+
+This is to emulate MyST syntax in Pandoc.
+(MyST is a CommonMark flavour with rST features mixed in.)
+
+Reference: https://myst-parser.readthedocs.io/en/latest/syntax/syntax.html#roles-an-in-line-extension-point
+]]
+
+function Inlines(inlines)
+  for i = #inlines-1,1,-1 do
+    local first = inlines[i]
+    local second = inlines[i+1]
+    local correct_tags = first.tag == 'Str' and second.tag == 'Code'
+    if correct_tags then
+      -- docutils supports alphanumeric strings separated by [-._:]
+      -- We are slightly more liberal for simplicity.
+      -- Allow preceding punctuation (eg '('), otherwise '({file}`...`)'
+      -- does not match. Also allow anything followed by a non-breaking space
+      -- since pandoc emits those after certain abbreviations (e.g. e.g.).
+      local prefix, role = first.text:match('^(.*){([-._+:%w]+)}$')
+      if role ~= nil and (prefix == '' or prefix:match("^.*[%p ]$") ~= nil) then
+        if prefix == '' then
+          inlines:remove(i)
+        else
+          first.text = prefix
+        end
+        second.attributes['role'] = role
+        second.classes:insert('interpreted-text')
+      end
+    end
+  end
+  return inlines
+end
diff --git a/flake-info/default.nix b/flake-info/default.nix
index c93e010..6d6461a 100644
--- a/flake-info/default.nix
+++ b/flake-info/default.nix
@@ -23,7 +23,6 @@ pkgs.rustPlatform.buildRustPackage rec {
   checkInputs = with pkgs; [ pandoc ];
 
   ROOTDIR = builtins.placeholder "out";
-  NIXPKGS_PANDOC_FILTERS_PATH = pkgs.path + "/doc/build-aux/pandoc-filters";
   LINK_MANPAGES_PANDOC_FILTER = import src/data/link-manpages.nix { inherit pkgs; };
 
   checkFlags = [
diff --git a/flake-info/src/data/pandoc.rs b/flake-info/src/data/pandoc.rs
index d533788..c43f315 100644
--- a/flake-info/src/data/pandoc.rs
+++ b/flake-info/src/data/pandoc.rs
@@ -1,15 +1,12 @@
 use lazy_static::lazy_static;
-use std::path::{Path, PathBuf};
+use std::path::PathBuf;
 
 use pandoc::*;
 
-const FILTERS_PATH: &str = env!("NIXPKGS_PANDOC_FILTERS_PATH");
-
 lazy_static! {
     static ref DOCBOOK_ROLES_FILTER: PathBuf =
-        Path::new(FILTERS_PATH).join("docbook-reader/citerefentry-to-rst-role.lua");
-    static ref MARKDOWN_ROLES_FILTER: PathBuf =
-        Path::new(FILTERS_PATH).join("myst-reader/roles.lua");
+        crate::DATADIR.join("data/docbook-reader/citerefentry-to-rst-role.lua");
+    static ref MARKDOWN_ROLES_FILTER: PathBuf = crate::DATADIR.join("data/myst-reader/roles.lua");
     static ref MANPAGE_LINK_FILTER: PathBuf = PathBuf::from(env!("LINK_MANPAGES_PANDOC_FILTER"));
     static ref XREF_FILTER: PathBuf = crate::DATADIR.join("data/fix-xrefs.lua");
 }