Fix citeproc regression with duplicate references.

- Use dev version of citeproc, which handles duplicate
  ids better, preferring the last one in the list
  and discarding the rest.
- Ensure that inline citations take priority over external
  ones.

See jgm/citeproc#36.

This restores the behavior of pandoc-citeproc.

1 files changed, 2 insertions, 1 deletions

diff --git a/src/Text/Pandoc/Citeproc.hs b/src/Text/Pandoc/Citeproc.hs
index 770d571a6..ca1ab9f96 100644
--- a/src/Text/Pandoc/Citeproc.hs
+++ b/src/Text/Pandoc/Citeproc.hs
@@ -119,7 +119,8 @@ processCitations (Pandoc meta bs) = do
                         Nothing -> return []
                     Nothing -> return []
   let refs = map (linkifyVariables . legacyDateRanges)
-                 (inlineRefs ++ externalRefs)
+                 (externalRefs ++ inlineRefs)
+                 -- note that inlineRefs can override externalRefs
   let otherIdsMap = foldr (\ref m ->
                              case T.words . extractText <$>
                                   M.lookup "other-ids"