neu-mods 0.2.0 → 0.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: fb4c30744c189d3e0b4bfc0f84bb75175685bcb142dd6dc35dc55cfbcbdaf0a2
-  data.tar.gz: a5cd73f764edf2da97e05787788124bed168c9dba8caf997307fc53591f05661
+  metadata.gz: ebf2bf82cd0b9c0e5593307ecc9a986cefcb16f4035a5004f233bdd7f9d4d8ea
+  data.tar.gz: 7087e35df1b2f34430a53acfa49e9c97be12dd9b8bdc36b3ffa69d05e98189e0
 SHA512:
-  metadata.gz: 785e3c7485886be0fd7946b61463e2e2b66319cf4c4f30980b6e4b2b0b8ed2ae4348be41a48caa9fbad536901a89e9c73d4d4fd005283bc2cc0dc9c9102d6ff6
-  data.tar.gz: 7a331ed220702dbab23bd6d16637f33a3da2f431a17939647ee7d1fd21898b9cf76a4594b26300eb01f8c9ad1b50d0cabcd298cdbff9c363a01a7f7bba6229a7
+  metadata.gz: 63b83e54cdf64af06856fe1de2cb3763cc36a1f1f0a73403fb7836fac9e910dcd2f0ebd0a4fd5fb928b7714ddec20ad281ac1cc42b47a18fcda21cbc8ad8589f
+  data.tar.gz: 3d6ecb5f48c1c4cff669add4adc7bfa2a90bf29045b17bb1b1047a58f02820c85fbe303ec0c0e7665f7045e27d03531fe32759f6bc023d91032038baf4466c65

data/.version

@@ -1 +1 @@
-0.2.0
+0.3.0

data/lib/neu/mods/projection.rb

@@ -45,7 +45,32 @@ module NEU
 
         optional = { ": " => parts[:subtitle], " - " => parts[:part_name], ", " => parts[:part_number] }
         suffix = optional.filter_map { |sep, val| "#{sep}#{val}" unless val.to_s.strip.empty? }.join
-        "#{parts[:non_sort]}#{parts[:title]}#{suffix}"
+        "#{join_non_sort(parts[:non_sort], parts[:title])}#{suffix}"
+      end
+
+      # Characters that bind a nonSort to the word after it. An elided article
+      # takes no space -- "L'Etranger", not "L' Etranger" -- and the same holds
+      # for a hyphenated prefix. U+2019 is the curly apostrophe, escaped rather
+      # than literal to keep lib/ pure ASCII (see the source-purity spec).
+      NON_SORT_BINDING = ["'", "\u2019", "-"].freeze
+
+      # MODS says a nonSort carries whatever separator it needs, so the historical
+      # composition simply concatenated. That only holds while the authored
+      # trailing space survives, and it does not: #child_text canonicalizes
+      # whitespace on read, so `<nonSort>The </nonSort>` arrives here as "The" and
+      # the title came out as "TheHobbit". Composing the separator instead makes
+      # the output right whether or not the source kept one -- which matters,
+      # because an invisible trailing space is not something a curator, a
+      # hand-edit or a third-party producer can be relied on to preserve.
+      #
+      # Callers that DO pass the space (Atlas's access-copy model) are unaffected:
+      # a nonSort already ending in whitespace is joined as-is.
+      def self.join_non_sort(non_sort, title)
+        prefix = non_sort.to_s
+        return title.to_s if prefix.empty?
+        return "#{prefix}#{title}" if prefix.end_with?(" ") || prefix.end_with?(*NON_SORT_BINDING)
+
+        "#{prefix} #{title}"
       end
 
       # --- Abstract / access ---------------------------------------------------

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: neu-mods
 version: !ruby/object:Gem::Version
-  version: 0.2.0
+  version: 0.3.0
 platform: ruby
 authors:
 - David Cliff
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2026-06-09 00:00:00.000000000 Z
+date: 2026-08-03 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: nokogiri