traject 3.8.1 → 3.8.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 0e54c076cdd70f6beacb8e9bdb6057c6ba809393398c7c786c6e7c943117581e
-  data.tar.gz: 77ed56fc2ee2355ab584af074fee656f8cfb543d4275a63384ce069a52a20f1a
+  metadata.gz: c4434e6a85dc5176d84dfb2c6e42c0882937cdffa37c24e5f5f06844ad0e542d
+  data.tar.gz: fd2da1d197ace61ab86e2308298e77ed64866ef0f1ffc49b88b344578900f76e
 SHA512:
-  metadata.gz: 3b3a6826b6825c08604c7ed8608b63fe98a889de676c2b0b7ba7eb1002eb029fe04f19b4fa57a480ff486f0acd911ddec749c2f6a3589581588d17e93648deeb
-  data.tar.gz: 1451ae18ada4181224d4a02269ba2532b0914d69ab7d07ab60d3390e12605c239e0304a93511ab3c3c854d9251623af94c852b2a97eab02fb19f969f43a374cb
+  metadata.gz: 3aad25487edbb9489a30f6d8a9c6c2fdb467683c6cd6a96199d7de10538f861fd92c2f25ec8213b5a753ef85bc7440c1d27ee48d65fc05d2f36e735a15a7b6a6
+  data.tar.gz: 7641b36d876e01d3b8188add71a4121a768116a8bd2462c1fcfb0e2652af7ddebfc9a17d03dd24c68fa56b8810e2b7cd865eea53ff905a53fb5f7675c5dc99e4

data/.github/workflows/ruby.yml

@@ -2,7 +2,7 @@ name: CI
 
 on:
   push:
-    branches: [ master ]
+    branches: [ main ]
   pull_request:
     branches: ['**']
 
@@ -12,7 +12,15 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        ruby: [ '2.4', '2.5', '2.6', '2.7', '3.0', '3.1', 'jruby-9.1', 'jruby-9.2' ]
+        ruby:
+          - '2.7'
+          - '3.0'
+          - '3.1'
+          - '3.2'
+          - '3.3'
+          - '3.4'
+          - 'jruby-9.3'
+          - 'jruby-9.4'
     name: Ruby ${{ matrix.ruby }}
     steps:
     - uses: actions/checkout@v2

data/CHANGES.md

@@ -1,10 +1,22 @@
 # Changes
 
-## NEXT
+## 3.8.3 Compatibility release for ruby 3.4
 
-*
+Note that some gems, `ruby-marc` in particular, will throw zillions of
+warnings about literal strings being frozen in future versions of ruby. 
 
-*
+- Update CI testing matrix
+  - removed MRI ruby 2.4, 2.5, 2.6, and jruby-9.2
+  - added MRI ruby 3.4
+- Add gem dependencies for `csv` and `mutex_m`, both of which were 
+  part of the standard library prior to ruby 3.4
+
+
+## 3.8.2
+
+Bug fix for the `#filing_version` logic, which was incorrectly assuming the 
+first subfield in a field would hold content (e.g., `$a`) and thus failed
+when it held a pointer to a linking field (e.g., `$6 245-01`)
 
 ## 3.8.1
 

data/lib/tasks/load_maps.rake

@@ -1,20 +1,16 @@
 require 'net/http'
 require 'open-uri'
+require 'csv'
 
 
 
+CODELIST_NS = 'info:lc/xmlns/codelist-v1'
 
 namespace :load_maps do
 
   desc "Load MARC geo codes by screen-scraping LC"
-  task :marc_geographic do
-    begin
-      require 'nokogiri'
-    rescue LoadError => e
-      $stderr.puts "\n  load_maps:marc_geographic task requires nokogiri"
-      $stderr.puts "  Try `gem install nokogiri` and try again. Exiting...\n\n"
-      exit 1
-    end
+  task :marc_geographic do |task|
+    require_nokogiri(task)
 
     source_url = "http://www.loc.gov/marc/geoareas/gacs_code.html"
 
@@ -45,4 +41,51 @@ namespace :load_maps do
     end
     $stderr.puts "Done."
   end
+
+  desc "Load MARC language codes from LOC and SIL"
+  task :marc_languages do |task|
+    require_nokogiri(task)
+    filename = ENV["OUTPUT_TO"] || File.expand_path("../../translation_maps/marc_languages.yaml", __FILE__)
+    file = File.open(filename, "w:utf-8")
+    $stderr.puts "Writing to `#{filename}` ..."
+    file.puts("# Map Language Codes (in 008[35-37], 041) to User Friendly Term\r")
+
+    marc_language_source_url = 'https://www.loc.gov/standards/codelists/languages.xml'
+    doc = Nokogiri::XML(URI.parse(marc_language_source_url).open)
+    marc_language_hash = doc.xpath('//codelist:language', codelist: CODELIST_NS)
+                            .to_h do |node|
+                              [node.xpath('./codelist:code/text()', codelist: CODELIST_NS).to_s,
+                               node.xpath('./codelist:name/text()', codelist: CODELIST_NS).to_s]
+                            end.reject { |key, _val| %w[und zxx].include? key }
+
+    file.puts "\r"
+    file.puts("# MARC language codes (including obsolete codes), from #{marc_language_source_url}\r\n")
+    marc_language_hash.sort_by { |k, _v| k }.each do |key, value|
+      file.puts("#{key}: #{escape_special_yaml_chars(value)}\r")
+    end
+
+    iso_639_3_url = 'https://iso639-3.sil.org/sites/iso639-3/files/downloads/iso-639-3.tab'
+    parsed_url = URI.parse(iso_639_3_url)
+    iso_languages = CSV.parse(parsed_url.read(encoding: 'UTF-8'), headers: true, col_sep: "\t", encoding: "UTF-8")
+    iso_language_hash = iso_languages.to_h { |row| [row['Id'], row['Ref_Name']] }
+                                     .reject { |key, _val| %w[und zxx].include? key }
+                                     .reject { |key, _val| marc_language_hash.keys.include? key }
+    file.puts "\r"
+    file.puts("# ISO 639-3 codes, from #{iso_639_3_url}\r")
+    iso_language_hash.sort_by { |k, _v| k }.each do |key, value|
+      file.puts("#{key}: #{escape_special_yaml_chars(value)}\r")
+    end
+  end
+
+  def require_nokogiri(task)
+    require 'nokogiri'
+  rescue LoadError
+    $stderr.puts "\n  #{task&.name} task requires nokogiri"
+    $stderr.puts "  Try `gem install nokogiri` and try again. Exiting...\n\n"
+    exit 1
+  end
+
+  def escape_special_yaml_chars(string)
+    string.match(/[\,\']/) ? %Q{"#{string}"} : string
+  end
 end

data/lib/traject/macros/marc21_semantics.rb

@@ -167,7 +167,13 @@ module Traject::Macros
       # (b) include the first subfield in the record
 
       subs = spec.subfields
-      return str unless subs && subs.include?(field.subfields[0].code)
+
+      # Get the code for the first alphabetic subfield, which would be
+      # the one getting characters shifted off
+
+      first_alpha_code = field.subfields.first{|sf| sf.code =~ /[a-z]/}.code
+
+      return str unless subs && subs.include?(first_alpha_code)
 
       # OK. If we got this far we actually need to strip characters off the string
 

data/lib/traject/version.rb

@@ -1,3 +1,3 @@
 module Traject
-  VERSION = "3.8.1"
+  VERSION = "3.8.3"
 end