Preserve local curation when importing LPSN

- Local changes are now preserved between LPSN imports through a marking fields for protection (`protect_from_lpsn`).
- To use, execute the `lpsn:protect` task with the last imported version of LPSN. Any fields that differ are assumed to have been manually updated and will be protected.

Author: lmrodriguezr

db/migrate/20250623185140_add_protect_from_lpsn_to_names.rb

@@ -0,0 +1,5 @@
+class AddProtectFromLpsnToNames < ActiveRecord::Migration[6.1]
+  def change
+    add_column :names, :protect_from_lpsn, :string, default: nil
+  end
+end

db/schema.rb

@@ -10,7 +10,7 @@
 #
 # It's strongly recommended that you check this file into your version control system.
 
-ActiveRecord::Schema.define(version: 2025_05_20_204421) do
+ActiveRecord::Schema.define(version: 2025_06_23_185140) do
 
   create_table "action_text_rich_texts", force: :cascade do |t|
     t.string "name", null: false
@@ -234,6 +234,7 @@
     t.string "nomenclatural_type_type"
     t.integer "nomenclatural_type_id"
     t.datetime "claimed_at"
+    t.string "protect_from_lpsn"
     t.index ["genome_id"], name: "index_names_on_genome_id"
     t.index ["name"], name: "index_names_on_name", unique: true
     t.index ["name_order"], name: "index_names_on_name_order"

lib/tasks/import_lpsn.rake

@@ -33,7 +33,7 @@ namespace :lpsn do
       type_name = nil
       if pars[:rank] == 'genus'
         type_name = row['nomenclatural_type']
-      else
+      else # LPSN only has genera, species, and subspecies
         pars[:nomenclatural_type_type] = 'Strain'
         pars[:nomenclatural_type_entry] =
           row['nomenclatural_type'].gsub('; ', ' = ')
@@ -48,7 +48,7 @@ namespace :lpsn do
       pars[:lpsn_url] = row['address']
       pars[:status] = 20
 
-      # Save data
+      # Check first the current name record
       name = Name.find_or_create_by(name: pars[:name])
       if name.status > 5 && name.status < 20
         warn "- Name in SeqCode, bypassing: #{name.name}"
@@ -58,6 +58,17 @@ namespace :lpsn do
         warn "- Name in a different code, bypassing: #{name.name}"
         next
       end
+      if name.redirect_id.present?
+        warn "- Name is deprecated, bypassing: #{name.name}"
+        next
+      end
+      name.protect_from_lpsn.to_s.split(',').each do |i|
+        pars.delete(i.to_sym)
+        parent = nil if i == 'parent'
+        row['record_lnk'] = nil if i == 'correct_name'
+      end
+
+      # Save data
       name.update!(pars)
       parsed_names[row['record_no']] = {
         name_id: name.id, parent: parent, correct_name: row['record_lnk'],

lib/tasks/protect_lpsn.rake

@@ -0,0 +1,76 @@
+
+require 'csv'
+
+namespace :lpsn do
+  desc 'Marks changes as protected from LPSN for future updates'
+  task :protect, [:csv] => :environment do |t, args|
+    def usage(t)
+      puts "Usage: rake #{t}[lpsn_gss.csv]"
+      exit 0
+    end
+
+    usage(t) unless args[:csv]
+    # First pass, discard repeats
+    $stderr.puts "Parsing CSV: #{args[:csv]}"
+    names = {}
+    CSV.foreach(args[:csv], headers: true).each_with_index do |row, k|
+      n = [row['genus_name'], row['sp_epithet'], row['subsp_epithet']]
+      names[n] = k
+    end
+
+    # Second pass, actually do work
+    CSV.foreach(args[:csv], headers: true).each_with_index do |row, k|
+      $stderr.print "- #{k} \r"
+
+      # Discard repeats
+      n = [row['genus_name'], row['sp_epithet'], row['subsp_epithet']]
+      next unless names[n] == k
+
+      # Get data
+      pars = { name: row['genus_name'] }
+      parent = nil
+      if row['sp_epithet'].present?
+        parent = { name: pars[:name] }
+        pars[:name] = pars[:name] + ' ' + row['sp_epithet']
+        if row['subsp_epithet'].present?
+          parent = { name: pars[:name] }
+          pars[:name] = pars[:name] + ' subsp. ' + row['subsp_epithet']
+        end
+      end
+      status = row['status'].split('; ')
+      pars[:proposal_kind] = status[1] if status[1].present?
+      pars[:nomenclatural_status] = status[2] if status[2].present?
+      pars[:taxonomic_status] = status[3] if status[3].present?
+      pars[:authority] = row['authors']
+
+      # Save data
+      name = Name.find_by(name: pars[:name])
+      next unless name.present?
+      next if name.status > 5 && name.status < 20
+      next if name.status > 20
+
+      protect = Set.new(name.protect_from_lpsn.to_s.split(','))
+      %i[
+        proposal_kind nomenclatural_status taxonomic_status authority
+      ].each do |i|
+        protect << i.to_s if name.send(i).to_s != pars[i].to_s
+      end
+      if row['record_lnk'].present? && (
+              protect.include?('nomenclatural_status') ||
+              protect.include?('taxonomic_status')
+            )
+        protect << 'correct_name'
+      end
+      if parent.present? && name.parent.present? &&
+            !name.parent.is_variant?(parent[:name])
+        protect << 'parent'
+      end
+      unless protect.empty?
+        warn "- Protecting: #{name.name}: #{protect.join(', ')}"
+        name.update_column(:protect_from_lpsn, protect.join(','))
+      end
+    end # CSV.foreach
+    $stderr.puts
+  end
+
+end