lex-knowledge 0.6.6 → 0.6.9

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 9db1bb96787265658439252f9632523d527f1b34b759769379a62b3577d64db5
-  data.tar.gz: c072f647cf7b91c55ebf1edb77c1ca2cbe57fde4e00a968f2e7b016502271c9c
+  metadata.gz: 29631e6f2778ceedf74b69197379edc4c215d9b7ff4b2e088790dedc227fd90f
+  data.tar.gz: d96dd08d5a1450e5e1f3db0936729d3a40c834edc166126d6c9dd73a1839518f
 SHA512:
-  metadata.gz: 7058729c26b22b3b85674650e91f1be9ab8869761cfae0ec710d3aafea306daad91e66a14d69fea55349666c135e8a6bd32a97068ac520adc74e34b47b1536b2
-  data.tar.gz: 78176d44e646f407bd7dfe3852d0249779404df4e690c16a2abd8a391a7f29e580bff4cdcd9cb893a7ee6849dfc1f01bfdfa27f1870a1d0aceb9aaae8f4f0524
+  metadata.gz: 971edc3c4347a949a89b690cccda16150e5644be27eb201d22daeaab0e293cc03ebf530e509712a514eb2f40e3e86c201ce96475c2cbf398252fcae73bc0af64
+  data.tar.gz: 980eb54c45d386db31c1ed817f41f2f12d6a7993e07db910c016398ddb5229274622c720e78eb1aabcb0920d01772de63627ecbff6653bb051b893b7e230d766

data/lib/legion/extensions/knowledge/helpers/manifest.rb

@@ -1,7 +1,6 @@
 # frozen_string_literal: true
 
 require 'digest'
-require 'find'
 
 module Legion
   module Extensions
@@ -12,19 +11,28 @@ module Legion
 
           def scan(path:, extensions: %w[.md .txt .docx .pdf])
             results = []
+            walk(path, extensions, results)
+            results
+          end
 
-            Find.find(path) do |entry|
-              basename = ::File.basename(entry)
-              Find.prune if basename.start_with?('.')
-
-              next unless ::File.file?(entry)
-              next unless extensions.include?(::File.extname(entry).downcase)
+          def walk(entry, extensions, results)
+            basename = ::File.basename(entry)
+            return if basename.start_with?('.')
 
+            if ::File.directory?(entry)
+              ::Dir.children(entry).each { |c| walk(::File.join(entry, c), extensions, results) }
+            elsif ::File.file?(entry) && extensions.include?(::File.extname(entry).downcase)
               results << build_entry(entry)
             end
+          rescue Errno::EPERM, Errno::EACCES, Errno::ELOOP, Errno::ENOENT => e
+            log.debug("[manifest] skipping unreadable #{entry}: #{e.class}: #{e.message}")
+          end
+          private_class_method :walk
 
-            results
+          def log
+            Legion::Logging
           end
+          private_class_method :log
 
           def diff(current:, previous:)
             current_map  = current.to_h { |e| [e[:path], e[:sha256]] }

data/lib/legion/extensions/knowledge/runners/maintenance.rb

@@ -47,16 +47,19 @@ module Legion
           end
 
           def health(path:)
-            scan_entries  = Helpers::Manifest.scan(path: path)
-            store_path    = Helpers::ManifestStore.store_path(corpus_path: path)
+            resolved = path || (Legion::Settings.dig(:knowledge, :corpus_path) if defined?(Legion::Settings))
+            return { success: false, error: 'corpus_path is required' } if resolved.nil? || resolved.to_s.empty?
+
+            scan_entries  = Helpers::Manifest.scan(path: resolved)
+            store_path    = Helpers::ManifestStore.store_path(corpus_path: resolved)
             manifest_file = ::File.exist?(store_path)
             last_ingest   = manifest_file ? ::File.mtime(store_path).iso8601 : nil
 
             {
               success: true,
-              local:   build_local_stats(path, scan_entries, manifest_file, last_ingest),
+              local:   build_local_stats(resolved, scan_entries, manifest_file, last_ingest),
               apollo:  build_apollo_stats,
-              sync:    build_sync_stats(path, scan_entries)
+              sync:    build_sync_stats(resolved, scan_entries)
             }
           rescue StandardError => e
             { success: false, error: e.message }

data/lib/legion/extensions/knowledge/runners/query.rb

@@ -69,11 +69,12 @@ module Legion
           def retrieve_chunks(question, top_k)
             return [] unless defined?(Legion::Extensions::Apollo)
 
-            Legion::Extensions::Apollo::Runners::Knowledge.retrieve_relevant(
+            result = Legion::Extensions::Apollo::Runners::Knowledge.retrieve_relevant(
               query: question,
               limit: top_k,
               tags:  ['document_chunk']
             )
+            result.is_a?(Hash) && result[:success] ? Array(result[:entries]) : []
           rescue StandardError => _e
             []
           end

data/lib/legion/extensions/knowledge/version.rb

@@ -3,7 +3,7 @@
 module Legion
   module Extensions
     module Knowledge
-      VERSION = '0.6.6'
+      VERSION = '0.6.9'
     end
   end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: lex-knowledge
 version: !ruby/object:Gem::Version
-  version: 0.6.6
+  version: 0.6.9
 platform: ruby
 authors:
 - Matthew Iverson