real_data_tests 0.3.6 → 0.3.8

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 179a591f6a485c397255c0cbec27e18757e2f7fdc450f7292f279cd681f91e0b
-  data.tar.gz: bb9b9f4f60b609f2ea8a901689b4b9b8d5ddfdc7d36e867c18ebbe24855cb5aa
+  metadata.gz: 573f1753777e04a6168a09385b8332cedbb4d521d11f1807012a25b8b846013c
+  data.tar.gz: b884098abbc9134a22479d5e9211a2d8bb153622022ba9116acb0057419cbe4a
 SHA512:
-  metadata.gz: a47d49d690958e70cd324ede48dadce7d3e9e7bd39471ee052879d6fe5646ad308aebbf05f79cb92ebe3c9b7ede94b0163c74efdddf3f518ef47518a489a2ee6
-  data.tar.gz: a4c6a97cff1bd02096064fd1da67580dd9307ce29f4e7d5a8be4d82c9f352b97009ef1fc22eb555dc39b1b87599947680a61bc2bc376c2cd14a7f8cc5397892d
+  metadata.gz: 1d4bcb84f7d3ab3ee0901ae46763c912f3c86935930cc17607d500ae7a2e26649f8c6ff629e8215437d8f08ae93d482870e253b294e84e720971c9bf681c7b50
+  data.tar.gz: 67cad3bda841d3e1de61d61dd30121ef3b6ddfc96a3162438a2b25f904185e6e4c59206d2626e6b40d44a72d123ed123e4073a4fdfe10d04fea807a8702bf689

data/CHANGELOG.md

@@ -1,5 +1,21 @@
 ## [Unreleased]
 
+## [0.3.8] - 2025-01-14
+### Fixed
+- Enhanced SQL statement parsing in native loader
+  - Improved handling of complex ON CONFLICT clauses with multiple closing parentheses
+  - Fixed spacing issues between VALUES and ON CONFLICT clauses
+  - Enhanced regex pattern for more precise conflict clause extraction
+  - Added proper statement reassembly for complex SQL structures
+
+## [0.3.7] - 2025-01-14
+### Fixed
+- Corrected SQL value handling in native loader
+  - Fixed boolean value handling, particularly at the end of VALUES clauses
+  - Improved handling of NULL values and numbers
+  - Added proper whitespace cleaning for values
+  - Enhanced value type detection for correct quoting
+
 ## [0.3.6] - 2025-01-14
 ### Fixed
 - Further enhanced SQL statement handling in native loader

data/lib/real_data_tests/rspec_helper.rb

@@ -103,81 +103,86 @@ module RealDataTests
       statements
     end
 
-          def clean_sql_statement(statement)
-        # Handle VALUES clause formatting
-        if statement.include?('VALUES')
-          # Split into pre-VALUES and VALUES parts
-          parts = statement.split(/VALUES\s*\(/i, 2)
-          if parts.length == 2
-            # Get the column names from the first part
-            column_names = parts[0].scan(/\((.*?)\)/).flatten.first&.split(',')&.map(&:strip) || []
-
-            # Split values into individual items
-            values_part = parts[1]
-            values = split_values(values_part)
-
-            # Process each value according to its column and position
-            values.each_with_index do |value, i|
-              # Skip if already properly quoted
-              next if value.start_with?("'") && value.end_with?("'")
-
-              # Quote UUIDs
-              if value.match?(/^[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}$/i)
-                values[i] = "'#{value}'"
-              # Quote any non-NULL string values that contain spaces or special characters
-              elsif !['NULL', 'true', 'false'].include?(value) &&
-                    !value.match?(/^\d+$/) && # not a number
-                    !value.start_with?("'{") && # not a JSON object
-                    !value.start_with?("'") # not already quoted
-                values[i] = "'#{value}'"
-              end
-            end
+    def extract_conflict_clause(statement)
+      # Use a more precise regex that handles multiple closing parentheses
+      if statement =~ /(.+?\))\s*(ON\s+CONFLICT\s+.*?)(?:;?\s*$)/i
+        [$1, $2.strip]
+      else
+        [statement.sub(/;?\s*$/, ''), nil]
+      end
+    end
 
-            # Reassemble the statement with any trailing ON CONFLICT clause
-            remainder = values_part.match(/\)\s*(ON\s+CONFLICT.*?)(;?\s*$)/i)
-            statement = parts[0] + 'VALUES (' + values.join(', ') + ')'
-            statement += " #{remainder[1]};" if remainder
-          end
-        end
+    def clean_sql_statement(statement)
+      # Extract the ON CONFLICT clause if it exists
+      statement, conflict_clause = extract_conflict_clause(statement)
+
+      # Handle VALUES clause formatting
+      if statement.include?('VALUES')
+        # Split into pre-VALUES and VALUES parts
+        parts = statement.split(/VALUES\s*\(/i, 2)
+        if parts.length == 2
+          # Clean and process the values
+          values = clean_values(parts[1].split(/\)\s*$/)[0])
 
-        statement
+          # Make sure we properly close the VALUES parentheses
+          statement = "#{parts[0]}VALUES (#{values})"
+        end
       end
 
-      def split_values(values_part)
-        values = []
-        current_value = ''
-        in_quotes = false
-        in_json = false
-        depth = 0
-
-        values_part.chars.each_with_index do |char, i|
-          case char
-          when "'"
-            # Toggle quote state if not escaped
-            if i == 0 || values_part[i-1] != '\\'
-              in_quotes = !in_quotes
-            end
-          when '{'
-            depth += 1 unless in_quotes
-            in_json = true unless in_quotes
-          when '}'
-            depth -= 1 unless in_quotes
-          when ','
-            if !in_quotes && depth == 0
-              values << current_value.strip
-              current_value = ''
-              next
-            end
+      # Add back the conflict clause if it existed
+      statement = "#{statement}#{conflict_clause ? ' ' + conflict_clause : ''}"
+
+      # Ensure the statement ends with a semicolon
+      statement += ";" unless statement.end_with?(';')
+
+      statement
+    end
+
+    def clean_values(values_str)
+      values = []
+      current_value = ''
+      in_quotes = false
+      nested_level = 0
+
+      values_str.chars.each do |char|
+        case char
+        when "'"
+          in_quotes = !in_quotes
+          current_value << char
+        when '{'
+          nested_level += 1
+          current_value << char
+        when '}'
+          nested_level -= 1
+          current_value << char
+        when ','
+          if !in_quotes && nested_level == 0
+            values << clean_value(current_value.strip)
+            current_value = ''
+          else
+            current_value << char
           end
+        else
           current_value << char
         end
+      end
+
+      values << clean_value(current_value.strip)
+      values.join(', ')
+    end
 
-        # Add the last value (remove trailing ';' or ')')
-        last_value = current_value.strip
-        last_value = last_value[0..-2] if last_value.end_with?(';', ')')
-        values << last_value
+    def clean_value(value)
+      return value if value.start_with?("'") # Already quoted
+      return value if value.start_with?("'{") # JSON object
+      return 'NULL' if value.upcase == 'NULL'
+      return value.downcase if ['true', 'false'].include?(value.downcase)
+      return value if value.match?(/^\d+$/) # Numbers
 
-        values
+      if value.match?(/^[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}$/i)
+        "'#{value}'" # UUID
+      else
+        "'#{value}'" # Other strings
       end
+    end
   end
 end

data/lib/real_data_tests/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module RealDataTests
-  VERSION = "0.3.6"
+  VERSION = "0.3.8"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: real_data_tests
 version: !ruby/object:Gem::Version
-  version: 0.3.6
+  version: 0.3.8
 platform: ruby
 authors:
 - Kevin Dias