Module: RubynCode::Context::SchemaFilter

Defined in:
lib/rubyn_code/context/schema_filter.rb

Overview

Extracts only the relevant table definitions from db/schema.rb based on which models are currently in context. Loading the full schema for a large Rails app can be 5-10K tokens; filtering to relevant tables typically reduces this to 200-500 tokens.

Constant Summary collapse

TABLE_PATTERN = 
/create_table\s+"([^"]+)"/
END_PATTERN = 
/\A\s+end\s*\z/

Class Method Summary collapse

Class Method Details

.filter(schema_path, table_names:) ⇒ String

Returns schema definitions for only the specified table names.

Parameters:

  • schema_path (String) —

    path to db/schema.rb

  • table_names (Array<String>) —

    table names to include

Returns:

  • (String) —

    filtered schema content



19
20
21
22
23
24
25
 
# File 'lib/rubyn_code/context/schema_filter.rb', line 19

def filter(schema_path, table_names:)
  return '' if table_names.empty?
  return '' unless File.exist?(schema_path)

  lines = File.readlines(schema_path)
  extract_tables(lines, table_names.to_set(&:to_s))
end

.filter_for_models(schema_path, model_names:) ⇒ Object

Convenience: filter schema by model names instead of table names.



36
37
38
39
 
# File 'lib/rubyn_code/context/schema_filter.rb', line 36

def filter_for_models(schema_path, model_names:)
  tables = tableize(model_names)
  filter(schema_path, table_names: tables)
end

.tableize(model_names) ⇒ Array<String>

Derives table names from model class names using Rails conventions.

Parameters:

  • model_names (Array<String>) —

    e.g., [“User”, “OrderItem”]

Returns:

  • (Array<String>) —

    e.g., [“users”, “order_items”]



31
32
33
 
# File 'lib/rubyn_code/context/schema_filter.rb', line 31

def tableize(model_names)
  model_names.map { |name| "#{name.gsub(/([a-z])([A-Z])/, '\1_\2').downcase}s" }
end