icql-dba

'use strict' ############################################################################################################ CND = require 'cnd' rpr = CND.rpr badge = 'ICQL/DBA' debug = CND.get_logger 'debug', badge warn = CND.get_logger 'warn', badge info = CND.get_logger 'info', badge urge = CND.get_logger 'urge', badge help = CND.get_logger 'help', badge whisper = CND.get_logger 'whisper', badge echo = CND.echo.bind CND #........................................................................................................... FS = require 'fs' #........................................................................................................... types = require './types' { freeze lets } = require 'letsfreezethat' L = @ { misfit } = require './common' E = require './errors' new_bsqlt3_connection = require 'better-sqlite3' PATH = require 'path' TMP = require 'tempy' { Import_export_mixin } = require './import-export-mixin' { Functions_mixin } = require './functions-mixin' { Checks_mixin } = require './checks-mixin' { Stdlib_mixin } = require './stdlib-mixin' guy = require 'guy' SQL = String.raw #=========================================================================================================== # #----------------------------------------------------------------------------------------------------------- class @Dba extends Stdlib_mixin Checks_mixin Functions_mixin Import_export_mixin() #--------------------------------------------------------------------------------------------------------- constructor: ( cfg ) -> super() guy.props.def @, 'types', { enumerable: false, value: types, } guy.props.def @, '_statements', { enumerable: false, value: {}, } guy.props.def @, 'sql', { enumerable: false, value: ( new ( require './sql' ).Sql() ), } @_schemas = freeze {} @cfg = freeze { @types.defaults.dba_constructor_cfg..., cfg..., } @types.validate.dba_constructor_cfg @cfg @_dbg = { debug: @cfg.debug, echo: @cfg.echo, } @_formats = freeze { @types.defaults.extensions_and_formats..., } throw new E.Dba_cfg_error '^dba@300^', "property `sqlt` not supported" if @cfg.sqlt? throw new E.Dba_cfg_error '^dba@301^', "property `schema` not supported" if @cfg.schema? throw new E.Dba_cfg_error '^dba@302^', "property `path` not supported" if @cfg.path? @_bsqlt3_cfg = freeze { readonly: @cfg.readonly fileMustExist: not @cfg.create timeout: @cfg.timeout } # verbose: ### TAINT to be done ### @_state = { in_unsafe_mode: false initialized: false stdlib_created: false } @_catalog = freeze {} ### NOTE: will hold data on user-defined functions, virtual tables ### #....................................................................................................... guy.props.def_oneoff @, 'sqlt', {}, => connection = new_bsqlt3_connection '', @_bsqlt3_cfg @initialize_sqlt connection @_state.initialized = true return connection #....................................................................................................... return undefined # #--------------------------------------------------------------------------------------------------------- # constructor: ( cfg ) -> # super() # guy.props.def @, 'types', { enumerable: false, value: types, } # guy.props.def @, '_statements', { enumerable: false, value: {}, } # guy.props.def @, 'sql', { enumerable: false, value: ( new ( require './sql' ).Sql() ), } # @_schemas = freeze {} # @cfg = freeze { @types.defaults.dba_constructor_cfg..., cfg..., } # @types.validate.dba_constructor_cfg @cfg # @_dbg = { debug: @cfg.debug, echo: @cfg.echo, } # @_formats = freeze { @types.defaults.extensions_and_formats..., } # # throw new E.Dba_cfg_error '^dba@300^', "property `sqlt` not supported" if @cfg.sqlt? # throw new E.Dba_cfg_error '^dba@300^', "property `schema` not supported" if @cfg.schema? # @_bsqlt3_cfg = freeze { # readonly: @cfg.readonly # fileMustExist: not @cfg.create # timeout: @cfg.timeout } # # verbose: ### TAINT to be done ### # @_state = { # in_unsafe_mode: false # initialized: false # stdlib_created: false } # @_catalog = freeze {} ### NOTE: will hold data on user-defined functions, virtual tables ### # @_state.initialized = true ### TAINT remove ### # @_initialize_prng() # #....................................................................................................... # { ram # name # path } = @cfg # if ram # ### formulate connection string à la 'file:memdb1?mode=memory&cache=shared' ### # { url # name } = dba._get_connection_url name ? null # debug '^3443^', name, url # if path? # ### File DB, Eventual Persistency; opened from file (may get created), mirror to RAM ### # null # else # ### RAM DB, No Persistency, starts empty ### # null # else # ### File DB, Continuous Persistency ### # null # @sqlt = new_bsqlt3_connection path, @_bsqlt3_cfg # @sqlt2 = new_bsqlt3_connection path, @_bsqlt3_cfg # @initialize_sqlt @sqlt # #....................................................................................................... # return undefined #--------------------------------------------------------------------------------------------------------- initialize_sqlt: ( sqlt ) -> ### NOTE do NOT use any instance methods in the initializer as this will lead to infinite regress b/c of the way the dynamic attribute `@sqlt` has been implemented. ### sqlt.pragma "foreign_keys = true;" return null #--------------------------------------------------------------------------------------------------------- open: ( cfg ) -> @types.validate.dba_open_cfg ( cfg = { @types.defaults.dba_open_cfg..., cfg..., } ) { path, schema, ram, } = cfg #....................................................................................................... ### TAINT troublesome logic with `path` and `saveas` ### if path? saveas = path else path = '' ### TAINT or ':memory:' depending on `cfg.disk` ### saveas = null #....................................................................................................... if ram then @_open_file_db_in_ram { path, schema, saveas, } else @_attach { path, schema, saveas, } #....................................................................................................... return null #--------------------------------------------------------------------------------------------------------- save: ( cfg ) -> ### TAINT could implement prohibition of `path` in type `dba_save_cfg` ### @types.validate.dba_save_cfg ( cfg = { @types.defaults.dba_export_cfg..., cfg..., } ) { schema path } = cfg throw new E.Dba_argument_not_allowed '^dba@303^', 'path', path if path? path = @_schemas[ schema ]?.path ? null throw new E.Dba_schema_unknown '^dba@304^', schema unless path? return @export { schema, path, format: 'sqlite', } #--------------------------------------------------------------------------------------------------------- export: ( cfg ) -> ### TAINT add boolean `cfg.overwrite` ### @types.validate.dba_export_cfg ( cfg = { @types.defaults.dba_export_cfg..., cfg..., } ) { schema path format } = cfg format ?= @_format_from_path path throw new E.Dba_extension_unknown '^dba@305^', path unless format? switch format when 'sqlite' then @_vacuum_atomically { schema, path, } ### TAINT when format derived from path, may be undefined, making the error message unintelligible ### else throw new E.Dba_format_unknown '^dba@306^', format return null #--------------------------------------------------------------------------------------------------------- _vacuum_atomically: ( cfg ) -> @types.validate.dba_vacuum_atomically ( cfg = { @types.defaults.dba_vacuum_atomically..., cfg..., } ) { schema path } = cfg schema_i = @sql.I schema try tmpdir_path = TMP.directory { prefix: @cfg._temp_prefix, } tmpfile_path = PATH.join tmpdir_path, PATH.basename path @run "vacuum #{schema_i} into ?;", [ tmpfile_path, ] FS.renameSync tmpfile_path, path finally FS.rmdirSync tmpdir_path return null #--------------------------------------------------------------------------------------------------------- is_ram_db: ( cfg ) -> @types.validate.dba_is_ram_db_cfg ( cfg = { @types.defaults.dba_is_ram_db_cfg..., cfg..., } ) { schema } = cfg sql = "select file from pragma_database_list where name = ? limit 1;" try return @types.isa.dba_ram_path @single_value @query sql, [ schema, ] catch error throw new E.Dba_schema_unknown '^dba@307^', schema if error instanceof E.Dba_expected_one_row throw error #--------------------------------------------------------------------------------------------------------- _list_temp_schema_numbers: -> matcher = @cfg._temp_prefix + '%' sql = """ select cast( substring( name, ? ) as integer ) as n from pragma_database_list where name like ?;""" return @all_first_values @query sql, [ @cfg._temp_prefix.length + 1, matcher, ] #--------------------------------------------------------------------------------------------------------- _max_temp_schema_number: -> matcher = @cfg._temp_prefix + '%' sql = """ select max( cast( substring( name, ? ) as integer ) ) as n from pragma_database_list where name like ?;""" return ( @first_value @query sql, [ @cfg._temp_prefix.length + 1, matcher, ] ) ? 0 #--------------------------------------------------------------------------------------------------------- _get_free_temp_schema: -> @cfg._temp_prefix + "#{( @_max_temp_schema_number() + 1 )}" #========================================================================================================= # DEBUGGING #--------------------------------------------------------------------------------------------------------- _echo: ( ref, sql ) -> return null unless @_dbg.echo echo ( CND.reverse CND.blue "^icql@888-#{ref}^" ) + ( CND.reverse CND.yellow sql ) return null #--------------------------------------------------------------------------------------------------------- _debug: ( P... ) -> return null unless @_dbg.debug debug P... return null #========================================================================================================= # QUERY RESULT ADAPTERS #--------------------------------------------------------------------------------------------------------- limit: ( n, iterator ) -> count = 0 for x from iterator return if count >= n count += +1 yield x return #--------------------------------------------------------------------------------------------------------- single_row: ( iterator ) -> throw new E.Dba_expected_one_row 'dba@763^', 0 if ( R = @first_row iterator ) is undefined return R #--------------------------------------------------------------------------------------------------------- all_first_values: ( iterator ) -> R = [] for row from iterator for key, value of row R.push value break return R #--------------------------------------------------------------------------------------------------------- first_values: ( iterator ) -> R = [] for row from iterator for key, value of row yield value return R #--------------------------------------------------------------------------------------------------------- first_row: ( iterator ) -> return row for row from iterator ### TAINT must ensure order of keys in row is same as order of fields in query ### single_value: ( iterator ) -> return value for key, value of @single_row iterator first_value: ( iterator ) -> return value for key, value of @first_row iterator list: ( iterator ) -> [ iterator..., ] #========================================================================================================= # QUERYING #--------------------------------------------------------------------------------------------------------- query: ( sql, P... ) -> @_echo 'query', sql statement = ( @_statements[ sql ] ?= @sqlt.prepare sql ) return statement.iterate P... #--------------------------------------------------------------------------------------------------------- run: ( sql, P... ) -> @_echo 'run', sql statement = ( @_statements[ sql ] ?= @sqlt.prepare sql ) return statement.run P... #--------------------------------------------------------------------------------------------------------- _run_or_query: ( entry_type, is_last, sql, Q ) -> @_echo '_run_or_query', sql statement = ( @_statements[ sql ] ?= @sqlt.prepare sql ) returns_data = statement.reader #....................................................................................................... ### Always use `run()` method if statement does not return data: ### unless returns_data return if Q? then ( statement.run Q ) else statement.run() #....................................................................................................... ### If statement does return data, consume iterator unless this is the last statement: ### if ( entry_type is 'procedure' ) or ( not is_last ) return if Q? then ( statement.all Q ) else statement.all() #....................................................................................................... ### Return iterator: ### return if Q? then ( statement.iterate Q ) else statement.iterate() #--------------------------------------------------------------------------------------------------------- execute: ( sql ) -> throw new E.Dba_argument_not_allowed '^dba@308^', "extra", rpr x if ( x = arguments[ 1 ] )? @_echo 'execute', sql return @sqlt.exec sql #--------------------------------------------------------------------------------------------------------- prepare: ( sql ) -> @_echo 'prepare', sql return @sqlt.prepare sql #========================================================================================================= # OTHER #--------------------------------------------------------------------------------------------------------- backup: ( P... ) -> @sqlt.backup P... checkpoint: ( P... ) -> @sqlt.checkpoint P... close: ( P... ) -> @sqlt.close P... read: ( path ) -> @sqlt.exec FS.readFileSync path, { encoding: 'utf-8', } load_extension: ( P... ) -> @sqlt.loadExtension P... pragma: ( P... ) -> @sqlt.pragma P... #========================================================================================================= # DB STRUCTURE REPORTING #--------------------------------------------------------------------------------------------------------- catalog: -> ### TAINT kludge: we sort by descending types so views, tables come before indexes (b/c you can't drop a primary key index in SQLite) ### throw new E.Dba_not_implemented '^dba@309^', "method dba.catalog()" @query "select * from sqlite_schema order by type desc, name;" #--------------------------------------------------------------------------------------------------------- walk_objects: ( cfg ) -> @types.validate.dba_walk_objects_cfg ( cfg = { @types.defaults.dba_walk_objects_cfg..., cfg..., } ) schema = cfg.schema ordering = cfg._ordering return @_walk_all_objects() unless schema? schema_i = @sql.I schema ordering_x = if ( ordering is 'drop' ) then 'desc' else 'asc' seq = @first_value @query \ @sql.SQL"select seq from pragma_database_list where name = #{@sql.L schema};" #....................................................................................................... return @query @sql.SQL""" select #{seq} as seq, #{@sql.L schema} as schema, name as name, type as type, sql as sql from #{@sql.I schema}.sqlite_schema order by seq, type #{ordering_x}, name;""" #--------------------------------------------------------------------------------------------------------- _walk_all_objects: -> schemas = {} parts = [] #....................................................................................................... ### TAINT use API ### for row from @query SQL"select seq, name, file as path from pragma_database_list order by seq;" schemas[ row.name ] = row #....................................................................................................... for schema, d of schemas schema_i = @sql.I schema schema_l = @sql.L schema parts.push SQL"""select #{d.seq} as seq, #{schema_l} as schema, name as name, type as type, sql as sql from #{schema_i}.sqlite_schema as d1""" parts = parts.join " union all\n" #....................................................................................................... sql = '' sql += parts sql += "\norder by seq, type, name;" return @query sql #--------------------------------------------------------------------------------------------------------- is_empty: ( cfg ) -> @types.validate.dba_is_empty_cfg ( cfg = { @types.defaults.dba_is_empty_cfg..., cfg..., } ) return ( has_schema = @_is_empty_schema @sql.I cfg.schema ) unless name? throw new E.Dba_not_implemented '^dba@310^', "dba.is_empty() for anything but schemas (got #{rpr cfg})" #--------------------------------------------------------------------------------------------------------- _is_empty_schema: ( schema_i ) -> ( @list @query "select 1 from #{schema_i}.sqlite_schema limit 1;" ).length is 0 #--------------------------------------------------------------------------------------------------------- list_schemas: -> @list @query "select * from pragma_database_list order by name;" list_schema_names: -> ( d.name for d in @list_schemas() ) #--------------------------------------------------------------------------------------------------------- has: ( cfg ) -> @types.validate.dba_has_cfg ( cfg = { @types.defaults.dba_has_cfg..., cfg..., } ) return cfg.schema in @list_schema_names() #--------------------------------------------------------------------------------------------------------- get_schemas: -> R = {} R[ row.name ] = row.file for row from @query "select * from pragma_database_list order by seq;" return R #--------------------------------------------------------------------------------------------------------- _path_of_schema: ( schema, fallback = misfit ) -> R = @first_value @query "select file from pragma_database_list where name = ?;", [ schema, ] return R if R? return fallback unless fallback is misfit throw new E.Dba_schema_unknown '^dba@311^', schema #--------------------------------------------------------------------------------------------------------- type_of: ( cfg ) -> @types.validate.dba_type_of_cfg ( cfg = { @types.defaults.dba_type_of_cfg..., cfg..., } ) { name, schema, } = cfg if name in [ 'sqlite_schema', 'sqlite_master', ] return 'table' row = @first_row @query SQL""" select type from #{@sql.I schema}.sqlite_schema where name = $name limit 1;""", { name, } throw new E.Dba_object_unknown '^dba@311^', schema, name unless row? return row.type #--------------------------------------------------------------------------------------------------------- fields_of: ( cfg ) -> @types.validate.dba_fields_of_cfg ( cfg = { @types.defaults.dba_fields_of_cfg..., cfg..., } ) { name, schema, } = cfg schema_i = @sql.I schema R = {} for d from @query SQL"select * from #{schema_i}.pragma_table_info( $name );", { name, } # { cid: 0, name: 'id', type: 'integer', notnull: 1, dflt_value: null, pk: 1 } type = if d.type is '' then null else d.type R[ d.name ] = { idx: d.cid type: type optional: !d.notnull default: d.dflt_value is_pk: !!d.pk } return R #--------------------------------------------------------------------------------------------------------- field_names_of: ( cfg ) -> # try @types.validate.dba_field_names_of_cfg ( cfg = { @types.defaults.dba_field_names_of_cfg..., cfg..., } ) { name, schema, } = cfg schema_i = @sql.I schema return ( d.name for d from @query SQL"select name from #{schema_i}.pragma_table_info( $name );", { name, } ) # catch error # throw new E.Dba_sqlite_error '^dba@111^', error #--------------------------------------------------------------------------------------------------------- dump_relation: ( cfg ) -> @types.validate.dba_dump_relation_cfg ( cfg = { @types.defaults.dba_dump_relation_cfg..., cfg..., } ) { schema name order_by limit } = cfg schema_i = @sql.I schema qname_i = schema_i + '.' + @sql.I name limit = if cfg.limit is null then 1e9 else cfg.limit order_by ?= 'random()' return @query SQL"select * from #{qname_i} order by #{order_by} limit #{limit};" #--------------------------------------------------------------------------------------------------------- _dependencies_of: ( table, schema = 'main' ) -> return @query "pragma #{@sql.I schema}.foreign_key_list( #{@sql.I table} )" #--------------------------------------------------------------------------------------------------------- dependencies_of: ( table, schema = 'main' ) -> @types.validate.ic_schema schema return ( row.table for row from @_dependencies_of table ) #========================================================================================================= # DB STRUCTURE MODIFICATION #--------------------------------------------------------------------------------------------------------- ### TAINT Error: index associated with UNIQUE or PRIMARY KEY constraint cannot be dropped ### clear: ( cfg ) -> @types.validate.dba_clear_cfg ( cfg = { @types.defaults.dba_clear_cfg..., cfg..., } ) { schema, } = cfg schema_i = @sql.I schema R = 0 fk_state = @get_foreign_keys_state() @set_foreign_keys_state off for { type, name, } in @list @walk_objects { schema, _ordering: 'drop', } statement = "drop #{type} if exists #{@sql.I name};" @execute statement R += +1 @set_foreign_keys_state fk_state return R #--------------------------------------------------------------------------------------------------------- _open_file_db_in_ram: ( cfg ) -> ### Given a `path` and a `schema`, create a temporary schema to open the file DB in as well as an empty in-memory schema; then copy all DB objects and their contents from the temporary file schema to the RAM schema. Finally, detach the file schema. Ensure the `path` given is kept around as the `saveas` (implicit) path to be used for eventual persistency (`dba.save()`). ### ### TAINT validate? ### schema_main_allowed = not @_state.initialized { path, schema, saveas, } = cfg return @_attach { schema, path, saveas, } if @types.isa.dba_ram_path path #....................................................................................................... tmp_schema = @_get_free_temp_schema() @_attach { schema: tmp_schema, path, } unless ( schema is 'main' ) and schema_main_allowed @_attach { schema, path: '', saveas, } @_copy_schema { from_schema: tmp_schema, to_schema: schema, } @_detach { schema: tmp_schema, } #....................................................................................................... @_schemas = lets @_schemas, ( d ) => d[ schema ] = { path: saveas, } ### TAINT use API call ### return null #--------------------------------------------------------------------------------------------------------- _attach: ( cfg ) -> ### Given a `path` and a `schema`, execute SQL"attach $path as $schema". `_attach()` will fail * if `schema` already exists, or * if the maximum number of schemas (10 by default) has already been attached, or * if the schema name is `main` and the DBA is `@_state.initialized`. If `@_state.initialized` is `false`, then a new `better-sqlite3` instance with a `main` schema will be created; * if the `schema` passed in is `main`, it will be opened from the `path` given. * If `schema` is not `main`, `main` will be opened as an empty RAM DB, and `schema` will be attached from the file given. ### @types.validate.dba_attach_cfg ( cfg = { @types.defaults.dba_attach_cfg..., cfg..., } ) { path, schema, saveas, } = cfg #....................................................................................................... unless @_state.initialized if schema is 'main' ### TAINT code duplication from oneoff handler ### connection = new_bsqlt3_connection path, @_bsqlt3_cfg @initialize_sqlt connection guy.props.def @, 'sqlt', enumerable: false, configurable: false, value: connection @_state.initialized = true @_schemas = lets @_schemas, ( d ) => d[ schema ] = { path: saveas, } ### TAINT use API call ### return null ignore = @sqlt ### NOTE retrieve dynamic attribute for side effect, ignore its value ### #....................................................................................................... if @has { schema, } throw new E.Dba_schema_exists '^dba@312^', schema #....................................................................................................... try @run "attach ? as ?;", [ path, schema, ] catch error throw error unless error.code is 'SQLITE_ERROR' throw new E.Dba_sqlite_too_many_dbs '^dba@313^', schema if error.message.startsWith 'too many attached databases' throw new E.Dba_sqlite_error '^dba@314^', error @_schemas = lets @_schemas, ( d ) => d[ schema ] = { path: saveas, } ### TAINT use API call ### return null #--------------------------------------------------------------------------------------------------------- _detach: ( cfg ) -> @types.validate.dba_detach_cfg ( cfg = { @types.defaults.dba_detach_cfg..., cfg..., } ) @execute @sql.SQL"detach #{@sql.I cfg.schema};" @_schemas = lets @_schemas, ( d ) => delete d[ cfg.schema ] return null #========================================================================================================= # IN-MEMORY PROCESSING #----------------------------------------------------------------------------------------------------------- _move_schema: ( cfg ) -> @_copy_or_move_schema cfg, true _copy_schema: ( cfg ) -> @_copy_or_move_schema cfg, false #----------------------------------------------------------------------------------------------------------- _copy_or_move_schema: ( cfg, detach_schema = false ) -> detach_from_schema = -> return null unless detach_schema return @_detach { schema: from_schema, } #....................................................................................................... @types.validate.copy_or_move_schema_cfg ( cfg = { @types.defaults.copy_or_move_schema_cfg..., cfg..., } ) { from_schema, to_schema, } = cfg #....................................................................................................... if from_schema is to_schema throw new E.Dba_schema_repeated '^dba@315^', from_schema #....................................................................................................... known_schemas = @list_schema_names() throw new E.Dba_schema_unknown '^dba@316^', from_schema unless from_schema in known_schemas throw new E.Dba_schema_unknown '^dba@317^', to_schema unless to_schema in known_schemas #....................................................................................................... to_schema_objects = @list @walk_objects { schema: to_schema, } if to_schema_objects.length > 0 throw new E.Dba_schema_nonempty '^dba@318^', to_schema #....................................................................................................... from_schema_objects = @list @walk_objects { schema: from_schema } return detach_from_schema() if from_schema_objects.length is 0 #....................................................................................................... to_schema_x = @sql.I to_schema from_schema_x = @sql.I from_schema inserts = [] fk_state = @get_foreign_keys_state() @set_foreign_keys_state off #....................................................................................................... for d in from_schema_objects continue if ( not d.sql? ) or ( d.sql is '' ) continue if d.name in [ 'sqlite_sequence', ] #..................................................................................................... ### TAINT consider to use `validate.ic_db_object_type` ### unless d.type in [ 'table', 'view', 'index', ] throw new E.Dba_unexpected_db_object_type '^dba@319^', d.type, d #..................................................................................................... ### TAINT using not-so reliable string replacement as substitute for proper parsing ### name_x = @sql.I d.name sql = d.sql.replace /\s*CREATE\s*(TABLE|INDEX|VIEW)\s*/i, "create #{d.type} #{to_schema_x}." #..................................................................................................... if sql is d.sql throw new E.Dba_unexpected_sql '^dba@320^', d.sql #..................................................................................................... @execute sql if d.type is 'table' inserts.push "insert into #{to_schema_x}.#{name_x} select * from #{from_schema_x}.#{name_x};" #....................................................................................................... @execute sql for sql in inserts @set_foreign_keys_state fk_state @pragma "#{@sql.I to_schema}.foreign_key_check;" if fk_state return detach_from_schema() #========================================================================================================= # RANDOM NUMBER GENERATION # seedable for testing purposes #--------------------------------------------------------------------------------------------------------- ### To obtain a class with a seedable PRNG that emits repeatable sequences, define class property `@_rnd_int_cfg: { seed, delta, }` where both seed and delta can be arbitrary finite numbers. **NOTE** very small `delta` values (like 1e-10) may cause adjacent numbers to be close together or even repeat. To use default values for both parameters, set `@_rnd_int_cfg: true`.### @_rnd_int_cfg: null _initialize_prng: -> clasz = @constructor if clasz._rnd_int_cfg? seed = clasz._rnd_int_cfg.seed ? 12.34 delta = clasz._rnd_int_cfg.delta ? 1 @_rnd_int = CND.get_rnd_int seed, delta else @_rnd_int = CND.random_integer.bind CND return null #--------------------------------------------------------------------------------------------------------- _get_connection_url: ( dbnick = null ) => ### TAINT rename `dbnick` to `dbnick` ### ### Given an optional `dbnick`, return an object with the `dbnick` and the `url` for a new SQLite connection. The url will look like `'file:your_name_here?mode=memory&cache=shared` so multiple connections to the same RAM DB can be opened. When `dbnick` is not given, a random dbnick like `_icql_6200294332` will be chosen (prefix `_icql_`, suffix ten decimal digits). For testing, setting class property `@_rnd_int_cfg` can be used to obtain repeatable series of random names. ### dbnick ?= "_icql_#{@_rnd_int 1_000_000_000, 9_999_999_999}" url = "file:#{dbnick}?mode=memory&cache=shared" return { url, dbnick, }