{"id":371,"date":"2024-10-08T18:29:27","date_gmt":"2024-10-08T18:29:27","guid":{"rendered":"https:\/\/permutationcity.co.uk\/bp\/?p=371"},"modified":"2024-10-01T19:01:17","modified_gmt":"2024-10-01T19:01:17","slug":"the-law-of-demeter","status":"publish","type":"post","link":"https:\/\/permutationcity.co.uk\/bp\/2024\/10\/08\/the-law-of-demeter\/","title":{"rendered":"The Law of Demeter"},"content":{"rendered":"\n

I’ve been doing some more reading and came across the\nLaw of Demeter<\/a> and train wrecks again.\nThe last time while while reading\nClean Code<\/a>.<\/p>\n\n\n\n

I’ve one book that tells me this is bad practice:<\/p>\n\n\n\n

    const auto path = context.GetOptions().GetScratchDirectory().GetAbsolutePath();<\/pre><\/div>\n\n\n\n
And that I should do this instead:<\/p>\n\n\n\n
    const auto options = context.GetOptions();\n    const auto directory = options.GetScratchDirectory();\n    const auto path = directory.getAbsolutePath();<\/pre><\/div>\n\n\n\n
My feeling was to prefer the first as more compact here and\nit avoids turning a simple context wrapper into a heavyweight class.\nNow another book that tells me that both<\/em> are bad.\nIt doesn’t feel like there was too much to back either up.\nI’ll dig in to it.<\/p>\n\n\n\n
Loose coupling and the law<\/h2>\n\n\n\n
The law is a specific case of loose coupling<\/a>.\nThe general goal with loose coupling is to ensure that components make use of little or\nno internal knowledge about other components.\nThat allows each component to be changed with little or no impact on other components.\nWith Demeter it’s specifically about a method M<\/em> within a class C<\/em> and what it can interact with:<\/p>\n\n\n\n