Changes between Version 55 and Version 56 of Records

Feb 24, 2012 1:29:09 AM (4 years ago)

add back Chris Done's project explanation


  • Records

    v55 v56  
    112112I couldn't find great specific information on record implementation ML variants. Best I can tell, SML does not allow records in the same module with the same field. Records from other modules require name-spacing or must be opened up similar to Agda.
     116=== Problems with using the current module namespace mechanism ===
     118Suppose I have 112 hand-crafted data types in my project (e.g. see attachment 51369.txt), this creates a lot of conflicts in field names and constructor names. For example:
     121data Comment = Comment {
     122      commentId           :: CommentId
     123    , commentContent      :: Content
     124    , commentReviewId     :: ReviewId
     125    , commentSubmissionId :: SubmissionId
     126    , commentConferenceId :: ConferenceId
     127    , commentDate         :: ISODate
     128    , commentReviewerNumber :: Int
     129  } deriving (Show)
     132This is a real type in my project. It has fields like “id”, “content”, “reviewId”, “submissionId”, “date”. There are seven other data types that have a field name “submissionId”. There are 15 with “conferenceId”. There are 7 with “content”. And so on. This is just to demonstrate that field clashes do occur a lot in a nontrivial project.
     134It also demonstrates that if you propose to put each of these 112 types into a separate module, you are having a laugh. I tried this around the 20 type mark and it was, apart from being very slow at compiling, very tedious to work with. Creating and editing these modules was a distracting and pointless chore.
     136It also demonstrated, to me, that qualified imports are horrible when used on a large scale. It happened all the time, that'd I'd import, say, 10 different data types all qualified. Typing map ( . BarMu?.thisField) and foo Bar.Zot{x=1,y=2} becomes tedious and distracting, especially having to add every type module when I want to use a type. And when records use other types in other modules, you have a lot of redundancy. With the prefixing paradigm I'd write fooId and barMuThisField, which is about as tedious but there is at least less . confusion and no need to make a load of modules and import lines. Perhaps local modules would solve half of this problem. Still have to write “ bar” rather than “mu bar”, but it'd be an improvement.
     138I also have 21 Enum types which often conflict. I end up having to include the name of the type in the constructor, or rewording it awkwardly. I guess I should put these all in separate modules and import qualified, too. Tedious, though. At least in this case languages like C# and Java also require that you type EnumName.EnumValue, so c‘est la vie.