#2776: Document -pgmL (Use cmd as the literate pre-processor)
---------------------------------+------------------------------------------
    Reporter:  Syzygies          |       Owner:                  
        Type:  bug               |      Status:  new             
    Priority:  normal            |   Component:  Documentation   
     Version:  6.10.1            |    Severity:  minor           
    Keywords:  pgmL literate     |    Testcase:                  
Architecture:  Unknown/Multiple  |          Os:  Unknown/Multiple
---------------------------------+------------------------------------------
 The GHC option -pgmL (Use cmd as the literate pre-processor) is
 undocumented, and behaves differently than -pgmF (Use cmd as the pre-
 processor).

 While one can count on the first three arguments to a -pgmF command giving
 file names, with optional arguments coming later, the ''last'' three
 arguments to a -pgmL command give the file names. An undocumented "-h"
 argument always comes ''before'' these file arguments, throwing off any
 command designed according to the documentation for -pgmF. Moreover, an
 optional argument provided using -opL comes ''before'' any of these
 arguments.

 The undocumented "-h" reminds me of the passage in Real World Haskell, how
 only one reasonable program could have type "(a, b) -> a", because there
 isn't enough information to do anything clever. Here, GHC wants the -pgmL
 command to emit standard Haskell, but for all it knows, the literate
 program is in Swahili. GHC isn't really in a good position to be offering
 advice via options on this conversion, because it has no idea what
 conversion is taking place. So the "-h" is inexplicable.

 Also, the example command in GHC manual section 5.10.4. doesn't use
 argument $1, and probably wants to use $1 rather than $2 for the echo, if
 in fact it is possible to create circumstance where $1 and $2 differ.

-- 
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/2776>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
_______________________________________________
Glasgow-haskell-bugs mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs

Reply via email to