[darcs-devel] Running darcs in GHCi

[Wim Lewis]

On Feb 8, 2007, at 8:13 PM, Zachary P. Landau wrote:
> That being said, I still wouldn't mind if I could load darcs through
> ghci when testing.  Is that just a pipe dream, given the dependencies
> that darcs pulls in and whatnot, or have others gotten it to work with
> recent versions?

Well, it seemed to me that it should work, since ghci does  
dynamically load ffi objects (as well as ghc-compiled haskell  
modules). After fiddling with it a little bit I've managed to run  
darcs in ghci. It works well enough to do a "darcs pull", incorporate  
some patches from darcs-unstable, then :reload to rerun darcs half- 
compiled-half-interpreted. I haven't tried anything more complicated.

The key seems to be to specify all the C extension modules on the  
command-line along with darcs.lhs. (But not rts.o.) The other haskell  
modules will get pulled in automatically.

On my system (MacOSX-ppc) it's also necessary to add a compiler flag  
to modify some bizarre behavior of Apple's gcc port. Adding CFLAGS='- 
mmacosx-version-min=10.4' when running ./configure will transform  
some symbols from _snprintf$LDBLStub (which statically-links OK, but  
breaks ghci's dynamic linking) into _snprintf$LDBL128 (which works  
fine either statically or dynamically).

After configuring with that flag added, I run 'make' to produce .o  
files for all the C modules (as well as the Haskell modules, of  
course... ghci seems happy to use them) and then

    ghci -cpp  -I. -DHAVE_CURSES -DHAVE_CURL  -package parsec -optl- 
lcurl -optl-lssl -optl-lcrypto -optl-lz -optl-lcurses  darcs.lhs  
maybe_relink.o atomic_create.o hscurl.o umask.o c_compat.o fpstring.o

will load everything into ghci and leave you at the *Main> prompt.