- C2hs - Haskell.org

patch applied (c2hs): Allow attrs in most places in declarators
by Duncan Coutts 19 Apr '07

19 Apr '07

Thu Apr 19 05:05:42 PDT 2007 Duncan Coutts <duncan.coutts(a)worc.ox.ac.uk> * Allow attrs in most places in declarators in (..) parenthesis and after '*' pointers. M ./c2hs/c/CParser.y -3 +53

1 0

patch applied (c2hs): Allow attrs in decl qualifier/specifiers and after declarators.
by Duncan Coutts 19 Apr '07

19 Apr '07

Thu Apr 19 04:58:14 PDT 2007 Duncan Coutts <duncan.coutts(a)worc.ox.ac.uk> * Allow attrs in decl qualifier/specifiers and after declarators. This is the most common case. Note the lexer still doesn't pass attribute tokens through. M ./c2hs/c/CParser.y -14 +36

1 0

patch applied (c2hs): Rename maybe_asm -> asm_opt for naming consistency
by Duncan Coutts 15 Apr '07

15 Apr '07

Sun Apr 15 03:18:38 PDT 2007 Duncan Coutts <duncan.coutts(a)worc.ox.ac.uk> * Rename maybe_asm -> asm_opt for naming consistency M ./c2hs/c/CParser.y -4 +4

1 0

patch applied (c2hs): Shorten names of the GNU C __attribute__ non-terminals
by Duncan Coutts 15 Apr '07

15 Apr '07

Sun Apr 15 03:15:45 PDT 2007 Duncan Coutts <duncan.coutts(a)worc.ox.ac.uk> * Shorten names of the GNU C __attribute__ non-terminals Since we're going to be sprinkling them round the rest of the grammar it's nicer if they don't have such huge names. Also uncomment those non-terminals. M ./c2hs/c/CParser.y -22 +22

1 0

patch applied (c2hs): Rename terminal ' extension' to ' __extension__'
by Duncan Coutts 15 Apr '07

15 Apr '07

Sun Apr 15 03:14:27 PDT 2007 Duncan Coutts <duncan.coutts(a)worc.ox.ac.uk> * Rename terminal 'extension' to '__extension__' That's what's it's really called anyway. M ./c2hs/c/CParser.y -8 +8

1 0

patch applied (c2hs): Apparently ' \f' and ' \v' are allowed in C strings and char literals
by Duncan Coutts 15 Apr '07

15 Apr '07

Sun Apr 15 00:58:32 PDT 2007 Duncan Coutts <duncan.coutts(a)worc.ox.ac.uk> * Apparently '\f' and '\v' are allowed in C strings and char literals According to both gcc and the C99 spec. Some odd programs really use it too, like flex and binutils. M ./c2hs/c/CLexer.x -2 +2

1 0

patch applied (c2hs): Allow all unnamed struct members, not just unamed nested structs
by Duncan Coutts 15 Apr '07

15 Apr '07

Sun Apr 15 00:54:06 PDT 2007 Duncan Coutts <duncan.coutts(a)worc.ox.ac.uk> * Allow all unnamed struct members, not just unamed nested structs This does make the grammar too liberal but it's much easier to parse this more general grammar than trying to select just the unnamed nested structs. Unnamed members other than structs/unions can be checked in the AST later. It worked before, but once we introduce GNU C attributes it becomes virtually impossible to parse unambigiously. M ./c2hs/c/CParser.y -15 +7

1 0

patch applied (c2hs): Allow nested function definitions (a GNU C extension)
by Duncan Coutts 15 Apr '07

15 Apr '07

Sun Apr 15 00:48:20 PDT 2007 Duncan Coutts <duncan.coutts(a)worc.ox.ac.uk> * Allow nested function definitions (a GNU C extension) Just extends the things can can appear in compound statements, as well as statements and declarations we can have function definitions This GNU C extension is quite rarely used. M ./c2hs/c/CAST.hs -7 +22 M ./c2hs/c/CParser.y -9 +31

1 0

patch applied (c2hs): Deal with the scope of the ' for' loop' s decl properly
by Duncan Coutts 15 Apr '07

15 Apr '07

Sun Apr 15 00:13:42 PDT 2007 Duncan Coutts <duncan.coutts(a)worc.ox.ac.uk> * Deal with the scope of the 'for' loop's decl properly typedef int n; // n declared as a type name for (int n = ...; ...; ...) stmt; // n now considered a normal ident n n = 0; // n reverts to being a type name the declaration can define a local var (eg 'n') that masks a type of the same name from an outer scope. The local var goes out of scope at the end of the for loop's statement, if we don't do this then we treat 'n' as a normal ident rather than a typeident after the 'for' statement. M ./c2hs/c/CParser.y -2 +2

1 0

Re: [Haskell-cafe] more problems with c2hs
by Duncan Coutts 13 Apr '07

13 Apr '07

On Fri, 2007-04-13 at 02:37 +1000, Ruben Zilibowitz wrote: > Hi, > > I've built and installed c2hs on my Mac OS X system now. I read > through the research paper on c2hs available on the website for c2hs > and decided I would need to try it out to learn it. But it gives me > errors which I am having trouble trying to understand. Here are the > error messages: > > c2hs file.chs > > powerpc-apple-darwin8-gcc-4.0.1: c: No such file or directory > powerpc-apple-darwin8-gcc-4.0.1: c: No such file or directory > powerpc-apple-darwin8-gcc-4.0.1: warning: '-x -x' after last input > file has no effect > powerpc-apple-darwin8-gcc-4.0.1: no input files > c2hs: Error during preprocessing custom header file c2hs runs cpp -x c ${header} for the header file you specify. You can either specify a header file on the commend line or use #include's in the .chs file. c2hs should generate a .h file itself in that case but it looks like it isn't for some reason. You can find out what it is really doing by running: c2hs --dump trace file.chs btw, the better mailing list for this stuff is the c2hs mailing list: c2hs(a)haskell.org Duncan

2 8