There's no flag to do this now, but we'll put something in the next release that is set by default in the Windows version.
For a work around, you can modify fileTable.c to downcase all the filenames. Here are the changes: 91a100 > /*!!!*/ abspath = cstring_downcase (abspath); 266c277,278 < cstringTable_insert (ft->htable, e->fname, ft->nentries); --- > /*!!! cstringTable_insert (ft->htable, e->fname, ft->nentries); */ > cstringTable_insert (ft->htable, cstring_downcase (e->fname), ft->nentries); (Its probably better to just run a script over your sources that canonicalizes the filenames though.) --- Dave On Tue, 23 Apr 2002, Martin Roehrig wrote: > Hi, > > I use splint 3.0.1.6 on Windows NT 4.0 SP6 (binary distribution from splint.org) and >get spurious redefinition warnings under the > following circumstances: > > ---------------- myheader.h -------------- > typedef int mytype; > ------------------------------------------ > > ---------------- mycode_1.c -------------- > #include "myheader.h" > ------------------------------------------ > > ---------------- mycode_2.c -------------- > #include "MYHEADER.H" > ------------------------------------------ > > > splint *.c > Splint 3.0.1.6 --- 11 Feb 2002 > > MYHEADER.H(1,13): Datatype mytype defined more than once > A function or variable is redefined. One of the declarations should use > extern. (Use -redef to inhibit warning) > myheader.h(1,13): Previous definition of mytype > > Finished checking --- 1 code warning > > > Obviously splint does not realize that myheader.h and MYHEADER.H is the same file on >Windows and therefore thinks mytype to be > redefined. > Of course I don't want to set -redef. On the other hand I have to cope with a large >software project where I just can not influence > how programmers write the include file names, so what can I do about that situation? >Is there a special switch? > > > Thanks in advance > Martin > > >
