It seems pcre new build is quite crufty. Comments? Looks like there should be a word_byteflip and dword_byteflip flavor. Also seems like a very endian-sensistive operation - hope our autodetects work properly.
Bill pcre.c G:\clean\httpd-2.1\srclib\pcre\pcre.c(622) : warning C4244: '=' : conversion from 'long ' to 'unsigned short ', possible loss of data G:\clean\httpd-2.1\srclib\pcre\pcre.c(623) : warning C4244: '=' : conversion from 'long ' to 'unsigned short ', possible loss of data G:\clean\httpd-2.1\srclib\pcre\pcre.c(624) : warning C4244: '=' : conversion from 'long ' to 'unsigned short ', possible loss of data G:\clean\httpd-2.1\srclib\pcre\pcre.c(625) : warning C4244: '=' : conversion from 'long ' to 'unsigned short ', possible loss of data G:\clean\httpd-2.1\srclib\pcre\pcre.c(627) : warning C4244: '=' : conversion from 'long ' to 'unsigned short ', possible loss of data G:\clean\httpd-2.1\srclib\pcre\pcre.c(629) : warning C4244: '=' : conversion from 'long ' to 'unsigned short ', possible loss of data G:\clean\httpd-2.1\srclib\pcre\pcre.c(630) : warning C4244: '=' : conversion from 'long ' to 'unsigned short ', possible loss of data study.c G:\clean\httpd-2.1\srclib\pcre\study.c(444) : warning C4090: 'function' : different 'const' qualifiers G:\clean\httpd-2.1\srclib\pcre\study.c(444) : warning C4022: 'pcre_fullinfo' : pointer mismatch for actual parameter 4
