"James Hawkins" truiken@gmail.com writes:
Hi,
Is there anything wrong with this patch?
It can't possibly work. Please test it properly before resending it once more.
On 7/12/06, Alexandre Julliard julliard@winehq.org wrote:
"James Hawkins" truiken@gmail.com writes:
Hi,
Is there anything wrong with this patch?
It can't possibly work. Please test it properly before resending it once more.
I did test it, and for the option I tested it does work. I should have done more thorough tests though, because looking more closely at why it might not be correct, I see that I don't need to advance the second string because we don't send in the slash or hyphen:
+ /* skip over possible hyphen or slash */ + return msi_strequal(str1 + 1, str2 + 1); << str2 + 1 should just be str2 ... - if(!msi_strequal(argvW[1], "/@")) + if(!msi_option_equal(argvW[1], "@")) << no hyphen or slash
It was working before because we call MultiByteToWideChar(...,str2,-1,...) which will return length 1 (null-terminator) for all options. We then try to compare str1 and str2W using the length value returned from MultiByteToWideChar minus 1 (zero) which always returns CSTR_EQUAL. I was testing to see if it worked with /i and -i, and that did the install correctly because the install option is the first case that uses msi_option_prefix. Any other value would cause the install too, which is why the testing was insufficient.