This is the mail archive of the cygwin-apps@cygwin.com mailing list for the Cygwin project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re[2]: URL paths in setup.exe


Hello Robert,

Tuesday, May 07, 2002, 12:45:33 PM, you wrote:

RC> I'd like to formalise what file:// and cygfile:// schemes mean.

RC> file:// is a native filesystem URL handler - whatever the OS may be.
RC> cygfile:// is a handler that only makes sense on mingw platforms, and
RC> access's the cygwin mount table.

RC> This means that:
RC> file:///foo/bar.txt is /foo/bar.txt on posix, and Current
RC> drive:\foo\bar.txt on mingw.

I dont understand what you're saying here ? Please explain a little
bit more.

The URLParsers parses URLs, obviously, so it will parse anything in the
for <schema>://<authority>/path. Still currently it wont parse
if authority starts whith underscore - should I chage this ?

RC> As for file:// + d: + \foo\bar.txt, can we normalise that as 
RC> file://d|/foo/bar.txt - that is what MS do, and will be less confusing
RC> for users of the codebase (IMO).

I understand this. The URLParser class will also parse it without a
problem. The URLParsers has a convinient method called GetAbsPath
which will return the absolute path for a given schema making some
assumptions based on the schema i.e. if it the schema is file it will
return the path like that <authority>:\<path>..

1) I can remove this method - each io_stream class will have its own
means to get the path from the URL.

2) I can tweak the GetAbsPath so it assumes different from what it
assumes now.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]