Anyone know if there's a hard limit for FTP url/file paths and, if so, what it is?

Cheers

Matthew

In Internet Explorer it's 2,083 Characters.

Zz

"Matthew" <matthew.hat@stlsoft.dot.org> wrote in message news:c9gbm2$1ovo$1@digitaldaemon.com...
> Anyone know if there's a hard limit for FTP url/file paths and, if so,
what it
> is?
>
> Cheers
>
> Matthew
>
>

In c++ Matthew <matthew.hat@stlsoft.dot.org> wrote:
> Anyone know if there's a hard limit for FTP url/file paths and, if so, what it is?

If you're talking about a URL, none that I've been able to find. Same goes for the FTP RFC.

So, no, probably not, except if your OS defines an upper limit on path names.

Looking at the question again for Internet Explorer its actually a maximum
URL length of 2083 and a maximum path length of 2048.
As far as I know there is no specified limit and some other products have
different length.

Zz

"Matthew" <matthew.hat@stlsoft.dot.org> wrote in message news:c9gbm2$1ovo$1@digitaldaemon.com...
> Anyone know if there's a hard limit for FTP url/file paths and, if so,
what it
> is?
>
> Cheers
>
> Matthew
>
>

Yeah, I've decided to not count any chickens.

Thanks to you all for your prompt answers.

:)

Matthew

"Scott Michel" <scottm@mordred.cs.ucla.edu> wrote in message news:c9gc5p$1p2d$1@digitaldaemon.com...
> In c++ Matthew <matthew.hat@stlsoft.dot.org> wrote:
> > Anyone know if there's a hard limit for FTP url/file paths and, if so, what
it
> > is?
>
> If you're talking about a URL, none that I've been able to find. Same goes for the FTP RFC.
>
> So, no, probably not, except if your OS defines an upper limit on path names.

Here's an interesting twist. The doc for FtpGetCurrentDirectory() says

    "Using a length of MAX_PATH is sufficient for all paths."



"Matthew" <matthew.hat@stlsoft.dot.org> wrote in message news:c9ghko$20n2$1@digitaldaemon.com...
> Yeah, I've decided to not count any chickens.
>
> Thanks to you all for your prompt answers.
>
> :)
>
> Matthew
>
> "Scott Michel" <scottm@mordred.cs.ucla.edu> wrote in message news:c9gc5p$1p2d$1@digitaldaemon.com...
> > In c++ Matthew <matthew.hat@stlsoft.dot.org> wrote:
> > > Anyone know if there's a hard limit for FTP url/file paths and, if so, what
> it
> > > is?
> >
> > If you're talking about a URL, none that I've been able to find. Same goes
for
> > the FTP RFC.
> >
> > So, no, probably not, except if your OS defines an upper limit on path names.
>
>

"Matthew" <matthew.hat@stlsoft.dot.org> wrote in message news:c9gbm2$1ovo$1@digitaldaemon.com...
> Anyone know if there's a hard limit for FTP url/file paths and, if so,
what it
> is?

The hard limit is when your stack buffer overflows and a virus gets written onto it.

Yes, very good. ;)

In fact, I've fixed up various assumptions of path length in STLSoft and recls, and there are now the classes basic_file_path_buffer in both WinSTL and UNIXSTL, which provide the appropriate sized buffers. For WinSTL, if the OS is NT, then it provides a buffer of 32001, 9x it provides 261. For UNIXSTL, if PATH_MAX is defined, then the length is 1 + PATH_MAX, otherwise it calls pathconf().

For InetSTL it's not assuming any length at all for most of the code, although things that deal with FTP directory lengths do assume _MAX_PATH, since that's documented in the WinInet API.

"Walter" <newshound@digitalmars.com> wrote in message news:c9ilrl$22iv$1@digitaldaemon.com...
>
> "Matthew" <matthew.hat@stlsoft.dot.org> wrote in message news:c9gbm2$1ovo$1@digitaldaemon.com...
> > Anyone know if there's a hard limit for FTP url/file paths and, if so,
> what it
> > is?
>
> The hard limit is when your stack buffer overflows and a virus gets written onto it.
>
>

Yes, very good. ;)

In fact, I've fixed up various assumptions of path length in STLSoft and recls, and there are now the classes basic_file_path_buffer in both WinSTL and UNIXSTL, which provide the appropriate sized buffers. For WinSTL, if the OS is NT, then it provides a buffer of 32001, 9x it provides 261. For UNIXSTL, if PATH_MAX is defined, then the length is 1 + PATH_MAX, otherwise it calls pathconf().

For InetSTL it's not assuming any length at all for most of the code, although things that deal with FTP directory lengths do assume _MAX_PATH, since that's documented in the WinInet API.

btw, the next version of recls, which I'm working on right now, will allow FTP searches on Win32. (I'm using ftp.digitaldaemon.com as my test site <g>) I'll naturally be updating the D mapping.

"Walter" <newshound@digitalmars.com> wrote in message news:c9ilrl$22iv$1@digitaldaemon.com...
>
> "Matthew" <matthew.hat@stlsoft.dot.org> wrote in message news:c9gbm2$1ovo$1@digitaldaemon.com...
> > Anyone know if there's a hard limit for FTP url/file paths and, if so,
> what it
> > is?
>
> The hard limit is when your stack buffer overflows and a virus gets written onto it.
>
>

The best approach is to do one's damdest in writing the code to avoid any need to even know the maximum path length.

"Matthew" <matthew.hat@stlsoft.dot.org> wrote in message news:c9is50$2bft$1@digitaldaemon.com...
> Yes, very good. ;)
>
> In fact, I've fixed up various assumptions of path length in STLSoft and
recls,
> and there are now the classes basic_file_path_buffer in both WinSTL and
UNIXSTL,
> which provide the appropriate sized buffers. For WinSTL, if the OS is NT,
then it
> provides a buffer of 32001, 9x it provides 261. For UNIXSTL, if PATH_MAX
is
> defined, then the length is 1 + PATH_MAX, otherwise it calls pathconf().
>
> For InetSTL it's not assuming any length at all for most of the code,
although
> things that deal with FTP directory lengths do assume _MAX_PATH, since
that's
> documented in the WinInet API.
>
> btw, the next version of recls, which I'm working on right now, will allow
FTP
> searches on Win32. (I'm using ftp.digitaldaemon.com as my test site <g>)
I'll
> naturally be updating the D mapping.