> As you will refer to the encoding rule for UTF-8,
> '?' and '*' must not appear in the #2, #3 and surrogate byte.
Good to know, thanks for the info, that makes the issue less bad.
> As for filename globbing operation, String.getBytes("UTF-8")
> will be enought.
Why are you against String.toCharArray() ? Do you fear about
performance when dealing with 16-bit chars? Personally, I
do not think that on modern machines the 16-bit operations
are any slower than 8-bit. Some may even be slower when iterating
over a packed byte array.
And, getting the 16-bit char array out of the String is a
very simple operation. Actually, I'd propose doing both
isPattern() and unquote() in a single method call, encaptured
in a local static class - then a single iteration over the
String is sufficient:
private static class Unquoter {
boolean isPattern = false;
boolean wasQuoted = false;
StringBuffer buf;
public Unquoter(String s) {
buf = new StringBuffer(s.length());
for(int i=0; i<s.length(); i++) {
}
}
public boolean isPattern() { return isPattern; }
public boolean wasQuoted() { return wasQuoted; }
public String getUnQuoted() { return buf.toString(); }
}
Cheers,
--
Martin Oberhuber
Wind River Systems, Inc.
Target Management Project Lead, DSDP PMC Member
http://www.eclipse.org/dsdp/tm
> -----Original Message-----
> From: Atsuhiko Yamanaka [mailto:[EMAIL PROTECTED]
> Sent: Thursday, September 27, 2007 4:47 AM
> To: Oberhuber, Martin
> Cc: jsch-users@lists.sourceforge.net
> Subject: Re: [JSch-users] Jsch bug: ChannelSftp cannot read
> contents ofdirectory with '?' in its name
>
> Hi,
>
> +-From: "Oberhuber, Martin" <[EMAIL PROTECTED]> --
> |_Date: Wed, 26 Sep 2007 18:02:08 +0200 _______________________
> |
> |Given a remote UNIX directory with an '?' or '*' in its name,
> |for instance /parent/a?b
> |Even if the path name is quoted ("/parent/a\\?b"), calling
> |ChannelSftp.ls(quotedDir) does not return the contents of
> |dir, but it returns the list of file which reside in
> |parent/ and match the pattern a?b.
> |The reason for this problem is that
> | ChannelSftp.isPattern()
> |just checks for occurrence of the characters '?' and '*',
> |but does not check whether these characters have been quoted
> |with a \ or not.
>
> You are right. There are
> ChannelSftp#isPattern(String path) and
> ChannelSftp#isPattern(byte[] path)
> ,but the former hamethod s not cheked carefully. It should be fixed.
>
> |A slightly related problem to this is that unquoting of
> |path names is done in the encoded byte[] array rather
> |than the Unicode String. Given remote path names that
> |are encoded in UTF-8 or similar, there can be characters
> |encoded as 2 or three bytes. It's possible that byte
> |#2 or byte #3 happen to look like a '?' or '*' character,
> |but are actually not.
>
> As you will refer to the encoding rule for UTF-8,
> '?' and '*' must not appear in the #2, #3 and surrogate byte.
> On the other hand, there are such possibilties in other encoding;
> for example, Shift_JIS, CP932(widely used in Japan).
>
> |Therefore, the unquoting must happen in the Unicode string
> |rather than the encoded byte[] array -- Util.unquote()
> |must use String.toCharArray() and work on that one,
> |rather than String.getBytes().
>
> As for filename globbing operation, String.getBytes("UTF-8")
> will be enought.
>
> Thank you for your feedback.
>
> Sincerely,
> --
> Atsuhiko Yamanaka
> JCraft,Inc.
> 1-14-20 HONCHO AOBA-KU,
> SENDAI, MIYAGI 980-0014 Japan.
> Tel +81-22-723-2150
> +1-415-578-3454
> Fax +81-22-224-8773
> Skype callto://jcraft/
>
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
JSch-users mailing list
JSch-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jsch-users
