A truly excelent new module has just appeared on CPAN, named Win32::StreamNames. It returns a list of file stream names for a specified file.
OK, I'll come clean. Yes it's one of mine. I was proudly showing off my blanket trick (see below) to another QA instructor when he said "I suppose you use a stream". Well, aside from a small river, a stream to me is an IO channel, or maybe a Unix device driver STREAM. This is neither. Enhanced at NTFS5 at Windows NT 5, correction, Windows 2000 ;-), a Windows File object points to Stream control blocks, which can include a Named stream. The names are of the form file-name:stream-name:$DATA. They are not normally visible to Windows Explorer, the dir command, Perl globbing, or readdir (using Win32 API). BackupRead Win32 API can retrieve the stream names, although it is not easy to use. Hence the module.
See it at your local CPAN mirror today! http://cpan.search.org
Wednesday, August 31, 2005
Friday, August 19, 2005
Blanket
While on the train, I decided to ammuse myself by writing a version of Damian Conway's 'bleach' program. Basically it "encrypts" a perl script with whitespace, so when it is viewed it looks blank.
The method I used (I couldn't remember Damian's exact code) is to take the ordinal number of each character in the character set, and write that number of space characters to the file.
For example, 'A' is decimal 65 in the character set (ISO Latin 1 and ASCII), so for that character I write a line contining 65 spaces. To get the number I use the Perl built-in function 'ord'. To decode I just have to get the length of each line and feed that into the 'chr' function to get back the original.
I showed this to my class as an ammusement, and one delegate (thank-you Daniel) played around with it. He discovered that if you zip the blanked file you get a very small archive. Smaller (about two-thirds) the size of the original, zipped. Must be the way the zip algorithm works with repeated characters.
I have just thought of another tweak I can do .............
The method I used (I couldn't remember Damian's exact code) is to take the ordinal number of each character in the character set, and write that number of space characters to the file.
For example, 'A' is decimal 65 in the character set (ISO Latin 1 and ASCII), so for that character I write a line contining 65 spaces. To get the number I use the Perl built-in function 'ord'. To decode I just have to get the length of each line and feed that into the 'chr' function to get back the original.
I showed this to my class as an ammusement, and one delegate (thank-you Daniel) played around with it. He discovered that if you zip the blanked file you get a very small archive. Smaller (about two-thirds) the size of the original, zipped. Must be the way the zip algorithm works with repeated characters.
I have just thought of another tweak I can do .............
Subscribe to:
Posts (Atom)