> How should I go about doing this in perl?  Obviously I don't want to
> read the whole thing into memory, and I need to stay away from dodgy
> functions that turn hex into ascii.

> If anybody's interested, it is 8 million digits of pi, in hex form (2
> digits per byte).  I already converted from ASCII into that packed
> format, but now I need to search it.

> There's a slow PHP version if you care.  http://ericjiang.co.nr/pi.php

> How should I go about doing this in perl?  Obviously I don't want to
> read the whole thing into memory, and I need to stay away from dodgy
> functions that turn hex into ascii.

> If anybody's interested, it is 8 million digits of pi, in hex form (2
> digits per byte).  I already converted from ASCII into that packed
> format, but now I need to search it.

> There's a slow PHP version if you care.  http://ericjiang.co.nr/pi.php

> this sounds cool, I bought a book on pi, from B&N, and it had I think 10000
> digits of pi.

>>I need a Perl script to search through a file to find specific hex
>>strings, NOT ASCII.  Ideally,  it should not read more of the file than
>>it needs, and to quit when it does find it. The file is about 4 mb, with
>>8 million hex digits.

>>How should I go about doing this in perl?  Obviously I don't want to
>>read the whole thing into memory, and I need to stay away from dodgy
>>functions that turn hex into ascii.

> look at the functions sysopen and sysread
> this will do the trick. on sysread you can specify how many bytes to read.

> How should I go about doing this in perl?  Obviously I don't want to
> read the whole thing into memory, and I need to stay away from dodgy
> functions that turn hex into ascii.

> If anybody's interested, it is 8 million digits of pi, in hex form (2
> digits per byte).  I already converted from ASCII into that packed
> format, but now I need to search it.

>>I need a Perl script to search through a file to find specific hex
>>strings, NOT ASCII.  Ideally,  it should not read more of the file
>>than it needs, and to quit when it does find it. The file is about 4
>>mb, with 8 million hex digits.

>>How should I go about doing this in perl?  Obviously I don't want to
>>read the whole thing into memory, and I need to stay away from dodgy
>>functions that turn hex into ascii.

>>If anybody's interested, it is 8 million digits of pi, in hex form (2
>>digits per byte).  I already converted from ASCII into that packed
>>format, but now I need to search it.

> If it's 2 digits per byte, it's not in hex form.  Hexadecimal is a
> (typically ASCII) text representation of base-16 numbers, using the
> characters '0'..'9', 'a'..'f'.  It typically takes 8 bits per
> hexadecimal digit.

> The format you're describing sounds like plain binary.  Or is it
> binary-coded decimal?

> If it's binary-coded decimal, it's going to look something like this:

>     0011 0001 0100 0001 0101 1001 ...
>     (3)  (1)  (4)  (1)  (5)  (9)  ...

> so.. (3.) 14159265 would be

> 14 15 92 65
> when viewed in a hex editor.

> So what's the best function to search for strings of numbers in that,
> asides from turning it into an ascii string in memory?  What kind of
> functions does Perl have that would be suited to that?  I searched
> perl's manual, but didn't find anything.

> Or would it be better to just expand the whole thing into human-readable
> ASCII string in memory?  It seems like that would take all the advantage
> out of it.

> > If you mean you want to search for a string of digits like 14159265
> > and there may be an odd number of digits, or you want to find them
> > whether they start at an odd or even position, you have enormously
> > complicated things by packing your data that way.

> "Enormously"???  I doubt it.  Fixed strings are easy:

>   /\x14\x15\x92\x65/ || /[\x01...\xF1]\x41\x59\x26[\x50-\x5F]/

> (of course, this leading [\x01...\xF1] should be completely spelled out).

> What *is* enormously complicated is looking for regular expressions in
> the expanded string - without expansion...