Re: Representing / Storing Text As a Unique Number Composed Of Prime Products

I should also clarify that storing text was meant to refer to storing
text as data for a computer. If you store text as Extended ASCII, it
requires 8 bits. If you store it as Unicode, it requires 16 bits.
Since most operating systems are 32 bits or 64 bits, I am not exactly
sure how text is stored on these systems. That is, how is a single
letter stored on such a system? On a 32 bit system, would a single
text letter be stored in a single 32 bit address space? How about on a
64 bit system?

Data compression might have some relevance with respect to these
answers. For example, if for some reason, a 32 bit operating system
ends up storing a text letter in a 32 bit address space, 2^32 =
4294967296, then both diet and edit could be stored in the space of one
letter:

2^32 = 4294967296
diet = 11*313*641*1109 = 2447521967
edit = 13*281*659*1109 = 2669725643

Of course, in a 64 bit system, if a single letter is stored in a 64 bit
address space, you could store substantially more words (although it
does not take many letters to exceed even a 64 bit space).

If text is not stored in this manner on such systems, then, obviously,
compression would, again, not be relevant.



Keith A. Lewis wrote:
> "rer" <reriker@xxxxxxxxxxxx> writes in article <1123822039.227407.206140@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> dated 11 Aug 2005 21:47:19 -0700:
> >Using the table (at http://www.maplenet.net/~reriker/primetext.html )
> >it is possible to formulate a unique number to represent the following
> >words:
> >
> >diet = 11*313*641*1109 = 2447521967
> >
> >edit = 13*281*659*1109 = 2669725643
> >
> >tide = 73*313*1013*1019 = 23585811703
> >
> >Any English word could be represented as a unique number in a manner
> >similar to this. And, it would be a relatively easy to reverse the
> >process -- generating the word from the number. Perhaps there is no
> >advantage to doing this. The original thought was that it might have
> >some potential for data compression and/or data encryption. Given that
> >the numbers become large rather quickly with the increasing length of
> >the word, it would appear that the latter (encryption) might be more
> >likely than the former (compression).
>
> I wouldn't call it compression because it (always!) gives a larger result
> than the straightforward base-26 representation.
>
> diet = 3*26^0 + 7*26^1 + 4*26^2 + 19*26^3 = 336833
>
> Because the process is easy to reverse, it cannot be used for public-key
> encryption, only shared-key. For example, the shared key could be a
> permutation of the first 26*n primes. But there are many shared-key
> algorithms that are more space-efficient...
>
> So even though the coding works, I don't see a useful app for it.
>
> p.s. Why exclude 2 from your list of primes?
>
> --Keith Lewis klewis {at} mitre.org
> The above may not (yet) represent the opinions of my employer.

.

Quantcast