ASCII coding/decoding

[Chronos]

I have an app that uses a SQL database. For fields like adress and phone number, it encodes the inputs.

I put all chars from 'a' to 'z' to find out what it becomes. Here I copied a table with the ASCII input codes and the output codes:

Input Output

97 147

98 199

99 129

100 242

101 69

102 20

103 211

104 217

105 153

106 207

107 237

108 36

109 102

110 16

111 164

112 78

113 58

114 121

115 75

116 26

117 201

118 30

119 245

120 204

121 244

122 186

I can't use a 'SELECT CASE' function. I must find the algorithm.

Thanks a lot for the help.

[Mobius]

Chronos,

You would like a script that decodes SQL encoding back to its original input?

If you could feed in Lowercase a - z why couldn't you pass a full character string and then work from that?

Where must the script read the encoded data from? Gui, File?

And where should the script export decoded data?

[Chronos]

Chronos,

You would like a script that decodes SQL encoding back to its original input?

If you could feed in Lowercase a - z why couldn't you pass a full character string and then work from that?

Where must the script read the encoded data from? Gui, File?

And where should the script export decoded data?

Mobius,

yes, I would like a decoder script.

Actually the connections for reading/exporting data are not a problem, they're working just fine, so assume the encoded data are in a internal variable and the decoded data will be set in another.

About passing a full character string, the problem is that in the second position, for example, the same char has a different code, that also depends on the previous one. So I hope that by decoding the first position I may find a way to decode the others just by changing a constant in the formula. That's why I said before I can't just create a table and use a 'Select Case' procedure.

[Mobius]

Chronos,

ah, I see, sorry dude. silly suggestion.

An algorithm for this would be no small feat!

going to have a dabble anyway...

beyond me...

Without a table consisting of all or most First instance output codes,

I could not hope to identify a pattern. (One for the future...)

Edited August 13, 2008 by Mobius