from tawk to gawk - multidimensional arrays

[Cornelius19]

Hi,

I use gawk but need to run an old awk script, written for the tawk compiler (

http://www.tasoft.com/).

The syntax is slightly different, for instance, they declare multi-dimensional arrays like this:

ary[word][dialect][24] = "A";

This does not compile under gawk, but I can easily convert it to standard awk syntax and it compiles (I hope this is the correct translation):

ary[word,dialect,24] = "A";

I could correct most of the errors with such a change. However, there are still a couple of errors I do not know how to correct (basically where the above mentioned tri-dimensional array is used with less than 3 dimensions):

for (dialect in ary[word])

or

if (24 in ary[word,dialect] && substr($0,24,1)!="C")

Do you have any idea?

Cornelius

 

[feherke]

Hi

That is impossible in [tt]gawk[/tt]. There are only simulated multi-dimensional arrays, not real ones.

ary[word,dialect,24] = "A";

[gray]# is equivalent with[/gray]

ary[word SUBSEP dialect SUBSEP 24] = "A";

[gray]# and by default is equivalent with[/gray]

ary[word "\034" dialect "\034" 24] = "A";

As you can see, the multiple indexes are concatenated into one and the array is infact single dimensional.

I am afraid, the closest dialect to convert that [tt]tawk[/tt] script into, would be [tt]perl[/tt]. :-(

Feherke.

http://rootshell.be/~feherke/

 

[Cornelius19]

Thanks for the post. I understood that this is only a simulated multi-dimensional array and in reality it is a single dimensional array with indexes concatenated into a single one. However, in this case, I might not need multi-dimensional arrays, just brake the single index into substrings. Could not something like that work?

Cornelius

 

[feherke]

Hi

Nope, the [tt]in[/tt] operator will not work on those values.

Sorry, but I was serious about [tt]perl[/tt] as best alternative.

Feherke.

http://rootshell.be/~feherke/

 

[Cornelius19]

Thanks a lot Feherke. I will try to explore the perl option.

Cornelius