I've to convert a linux dash script because it is too slow and i decded to do it in D. I'm totally new to this and i think it could be a good exercise to learn this language.

The shell script does some simple jobs like:
0) Run the script with some options
1) sed/grep regex to catch a portion of a file.
    For example: it finds the line that match "1234" and take all the lines until the line that match "abcd".
2) sed regex to catch some strings
    For example: "abc sdfs#=8 // some text" i've to take "8" and "some text"
3) Creates dirs and copy files
4) Add specific char to a specific column position at every row of a file
    Original file:
    abcdefghij
    1234567890
    c34vt59erj
    04jèoàòr4t
    14sdf7g784

    Edited file:
    ab;cde;f;g;hij
    12;345;6;7;890
    c3;4vt;5;9;erj
    04;jèo;à;ò;r4t
    14;sdf;7;g;784

I would like to know what could be the best approach i would have to take with D to write simple, elegant and fast code, scanning files with more than 3000+ columns per line.

Thank you,
Loris

On Monday, 16 November 2020 at 10:51:51 UTC, Bloris wrote:
> I've to convert a linux dash script because it is too slow and i decded to do it in D. I'm totally new to this and i think it could be a good exercise to learn this language.
>
> The shell script does some simple jobs like:
> 0) Run the script with some options
> 1) sed/grep regex to catch a portion of a file.
>     For example: it finds the line that match "1234" and take all the lines until the line that match "abcd".
> 2) sed regex to catch some strings
>     For example: "abc sdfs#=8 // some text" i've to take "8" and "some text"
> 3) Creates dirs and copy files
> 4) Add specific char to a specific column position at every row of a file
>     Original file:
>     abcdefghij
>     1234567890
>     c34vt59erj
>     04jèoàòr4t
>     14sdf7g784
>
>     Edited file:
>     ab;cde;f;g;hij
>     12;345;6;7;890
>     c3;4vt;5;9;erj
>     04;jèo;à;ò;r4t
>     14;sdf;7;g;784
>
> I would like to know what could be the best approach i would have to take with D to write simple, elegant and fast code, scanning files with more than 3000+ columns per line.
>
> Thank you,
> Loris

regex you can use std.regex module https://dlang.org/phobos/std_regex.html
IO stuff, read files, creatoe folders, etc: https://devdocs.io/d/std_stdio
I'm not sure about performance, if you find it be slow, maybe there are something better at https://code.dlang.org/ it also depends on your algorithm/code, of course