How can I colorize my help? 1

[jdespres]

I'm trying to create a colorful help for all the functions I'm creating...

I have the following ::--->

gethelp () {
cat << EOF
act ::---> Get one or more clients activity. act <Client_name or Client_list>
his ::---> Get one or more client backup history. his <Client_name or Client_list>
logs ::---> Get logs for one or more clients. logs <Client_name or Client_list>
cl ::---> Generate a client list. cl
gf ::---> Get clients failures from the past 2 days. gf
csv ::---> Generate a csv file containing all clients. csv
cc ::---> Generate a csv file for one or many clents. cc <Client_name or Client_list>
tbw ::---> Test Backup for a windows client. tbw <Client_name>
gln ::---> Get Label number for the last backup. gln <Client_name>
wgbu ::---> What got backed up. wgbu <label_number> </Domain/Client_name>
ptf ::---> Perform the test restore. ptf <label_number> </Domain/Client_name>
gip ::---> Get ip for one or many clients. gip <Client_name or Client_list>
cu ::---> Client usage for one or many clients. cu <Client_name or Client_list>
EOF
}

I would like to make the 1st column from the left bold..

And a separate color for the following ::--->

::--->
Everything up to the period
From the period to the end

Thanks.....

Joe Despres

 

[feherke]

Hi

This colorizes the text only if the output goes to the terminal. So if gets redirected, the plain version is output ( for example so you can [tt]grep[/tt] in it ) :

[COLOR=darkgoldenrod]gethelp()[/color]
{
  {
    [teal][[/teal] -t [purple]1[/purple] [teal]][/teal] [teal]&&[/teal]
    sed [green][i]'s/^\(\w\+\)\(\s\+\)\(::--->\)\(\s\+\)\([^.]*.\)\(\s\+\)\(.*\)$/[COLOR=lime black]^[[/color][31m\1[COLOR=lime black]^[[/color][0m\2[COLOR=lime black]^[[/color][32m\3[COLOR=lime black]^[[/color][0m\4[COLOR=lime black]^[[/color][33m\5[COLOR=lime black]^[[/color][0m\6[COLOR=lime black]^[[/color][34m\7[COLOR=lime black]^[[/color][0m/'[/i][/green] [teal]||[/teal]
    cat
  } [teal]<<[/teal] EOF
act ::---> Get one or more clients activity. act <Client_name or Client_list>
his ::---> Get one or more client backup history. his <Client_name or Client_list>
logs ::---> Get logs for one or more clients. logs <Client_name or Client_list>
cl ::---> Generate a client list. cl
gf ::---> Get clients failures from the past 2 days. gf
csv ::---> Generate a csv file containing all clients. csv
cc ::---> Generate a csv file for one or many clents. cc <Client_name or Client_list>
tbw ::---> Test Backup for a windows client. tbw <Client_name>
gln ::---> Get Label number for the last backup. gln <Client_name>
wgbu ::---> What got backed up. wgbu <label_number> </Domain/Client_name>
ptf ::---> Perform the test restore. ptf <label_number> </Domain/Client_name>
gip ::---> Get ip for one or many clients. gip <Client_name or Client_list>
cu ::---> Client usage for one or many clients. cu <Client_name or Client_list>
EOF
}

Tested with Bash, Dash and MKsh.

Unless this is the only place you need it, i would think to develop a generic way to do this.


Feherke.
[link feherke.github.com/]

http://feherke.github.com/

[/url]

 

[jdespres]

hhhhmmmmm

I must of copied wrong...

I'm getting a bunch of characters before and after..

Thanks!

Joe Despres

 

[feherke]

Hi

Maybe because I forgot to add the explanation : the [tt][COLOR=lime black]^[[/color][/tt] in the text are ESC ( [tt]\e[/tt], [tt]\033[/tt], [tt]\x1b[/tt] ) characters. You type them as [kbd]Ctrl[/kbd]-[kbd]v[/kbd],[kbd]Esc[/kbd]. So the "^[" is only the displayed image of the character, you can not get the character itself by by copy & pasting its displayed image. ( The way to type such characters may depend on the editor you use. For example in Midnight Commander's built-in editor you have to type [kbd]Ctrl[/kbd]-[kbd]q[/kbd],[kbd]Esc[/kbd],[kbd]Esc[/kbd]. )

To avoid typing ESC characters, you can use this code instead, but I am not sure whether is understood by other implementations than GNU [tt]sed[/tt] :

sed [green][i]'s/^\(\w\+\)\(\s\+\)\(::--->\)\(\s\+\)\([^.]*\.\)\(\s\+\)\(.*\)$/\o033[31m\1\o033[0m\2\o033[32m\3\o033[0m\4\o033[33m\5\o033[0m\6\o033[34m\7\o033[0m/'[/i][/green]

Or if you use Bash, you can use it to interpret the escape sequences :

sed -r $[green][i]'s/^(\\w+)(\\s+)(::--->)(\\s+)([^.]*\\.)(\\s+)(.*)$/\e[31m\\1\e[0m\\2\e[32m\\3\e[0m\\4\e[33m\\5\e[0m\\6\e[34m\\7\e[0m/'[/i][/green]

Feherke.
[link feherke.github.com/]

http://feherke.github.com/

[/url]

 

[PHV]

Why not simply use the tput command ?

Hope This Helps, PH.
FAQ219-2884
FAQ181-2886

 

[feherke]

Hi

Why not simply use the tput command ?

1) Personal habit since DOS times; 2) never found a good documentation; 3) speed consideration.

By the way, how would you use [tt]tput[/tt] in this case ? In the [tt]sed[/tt] code like I used the ANSI escapes or directly in the [tt]cat[/tt]ed text ?


Feherke.
[link feherke.github.com/]

http://feherke.github.com/

[/url]