regex in tcl script

[apertosa]

Hi folks and thanks for reading.
I am a newbie in both TCL and REGEX (ugh...lots to learn), so please be forgiving.
I did my research but couldn't find a match (hehe!) of what i need to do.
The problem:
I have a file that I'm parsing line by line in TCL
The file looks like this:

'point1 ' 'left/right' 0.0 -200.0 225.0
'point2 ' 'left/right' -200.0 -400.0 225.0
...

-> a name, a string with / and three numerical values.

TCL allows me to directly call the regexp command.
I would like to store the 5 entities in appropriate variables: name, symmetry, x y and z.

   while {![eof $raw] } {
      gets $raw line
      if { [regexp \
         {  --a string---a string--- a signed float --- a signed float --- a signed float  }\
             $line name symmetry x y z ]} {
             puts $name
         }
   }
   close $raw

I have absolutely no idea on where to go from here. Can someone throw me a bone?
thanks!
Andrea

 

[apertosa]

With someone's help I was able to get what I wanted:
here is the solution:

if {[string compare $u sub] == 0} {
   set raw [open $file r]
   while {![eof $raw] } {
      gets $raw line
      if { [regexp \
         {^\s*'([^']*)'\s+'([^']*)'\s+(-?(?!0\d)\d+\.\d*)\s+(-?(?!0\d)\d+\.\d*)\s+(-?(?!0\d)\d+\.\d*)\s*$} \
            $line hp_name hp_symmetry x y z] } {
            puts $hp_name
            puts $hp_symmetry
            puts $x
            puts $y
            puts $z 
         }
      }
   }
   close $raw

 

[Bong]

regexp is very powerfull and I admit, it still gives me a thrill when I get the syntax right. That said, however, Tcl lists are a more direct way to do what you want:

[red]#the following line is as you had it[/red]
set raw [open $file r]
[red]#the next line puts every line in the file in a list[/red]
set linelist [split [read $raw] /n]
close $raw
[red]loop on each line in the list[/red]
foreach line $linelist {
   [red]#Tcl will treat a string as a list so this next is technically unnecessary[/red]
   set list1 [split $line " "]
   foreach elm $list1 {puts $elm}
}

_________________
Bob Rashkin

 

[apertosa]

OK. I follow you.
I used regex because (I think) it automatically sets the variable for the matches that it finds.
If I use your approach and do this:

set line " USAGE           =  'entity_name'"

set list1 [split $line " "]

and want to extract the string entity_name I have to remove trailing, middle and blank characters if I want to make sure that

lindex $list1 0
is USAGE and
lindex $list1 1 is =
etc...
right?

So my question is: is there a way to remove all the blank chars in one go?

Thanks!
Andrea.

 

[Bong]

Not so much, but it would be pretty much what you want if you omit the step I called superflous:

....
foreach line $linelist {
   foreach elm [red]$line[/red] {puts $elm}
}

_________________
Bob Rashkin