tcl search and replace - search

Question

I have an input file and I would like to do a search/replace and multiply and dump out the output file. How do I do that in TCL?
All the digit needs to be multiply by a multiplier of 10.
It needs to look for SECTION and END SECTION and find the word 'shape' and multiply all the digit with 10.
Input File
heading
size 9 XY 9
section1
shape name 1 2 3 4
end section1
section 2
shape name 1 2 3 4
end section2
Output file:
heading
size 90 XY 90
section1
shape name 5 10 15 20
end section1
section 2
shape name 100 200 300 400
end section2
tcl
set multiplier1 10
set multiplier2 5
set multiplier3 100
while {[gets $infile1] > 0} {
if {[regexp "size" $value]} {
}

Answer 1

Firstly, you'd be much better off defining the multipliers as an array. Using variable-named variables is usually a bad idea (unless you're about to upvar them). Also, remember that they're associative arrays, so you can use any string as an index and not just numbers; that's sometimes useful.
set multiplier(1) 10
set multiplier(2) 5
set multiplier(3) 100
Secondly, doing the multiplications for a line of numbers is best with a helper procedure:
proc ApplyMultiplies {line multiplier} {
set NUMBER_RE {-?\d+}
# For all locations of numbers, in *reverse* order
foreach location [lreverse [regexp -all -indices -inline -- $NUMBER_RE $line]] {
# Get the number
set value [string range $line {*}$location]
# Multiply it
set value [expr {$value * $multiplier}]
# Write it back into the string
set line [string replace $line {*}$location $value]
}
return $line
}
Testing that interactively:
% ApplyMultiplies {shape name 1 2 3 4} 5
shape name 5 10 15 20
% ApplyMultiplies "tricky_case\"123 yo" 17
tricky_case"2091 yo
In Tcl 8.7, you'll instead be able to do this as a one-liner because of the new -command option to regsub:
proc ApplyMultiplies {line multiplier} {
regsub -all -command -- {-?\d+} $line [list ::tcl::mathop::* $multiplier]
}
I do not understand the conditions under which you are deciding whether to apply the operation. Are the indices to multiplier meant to be section names, but are somehow a bit off? Why are we multiplying values on the size line? Without understanding that, writing the outer control code is impossible for me.