How do you split a paragraph or page into 2 variables using TCL?

Question

I know you can find a string and place everything after it into a variable using TCL and regex like the following

regexp "Shipping \(\[^\f]+)" $VarWithWholePage match shipinfo;

But is it possible to place everything before the string into a variable?

For example.

Sentence number 1.
Sentence number 2.
Sentence number 3.
Sentence number 4.
Shipping
Sentence number 5.

My example would place "Sentence number 5." into shipinfo , but I would like to be able to place

Sentence number 1.
Sentence number 2.
Sentence number 3.
Sentence number 4.

into another variable.

Answer 1

regexp "\(.*\)Shipping \(\[^\f]+)" $VarWithWholePage match before after;

should put the text before "Shipping" in before and the part after in after .

PS I am not sure why you are using ^\\f , but if you have a good reason you can use it also in the first subexpression.

Answer 2

But is it possible to place everything before the string into a variable?

The simplest way to do this is to get regexp to tell you the indices of the match instead of the matched substring. You can then use string range to get the parts you want.

regexp -indices "Shipping \(\[^\f]+)" $VarWithWholePage match shipinfo

At this point, match and shipinfo will have pairs of numbers indicating exactly where the match happened within the input string. If the match succeeded.

# Now we can get the bits before (and after) the match with simple operations
set beforeMatch [string range $VarWithWholePage 0 [expr {[lindex $match 0] - 1}]]
set afterMatch [string range $VarWithWholePage [expr {[lindex $match 1] + 1}] end]

# Convert the string ranges into the matched substrings so that your code still works
set match [string range $VarWithWholePage {*}$match]
set shipinfo [string range $VarWithWholePage {*}$shipinfo]

Footnote: your RE would be idiomatically written as {Shipping ([^\\f]+)} as it's usually a very good idea to put regular expressions in braces. It combats backslash-itis.