Parse txt file on basis of occurrence of a tag in Linux

I am trying to parse a txt file containing xml "messages"in linux, something like this

xyz     xyz     xyz     xyz   and so on

The code will read file, extract each section from

till

and put each section into a separate file. My code for this is as below

input_file="input_file.txt"

# Extracting Document parts
sed -n '//p' "$input_file" > temp_output.txt

# Splitting into Different Files
csplit -f output -b %d.txt -z temp_output.txt '//' '{*}'

# Cleaning up temporary files
rm temp_output.txt

However, this code is extracting several xml messages into one file, particularly the ones with no line break. Could someone suggest what can be rectified in the above code?