One of my earliest development projects circa 2002 was to create a complex awk & sed program that took tons of database blobs with inconsistent semi-structured test from a publishing company and converted it to XML, automatically prompting the user for input (using the Pico text editor) whenever there was something that couldn't be resolved. It saved thousands of hours as part of a metadata digitization project that got their journals into ProQuest.
As only a semi-literate programmer at the time, it was significantly easier than trying the same thing with a "traditional" programming language. On occasion I still call to awk withing python code when I need to do a one-off quick & dirty <something>
For those interested, if you can find a copy, there'a truly excellent book on awk:
"The Awk Programming Language", by Alfred Aho, Brian Kernighan, and Peter J. Weinberger