If the task can be fully automated, means there is a unique algorithm to transform a resource A to A' based on the content of A than you will look for available methods and tools supporting you to do this kind of operation.
Regular expressions are very powerful rules to express not only finding common pattern in text based resources but also a good foundation to replace or extend existing content.
Compared to simple phrase based pattern most imaginable rules can be expressed and used as a source for the required transformation.
But regular expressions come with high cost of complexity. It is very likely to defines rules which results in "false positives", means matches that you didn't want to match.
Doing Search & Replace in the file system on multiple resources (files) is easy for IT people using linux tools like grep, ....
On Windows you also can install those tools and make them a powerful foundation for those kind of operations (see http://gnuwin32.sourceforge.net/packages/grep.htm).
But not all people like to become an IT expert for simple replacing the term "foo" with "delicious" . On Windows you can use TextCrawler for this. One of the best UI based tools I'm aware of.
- simple phrase based operation "Replace phrase A with B" on multiplier files
- more complex regular expression based operation
- and in addition a fuzzy search operation for more advanced search operations
To avoid false positives you can
- preview the hits before actual performing the replace operation
- use a dedicated regular expression tester to see what exactly match and what will replace