Script FMfindRepl version 0.8 - questions concerning regular expression handling
Dear friends,
In my project FMfindRepl i want to build RegEx by selection items from drop-down lists. Experiments show that there are more capabilities in the FM Find process than documented in Help:
- Character classes \D (non-digit), \w (word character) and \W (non-word-character)
- Anchor \< for beginning of word, but not \> for end of word.
- 'lazy' multipliers *?, +? and ?? to get the shortest possible match
- DOTALL mode (?s)
- Case mode (sensitive, insensitive) by ?-i: and ?i-:
- Named capture groups and hence also named back references by ?<name> and \k<name>
About the replacement function we only know about the 'back reference' $0 … $9, but not of any (useful modifier such as upper-casing the match.
→ Where the heck can we get a complete description of the implemented RegEx engine?
Of course I know the phrase from the Help:
«FrameMaker uses the boost engine for Regex. For more information on Boost, see http://www.boost.org/.
You can write the regular expressions using the following syntax: Perl (default): See Perl doc – Regular expressions»
But this is only of limited value:
- The boost library can be implemented in various ways. What is the result for the end user in case of FM?
- There are many 'levels' of PERL syntax defined. Which one is implemented in FM?
- PERLprovides Regex features, which obviously are not implemented in FM. Some which one fare and which one aren't?
The UI for this secondary panel will look something like this:


