| A case for lookarounds in StRegEx() [message #88857] |
Thu, 26 June 2014 17:53 |
Matthew Argall
Messages: 286
Registered: October 2011
|
Senior Member |
|
|
I want to make a case for the stregex function to recognize lookarounds.
Say I have a list of tokens YMd. The tokens are identifiable because they are preceded by %. The "%" character can be escaped by "\". Try to extract the tokens following %.
The following case is successful. There are three tokens I want to find, so I search for "%" followed by any one of the three characters "[YMd]" and extract it with "()", then eat up any extra characters that are not % with "[^%]*".
IDL> print, stregex('file_%Y%M%d.txt', strjoin(replicate('%([YMd])[^%]*', 3)), /SUBEXP, /EXTRACT)
%Y%M%d.txt Y M d
Now I want to change the "%Y" character to "\%Y" so that the % is escaped and Y is excluded from the search. The following successfully skips "\%Y" and finds "%M", but fails to find "%d" because the "%" character that precedes "d" has been eaten up by a search for "[^\]" -- i.e. "[^\]" is of length one, whereas a negative lookbehind is of length zero.
IDL> print, stregex('file_\%Y%M%d.txt', strjoin(replicate('%([YMd])[^%]*', 3)), /SUBEXP, /EXTRACT)
IDL> print, stregex('file_\%Y%M%d.txt', strjoin(replicate('[^\]%([YMd])[^%]*', 1)), /SUBEXP, /EXTRACT)
Y%M M
Using the Python negative lookbehind notation "(?<!\\)%[YMd]" avoids %Y and matches %M and %d successfully (test here: https://www.debuggex.com/)
This is just one example of where they are useful.
---------------
TLDR; negative lookbehinds make searching for escaped characters really easy.
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
