Does anyone have any good examples of normalizing paths in files?
I have files that have paths like this in them:
../../foo.htm
Which I process to make fully-qualified like this:
/abc/def/ghi/jkl/mno/../../foo.htm
Now I want to normalize the fully-qualified paths to remove the "../" parts
to get a result like this:
/abc/def/ghi/foo.htm
I've found I can use <replaceregexp> like this to replace one set of "../"
pairs:
<replaceregexp
match="${util.includeTagStartRegExp}(.*?)/([^/.]*?)/\.\./(.*?)${util.includeTagEndRegExp}"
replace="${util.includeTagStartRegExp}\1/\3${util.includeTagEndRegExp}"
flags="sg" byline="false">
And if I enclose this in a for loop, I can call it multiple times to remove
multiple sets of "../" pairs. However, I don't like hardcoding the number
of loops. Because it would be complicated to actually scan the files to get
the maximum number of "../" path components, the simplest thing for me to
use as my loop count is the maximum depth of the directory tree. Is there a
way to compute this?
I've tried looping over the directory tree, getting the directories,
stripping out everything that isn't a "/". However, I can't figure out how
to determine which of these strings is the longest inside a for loop so I
know what my loop count will be.
--
View this message in context:
http://www.nabble.com/Normalizing-paths-in-files-tp22695672p22695672.html
Sent from the Ant - Users mailing list archive at Nabble.com.
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
