Hi everyone,
I'm trying to create a script that will compare the contents of two directories and delete all of the
duplicated files in the first.
The problem comes with long filenames in OS X. The OS mangles filenames for FSSpec
representation (which I assume MPW is using) and changes names such as "This is a really long
filename that will cause OS name mangling.png" into "This is a really lon#33C91E.png"
My attempts to deal with this problem involves the following attempt (apologies if high-ASCII
characters don't make it through the list server):
echo "### DeleteEqualFiles {1} {2}"
for f in `files -o -s "{{1}}"`
set ismatch 0
if "{f}" =~ /(?«18,26»[¬#])®2?#[0-9A-F]«2,10»(.?«1,6»)*®3/
# special-case processing for long file names
for g in `files -o "{{2}}{{®2}}"?"{®3}`
# compare data fork only
equal -d -q "{{1}}{{f}}" "{{g}}"
if {status} == 0
# compare resource forks. if neither has an RF, succeeds
equal -r -q "{{1}}{{f}}" "{{g}}"
if {status} != 0
resequal -fFlags -skip 'ckid' -skip 'vers' "{{1}}{{f}}" "{{g}}"
end
if {status} == 0
set ismatch 1
break
end
end
end
else
# compare data fork only
equal -d -q "{{1}}{{f}}" "{{2}}{{f}}"
if {status} == 0
# compare resource forks. if neither has an RF, succeeds
equal -r -q "{{1}}{{f}}" "{{2}}{{f}}"
if {status} != 0
resequal -fFlags -skip 'ckid' -skip 'vers' "{{1}}{{f}}" "{{2}}{{f}}"
end
if {status} == 0
set ismatch 1
end
end
end
if {ismatch}
echo "{{f}}"
# delete "{{1}}{{f}}"
end
end
As you can see in line 4, I'm trying to detect a mangled name and do a recursive check of matches in
the target directory. If no mangling is detected, I fall through to equivalent code to do duplication
checks on the complete filename.
My test folder consists of the following files, a folder from my daughter's iBook:
files 'Macintosh HD:Users:petersb:Desktop:Stuff1:'
'"If you would make p#33C809.rtf' # problem file 1
'"Son," the man said,#33C80A.rtf'
'*How will understand#33C80B.rtf'
.DS_Store
'Ah, well, after the #33C80D.rtf'
'And wiping the dust #33C811.rtf'
'BLUE interrupted, "Y#33C813.rtf' # problem file 2
'God'?''s word is clear #33C816.rtf'
'I shout it to you in#33C817.rtf'
Journals....rtf
'Letter To Toni (Day 2) .rtf'
'Little Lines.rtf'
'Little Secret Sister Notes.rtf'
'My dreams for my Writing.rtf'
'Quote Book.rtf'
'The Piano'
'Value Journals.rtf'
Apparently she likes to be liberal in her use of spaces and single/double quote characters in her
filenames, and tends to make them very descriptive.
Unfortunately I am having problems in my mangled-name detector. For example, the filenames
noted above as "problem files" both generate an error in the Shell. (### MPW Shell - "s must occur in
pairs).
Also, the pattern doesn't seem to catch the mangled names that have extensions. I have been able to
catch non-extension mangles with this pattern
if "{f}" =~ /(?«18,26»[¬#])®2?#[0-9A-F]«2,10»/
but catching all cases and getting the proper quoting is proving to be quite elusive.
Does anyone have a good mangled-name filename detector that you would be willing to share?
(And before anyone says to "just use diff in the Terminal", AFAIK diff doesn't do any checks with
resource forks, even in 10.4.x . :-)
Thanks!
Bill
_______________________________________________
Do not post admin requests to the list. They will be ignored.
MPW-Dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden
