How to renumber files? - comp.os.linux.misc

Hi: Photos from my digital camera are named like this: crw_0001.crw crw_0002.crw crw_0003.crw crw_0004.crw etc. Let's say I have two CF cards, both with a name sequence that begins at _0001. I want to merge the contents of the two cards into one directory without the second card's files overwriting the first. How can I rename the files from the second card so that the 4-digit numeric field is renumbered starting at a value of my choosing (which would likely be 1+ the last number from the first card)? For instance, I want to rename the above files to: crw_0005.crw crw_0006.crw crw_0007.crw crw_0008.crw etc. Any ideas? Thanks. -- _____________________ Christopher R. Carlen crobc@sbcglobal.net SuSE 9.1 Linux 2.6.5

On Fri, 21 Jan 2005 11:21:58 -0800, Chris Carlen staggered into the Black Sun and said: > Photos from my digital camera are named like this: > crw_0001.crw crw_0002.crw crw_0003.crw crw_0004.crw > > Let's say I have two CF cards, both with a name sequence that begins > at _0001. I want to merge the contents of the two cards into one > directory without the second card's files overwriting the first. How > can I rename the files from the second card so that the 4-digit > numeric field is renumbered starting at a value of my choosing? > > For instance, I want to rename the above files to: > crw_0005.crw crw_0006.crw crw_0007.crw crw_0008.crw > Any ideas? Try something like this: #!/bin/bash # place in ~/bin/renamecrw.sh ; chmod +x it # call with "renamecrw.sh 5" to add 5 to the index number of every .crw # file in the current directory for FILE in *.crw ; do NUM=`echo $FILE | sed -e 's/crw_//' -e 's/.crw//' ` let NUM=$NUM+$1 NEWFILE=`printf "crw_%04d.crw" $NUM ` mv $FILE $NEWFILE done # end of script ....there are probably better ways to do this; this is just what I thought of in ~30 seconds. I think "rename" won't work here since it's not complex enough, though there's probably a more intelligent file-renaming thingy out there--I've just forgotten its name. HTH, -- Matt G|There is no Darkness in Eternity/But only Light too dim for us to see Brainbench MVP for Linux Admin / mail: TRAP + SPAN don't belong http://www.brainbench.com / Hire me! -----------------------------/ http://crow202.dyndns.org/~mhgraham/resume

On Fri, 21 Jan 2005 at 19:21 GMT, Chris Carlen wrote: > Hi: > > Photos from my digital camera are named like this: > > crw_0001.crw > crw_0002.crw > crw_0003.crw > crw_0004.crw > etc. > > Let's say I have two CF cards, both with a name sequence that begins at > _0001. I want to merge the contents of the two cards into one directory > without the second card's files overwriting the first. > > How can I rename the files from the second card so that the 4-digit > numeric field is renumbered starting at a value of my choosing (which > would likely be 1+ the last number from the first card)? > > For instance, I want to rename the above files to: > > crw_0005.crw > crw_0006.crw > crw_0007.crw > crw_0008.crw > etc. DEST=/path/to/new/directory X=4 ## Amount to increment each number for file in *.crw do num=${file%.crw} ## remove .crw from end num=${num#crw_} ## remove crw_ from beginning case $num in ## Remove leading zeros and add increment 0000) newnum=$X ;; 000?) newnum=$(( ${num#000} + $X )) ;; 00??) newnum=$(( ${num#00} + $X )) ;; 0???) newnum=$(( ${num#0} + $X )) ;; *) newnum=$(( $num + $X )) ;; esac mv "$file" "$DEST/crw_$newnum.crw" done -- Chris F.A. Johnson http://cfaj.freeshell.org/shell =================================================================== My code (if any) in this post is copyright 2004, Chris F.A. Johnson and may be copied under the terms of the GNU General Public License

On Fri, 21 Jan 2005 20:22:00 +0000, Dances With Crows wrote: > On Fri, 21 Jan 2005 11:21:58 -0800, Chris Carlen staggered into the > Black Sun and said: >> Photos from my digital camera are named like this: >> crw_0001.crw crw_0002.crw crw_0003.crw crw_0004.crw >> >> Let's say I have two CF cards, both with a name sequence that begins >> at _0001. I want to merge the contents of the two cards into one >> directory without the second card's files overwriting the first. How >> can I rename the files from the second card so that the 4-digit >> numeric field is renumbered starting at a value of my choosing? >> >> For instance, I want to rename the above files to: >> crw_0005.crw crw_0006.crw crw_0007.crw crw_0008.crw >> Any ideas? > > Try something like this: > > #!/bin/bash > # place in ~/bin/renamecrw.sh ; chmod +x it > # call with "renamecrw.sh 5" to add 5 to the index number of every .crw > # file in the current directory > for FILE in *.crw ; do > NUM=`echo $FILE | sed -e 's/crw_//' -e 's/.crw//' ` > let NUM=$NUM+$1 > NEWFILE=`printf "crw_%04d.crw" $NUM ` > mv $FILE $NEWFILE > done > # end of script > > ...there are probably better ways to do this; this is just what I > thought of in ~30 seconds. I think "rename" won't work here since it's > not complex enough, though there's probably a more intelligent > file-renaming thingy out there--I've just forgotten its name. HTH, I recall doing this kind of thing using "expr" (yeh, old timer, I know you use the shell more effectively than I do) to calculate the new sequence number. However, in situations like this, I personally would use rename, starting the next sequence at some convenient boundary, e.g. 0100? 1000?. In the past, I have found it useful to retain a hint of where the boundary was, by retaining a gap in the numbering. In retrospect, you may discover there are differences between the two sets, and you may want to retroactively apply some selective processing... but, as they say, YMMV. -- Juhan Leemet Logicognosis, Inc.

Dances With Crows wrote: > Try something like this: > > #!/bin/bash > # place in ~/bin/renamecrw.sh ; chmod +x it > # call with "renamecrw.sh 5" to add 5 to the index number of every .crw > # file in the current directory Ok, I like the idea of being able to pass the increment amount as a parameter. Lemme see if I understand this right: > for FILE in *.crw ; do The *.crw is expanded by bash to the list of files, and each is in turn assigned to var FILE upon which... > NUM=`echo $FILE | sed -e 's/crw_//' -e 's/.crw//' ` The var NUM is assigned a value equal to just the numeric digits portion of the file name by the magic sed command whose syntax I don't understand. I surmise it is doing something like: strip away the "crw_" part and strip away the ".crw" part, leaving just the number digits. > let NUM=$NUM+$1 Add the parameter from the command line. Oh god is the shell wierd. Being a moderately experienced C programmer for mostly embedded stuff, this being able to add two variables that a moment ago were used as if their types were strings, is very hard to digest. > NEWFILE=`printf "crw_%04d.crw" $NUM ` Make a new file name by printf-ing a string containing the before and after text fields again, and sticking the value of NUM in the middle as a 4-digit decimal value. Could have some fun with this, and make my wife crazy, by choosing a hex or binary format spec. ;-) > mv $FILE $NEWFILE > done > # end of script Ok, I get the point. Hmm, I can almost figure out how to modify this to do the other steps in my process of straightening out the filenames. I also usually do a: rename crw_ 012005- * to strip the unneeded pre-text, and replace it with the date. I suppose I can try to make the script do this automagically for me. If I can't figure it out, I'll be back. But I won't complain if the clues are provided anyway. Thanks for the input! Good day! -- _____________________ Christopher R. Carlen crobc@sbcglobal.net SuSE 9.1 Linux 2.6.5

Dances With Crows wrote: > NUM=`echo $FILE | sed -e 's/crw_//' -e 's/.crw//' ` Is there a way to not hardcode dependence on a specific initial text field, and .extension field, but instead just pick out the numeric part of any filename of the basic structure: beforeNNNNafter Where NNNN is the numeric digits part which may be any number of digits. "before" is alphabetic or punctuation before the numbers, that may also not even exist. "after" is also alphabetic or punctuation after the numbers which may also not exist. So it could work correctly on any of these: 1 003.pic file0001.ext data456.bin stuff-99 Also what would happen if we incremented the "99" in the last example to a three digit number? If we can do this, it's a fairly presentable "renum" command! Thanks for input. Good day. -- _____________________ Christopher R. Carlen crobc@sbcglobal.net SuSE 9.1 Linux 2.6.5

In comp.os.linux.misc Chris Carlen <crobc@bogusfield.sbcglobal.net>: > Hi: > Photos from my digital camera are named like this: > crw_0001.crw > crw_0002.crw > crw_0003.crw [..] > How can I rename the files from the second card so that the 4-digit > numeric field is renumbered starting at a value of my choosing (which > would likely be 1+ the last number from the first card)? > For instance, I want to rename the above files to: > crw_0005.crw > crw_0006.crw [..] ls crw_[0-9]* | awk 'BEGIN{FS="[0-9]+"}{printf "mv "$0" "; printf "%s%4.4d%s",$1,NR+4,$NF"\n"}' > file The above should be a single line, got wrapped. Check file if things are fine then run it: sh file Might give some problems if numbers are missing.;) Good luck -- Michael Heiming (X-PGP-Sig > GPG-Key ID: EDD27B94) mail: echo zvpunry@urvzvat.qr | perl -pe 'y/a-z/n-za-m/' #bofh excuse 441: Hash table has woodworm

Dances With Crows wrote: > Try something like this: > > #!/bin/bash > # place in ~/bin/renamecrw.sh ; chmod +x it > # call with "renamecrw.sh 5" to add 5 to the index number of every .crw > # file in the current directory > for FILE in *.crw ; do > NUM=`echo $FILE | sed -e 's/crw_//' -e 's/.crw//' ` > let NUM=$NUM+$1 > NEWFILE=`printf "crw_%04d.crw" $NUM ` > mv $FILE $NEWFILE > done > # end of script Problems: user1@ting:/photos/300d-photos/raw-jpeg/test> l crw_0012.crw renum user1@ting:/photos/300d-photos/raw-jpeg/test> ./renum 10 user1@ting:/photos/300d-photos/raw-jpeg/test> l crw_0020.crw renum user1@ting:/photos/300d-photos/raw-jpeg/test> ./renum 100 user1@ting:/photos/300d-photos/raw-jpeg/test> l crw_0116.crw renum user1@ting:/photos/300d-photos/raw-jpeg/test> ./renum 100 user1@ting:/photos/300d-photos/raw-jpeg/test> l crw_0178.crw renum user1@ting:/photos/300d-photos/raw-jpeg/test> ./renum 10 ../renum: line 8: let: NUM=0178: value too great for base (error token is "0178") ../renum: line 1: printf: 0178: invalid number user1@ting:/photos/300d-photos/raw-jpeg/test> l crw_0000.crw renum It would seem that the math on text strings is a little dicey. Any ideas how to fix? Thanks. -- _____________________ Christopher R. Carlen crobc@sbcglobal.net SuSE 9.1 Linux 2.6.5

On Fri, 21 Jan 2005 16:26:40 -0800, Chris Carlen staggered into the Black Sun and said: > Dances With Crows wrote: >> #!/bin/bash >> # place in ~/bin/renamecrw.sh ; chmod +x it >> # call with "renamecrw.sh 5" to add 5 to the index number of every .crw >> # file in the current directory > > Lemme see if I understand this right: >> for FILE in *.crw ; do >> NUM=`echo $FILE | sed -e 's/crw_//' -e 's/.crw//' ` > The var NUM is assigned a value equal to just the numeric digits > portion of the file name by the magic sed command whose syntax I don't > understand. sed -e $REGEX : apply regular expression REGEX to stardard input. You can use multiple -e arguments to apply multiple regexes to the input. There's no man page for regular expressions, but "man perlre" provides a nice reference for Perl-compatible regular expressions. sed doesn't really use PCRE AFAIK, but the basics hold. > I surmise it is doing something like: strip away the "crw_" part and > strip away the ".crw" part, leaving just the number digits. Yep, that's it. >> let NUM=$NUM+$1 > Add the parameter from the command line. Oh god is the shell wierd. > Being a moderately experienced C programmer for mostly embedded stuff, > this being able to add two variables that a moment ago were used as if > their types were strings, is very hard to digest. Type-safety is mostly a C and Java preoccupation. Calling atoi() would be... safer, but more work. I find bash's syntax kind of weird, but manageable for small things. For anything long, or anything complicated, Perl works better. (If you think bash's syntax is "weird" or "sloppy", you'll really get a kick out of Perl's TMTOWTDI approach.) >> NEWFILE=`printf "crw_%04d.crw" $NUM ` >> mv $FILE $NEWFILE >> done >> # end of script > Hmm, I can almost figure out how to modify this to do the other steps > in my process of straightening out the filenames. I also usually do > > rename crw_ 012005- * > > to strip the unneeded pre-text, and replace it with the date. I > suppose I can try to make the script do this automagically for me. If > I can't figure it out, I'll be back. But I won't complain if the > clues are provided anyway. OK, one thing that I didn't test in the ~2 minutes I spent writing the original message was longer numbers. bash interprets integers that start with 0 as octal (for obvious reasons) so that's behind the "weird math" you were seeing earlier. You can run NUM through a sed command that removes leading zeroes, and that'd make things work much better. Like so: NUM=`echo $NUM | sed -e 's/^0*//' ` ....put that before the "let" and things should work much better. HTH, -- Matt G|There is no Darkness in Eternity/But only Light too dim for us to see Brainbench MVP for Linux Admin / mail: TRAP + SPAN don't belong http://www.brainbench.com / Hire me! -----------------------------/ http://crow202.dyndns.org/~mhgraham/resume

On Sat, 22 Jan 2005 at 01:21 GMT, Chris Carlen wrote: > user1@ting:/photos/300d-photos/raw-jpeg/test> ./renum 10 > ./renum: line 8: let: NUM=0178: value too great for base (error token is > "0178") > ./renum: line 1: printf: 0178: invalid number > user1@ting:/photos/300d-photos/raw-jpeg/test> l > crw_0000.crw renum > > > It would seem that the math on text strings is a little dicey. > > Any ideas how to fix? The leading zeros must be removed, or they will be interpreted as octal. and 08 and 09 are not valid octal numbers. Here's my solution, again, which deals with the issue (note that there is no need for any external utility to break apart the string and do the arithmeitc. Also note that I do not use command substitution (`...`) as it is very slow in bash: DEST=/path/to/new/directory X=4 ## Amount to increment each number for file in *.crw do num=${file%.crw} ## remove .crw from end num=${num#crw_} ## remove crw_ from beginning case $num in ## Remove leading zeros and add increment 0000) newnum=$X ;; 000?) newnum=$(( ${num#000} + $X )) ;; 00??) newnum=$(( ${num#00} + $X )) ;; 0???) newnum=$(( ${num#0} + $X )) ;; *) newnum=$(( $num + $X )) ;; esac mv "$file" "$DEST/crw_$newnum.crw" done If you want to generalize it, put crw_ and .crw in variables, instead of hard coding them. -- Chris F.A. Johnson http://cfaj.freeshell.org/shell =================================================================== My code (if any) in this post is copyright 2004, Chris F.A. Johnson and may be copied under the terms of the GNU General Public License

On Sat, 22 Jan 2005 at 02:46 GMT, Chris F.A. Johnson wrote: > On Sat, 22 Jan 2005 at 01:21 GMT, Chris Carlen wrote: >> user1@ting:/photos/300d-photos/raw-jpeg/test> ./renum 10 >> ./renum: line 8: let: NUM=0178: value too great for base (error token is >> "0178") >> ./renum: line 1: printf: 0178: invalid number >> user1@ting:/photos/300d-photos/raw-jpeg/test> l >> crw_0000.crw renum >> >> >> It would seem that the math on text strings is a little dicey. >> >> Any ideas how to fix? > > The leading zeros must be removed, or they will be interpreted as > octal. and 08 and 09 are not valid octal numbers. > > Here's my solution, again, which deals with the issue (note that > there is no need for any external utility to break apart the string > and do the arithmeitc. Also note that I do not use command > substitution (`...`) as it is very slow in bash: > > DEST=/path/to/new/directory > X=4 ## Amount to increment each number > for file in *.crw > do > num=${file%.crw} ## remove .crw from end > num=${num#crw_} ## remove crw_ from beginning > case $num in ## Remove leading zeros and add increment > 0000) newnum=$X ;; > 000?) newnum=$(( ${num#000} + $X )) ;; > 00??) newnum=$(( ${num#00} + $X )) ;; > 0???) newnum=$(( ${num#0} + $X )) ;; > *) newnum=$(( $num + $X )) ;; > esac > mv "$file" "$DEST/crw_$newnum.crw" > done > > If you want to generalize it, put crw_ and .crw in variables, > instead of hard coding them. It doesn't deal with padding, however. That can be done with another case statement: case $newnum in ?) newnum=000$newnum ;; ??) newnum=00$newnum ;; ???) newnum=0$newnum ;; esac Or with parameter expansion: newnum=0000$newnum newnum=${newnum#${newnum%????}} -- Chris F.A. Johnson http://cfaj.freeshell.org/shell =================================================================== My code (if any) in this post is copyright 2004, Chris F.A. Johnson and may be copied under the terms of the GNU General Public License

� Chris Carlen <crobc@BOGUSFIELD.sbcglobal.net>: > Hi: > > Photos from my digital camera are named like this: > > crw_0001.crw > crw_0002.crw > crw_0003.crw > crw_0004.crw > etc. > > Let's say I have two CF cards, both with a name sequence that begins at > _0001. I want to merge the contents of the two cards into one directory > without the second card's files overwriting the first. > > How can I rename the files from the second card so that the 4-digit > numeric field is renumbered starting at a value of my choosing (which > would likely be 1+ the last number from the first card)? > > For instance, I want to rename the above files to: > > crw_0005.crw > crw_0006.crw > crw_0007.crw > crw_0008.crw > etc. > > > Any ideas? Yep. #! /bin/bash current=1 start=4 for image in crw_*.crw; do number=$(printf %04d $(($start + $current)) ) mv "$image" "crw_${number}.crw" current=$(($current + 1)) done Alexander Skwar -- Be careful what you set your heart on -- for it will surely be yours. -- James Baldwin, "Nobody Knows My Name" ________________________________________________________________________

In article <35dt3uF4ki4o5U1@individual.net>, Chris F.A. Johnson <cfajohnson@gmail.com> wrote: : : The leading zeros must be removed, or they will be interpreted as : octal. and 08 and 09 are not valid octal numbers. : : Here's my solution, again, which deals with the issue (note that : there is no need for any external utility to break apart the string : and do the arithmeitc. Also note that I do not use command : substitution (`...`) as it is very slow in bash: : :DEST=/path/to/new/directory :X=4 ## Amount to increment each number :for file in *.crw :do : num=${file%.crw} ## remove .crw from end : num=${num#crw_} ## remove crw_ from beginning : case $num in ## Remove leading zeros and add increment : 0000) newnum=$X ;; : 000?) newnum=$(( ${num#000} + $X )) ;; : 00??) newnum=$(( ${num#00} + $X )) ;; : 0???) newnum=$(( ${num#0} + $X )) ;; : *) newnum=$(( $num + $X )) ;; : esac : mv "$file" "$DEST/crw_$newnum.crw" :done Why not simply tell bash that it's a decimal number: newnum=$(( 10#$num + X )) But, since the OP indicated a wish to preserve the leading zeros: DEST=/path/to/new/directory X=4 ## Amount to increment each number ZPAD=0000000000 for file in *.crw do num=${file%.crw} ## remove .crw from end num=${num#crw_} ## remove crw_ from beginning newnum=$(( 10#$num + $X )) pad=$(( ${#num} - ${#newnum} )) ## the length difference [ $pad -gt 0 ] && newnum=${ZPAD:0:$pad}$newnum mv "$file" "$DEST/crw_$newnum.crw" done -- Bob Nichols AT comcast.net I am "rnichols42"

On Sat, 22 Jan 2005 at 16:06 GMT, Robert Nichols wrote: > In article <35dt3uF4ki4o5U1@individual.net>, > Chris F.A. Johnson <cfajohnson@gmail.com> wrote: >: >: The leading zeros must be removed, or they will be interpreted as >: octal. and 08 and 09 are not valid octal numbers. >: >: Here's my solution, again, which deals with the issue (note that >: there is no need for any external utility to break apart the string >: and do the arithmeitc. Also note that I do not use command >: substitution (`...`) as it is very slow in bash: >: >:DEST=/path/to/new/directory >:X=4 ## Amount to increment each number >:for file in *.crw >:do >: num=${file%.crw} ## remove .crw from end >: num=${num#crw_} ## remove crw_ from beginning >: case $num in ## Remove leading zeros and add increment >: 0000) newnum=$X ;; >: 000?) newnum=$(( ${num#000} + $X )) ;; >: 00??) newnum=$(( ${num#00} + $X )) ;; >: 0???) newnum=$(( ${num#0} + $X )) ;; >: *) newnum=$(( $num + $X )) ;; >: esac >: mv "$file" "$DEST/crw_$newnum.crw" >:done > > Why not simply tell bash that it's a decimal number: > > newnum=$(( 10#$num + X )) > > But, since the OP indicated a wish to preserve the leading zeros: > > DEST=/path/to/new/directory > X=4 ## Amount to increment each number > ZPAD=0000000000 > for file in *.crw > do > num=${file%.crw} ## remove .crw from end > num=${num#crw_} ## remove crw_ from beginning > newnum=$(( 10#$num + $X )) > pad=$(( ${#num} - ${#newnum} )) ## the length difference > [ $pad -gt 0 ] && newnum=${ZPAD:0:$pad}$newnum > mv "$file" "$DEST/crw_$newnum.crw" > done This is fine in bash, but I prefer to use a more portable syntax whenever possible. The two lines that are not POSIX compatible are: newnum=$(( 10#$num + $X )) [ $pad -gt 0 ] && newnum=${ZPAD:0:$pad}$newnum -- Chris F.A. Johnson http://cfaj.freeshell.org/shell =================================================================== My code (if any) in this post is copyright 2004, Chris F.A. Johnson and may be copied under the terms of the GNU General Public License

In article <35fri3F4jhg4qU1@individual.net>, Chris F.A. Johnson <cfajohnson@gmail.com> wrote: :On Sat, 22 Jan 2005 at 16:06 GMT, Robert Nichols wrote: :> :> Why not simply tell bash that it's a decimal number: :> :> newnum=$(( 10#$num + X )) :> :> But, since the OP indicated a wish to preserve the leading zeros: :> :> DEST=/path/to/new/directory :> X=4 ## Amount to increment each number :> ZPAD=0000000000 :> for file in *.crw :> do :> num=${file%.crw} ## remove .crw from end :> num=${num#crw_} ## remove crw_ from beginning :> newnum=$(( 10#$num + $X )) :> pad=$(( ${#num} - ${#newnum} )) ## the length difference :> [ $pad -gt 0 ] && newnum=${ZPAD:0:$pad}$newnum :> mv "$file" "$DEST/crw_$newnum.crw" :> done : : This is fine in bash, but I prefer to use a more portable syntax : whenever possible. The two lines that are not POSIX compatible are: : :newnum=$(( 10#$num + $X )) :[ $pad -gt 0 ] && newnum=${ZPAD:0:$pad}$newnum A reasonable goal. I pondered whether I should restrict myself to less bash-specific syntax, but then noted that this is a Linux newsgroup and the proposed script that started this subthread began with "#!/bin/bash". Actually, I believe that syntax also requires BASH_VERSION >= 2.00. -- Bob Nichols AT comcast.net I am "rnichols42"