syntax for varaiable with spaces

[Michael Allmer]

I have found many threads that talk about spaces in file name and directories. I though that I was following the information in them but I am sptill having issues, and I must have sothing wrong in my syntax.
Script is as follows;
#!/usr/bin/sh
#
COPY_CMD=`rsync -avz`
#
LIST=`cat /home/allmerms/office_move/raid_mdb.txt`
#
DEST="/office"
#
#
for l in $LIST
do
$COPY_CMD "${l}" $DEST
done

[James R. Ferguson · Accepted solution]

Hi Mike:

# COPY_CMD=`rsync -avz`

...should be:

# COPY_CMD="rsync -avz"

Regards!

...JRF...

[Michael Allmer]

Here is the example of the error that I am seeing;
+ /raid/documents/actuation/eng_pdu/Tony /office
./test_copy.sh[13]: /raid/documents/actuation/eng_pdu/Tony: not found.
+ Li/saved/MoveNewDesktop/work/PNcrossReference.mdb /office
./test_copy.sh[13]: Li/saved/MoveNewDesktop/work/PNcrossReference.mdb: not found.



The full path is;
"/raid/documents/actuation/eng_pdu/Tony Li/saved/work/PNcrossReference.mdb"

[Steven Schweda]

> [...] but I am sptill having issues [...]

Are those anything like _problems_? And do
we need to guess what they were, or might you
be persuaded to tell us?

> LIST=`cat /home/allmerms/office_move/raid_mdb.txt`

You have a file named "cat"?

[James R. Ferguson]

Hi (again) Mike:

By the way, the use of backticks is deprecated in favor of '$()' as:

# LIST=`cat /home/allmerms/office_move/raid_mdb.txt`

...is better written:

# LIST=$(cat /home/allmerms/office_move/raid_mdb.txt)

...which, is more _efficiently_ done as:

# LIST=$(< /home/allmerms/office_move/raid_mdb.txt)

...which causes the shell to perform the read of the file without forking a separate process (here, 'cat').

Regards!

...JRF...

[Michael Allmer]

Thanks James, you help was what I needed.

[Michael Allmer]

After changing script to follwo James' suggestions:
#!/usr/bin/sh
#
COPY_CMD="rsync -avz"
#
LIST=$(#
DEST="/office"
#
#
for l in $LIST
do
$COPY_CMD "${l}" $DEST
done

I am still getting the following error form the output:

sent 29 bytes received 20 bytes 98.00 bytes/sec
total size is 0 speedup is 0.00
rsync error: some files could not be transferred (code 23) at main.c(892) [sender=2.6.8]
+ rsync -avz /raid/documents/actuation/eng_pdu/Tony /office
building file list ... rsync: link_stat "/raid/documents/actuation/eng_pdu/Tony" failed: No such file or directory (2)
done

sent 29 bytes received 20 bytes 98.00 bytes/sec
total size is 0 speedup is 0.00
rsync error: some files could not be transferred (code 23) at main.c(892) [sender=2.6.8]
+ rsync -avz Li/saved/move/Program /office
building file list ... rsync: link_stat "/home/allmerms/office_move/Li/saved/move/Program" failed: No such file or directory (2)
done


In additions some of the file that where succesful are not what I expected. I wanted the full path created after /office
Exapmle:
"/raid/unigraphics/utils/var/tables/Teamcenter-DB.mdb"
Was created in /office as; /office/Teamcenter-DB.mdb

I was expecting;
"/office/raid/unigraphics/utils/var/tables/Teamcenter-DB.mdb"

I hope that I have explained my expectation correctly.

[James R. Ferguson]

Hi (again) Mike:

A failed stat() means that the file or directory isn't found (or readable). See if this helps you:

...
for l in ${LIST}
do
[ -f "${l}" ] || { echo "missing file: ${l}"; continue; }
DIR=$(dirname ${l})
echo ${COPY_CMD} "${l}" "${DEST}${DIR}"
done

Regards!

...JRF...

[Michael Allmer]

James,

Here is a log of the output from you latest suggested change.

[James R. Ferguson]

Hi Mike:

> Here is a log of the output from you latest suggested change.

Without seeing what the input file for that run looks like, it is difficult to comment.

Regards!

...JRF...

[Michael Allmer]

Sorry, here it is

[Michael Allmer]

here is the "LIST" file as well.

[Michael Allmer]

I was running against an old list. Sorry for the confusion. here is the updated list as well as the attached output.

/raid/cad/777-lowerlobe/200-lr/docs/All_Side_Guide_BOMs.mdb
/raid/unigraphics/utils/var/tables/Teamcenter-DB.mdb
/raid/documents/actuation/eng_pdu/Tony_Li/saved/MoveNewDesktop/work/PNcrossReference.mdb
/raid/documents/actuation/eng_pdu/Tony_Li/saved/move/Program Files/Visual Sizer/VisualSizer.mdb
/raid/documents/actuation/eng_pdu/Tony_Li/saved/move/Program Files/Visual Sizer/appdata.mdb
/raid/documents/actuation/eng_pdu/Tony_Li/saved/move/Program Files/Visual Sizer/datatemp.mdb
/raid/documents/actuation/eng_pdu/Tony_Li/saved/move/Program Files/speed/spud/tuts/mag10.mdb
/raid/documents/actuation/eng_pdu/Tony_Li/saved/work/PNcrossReference.mdb
/raid/documents/actuation/eng_pdu/Tony_Li/saved/workold/PNcrossReferenceCustomerPN-1.mdb
/raid/documents/actuation/eng_pdu/Tony_Li/saved/workold/PNcrossReferenceCustomerPN.mdb

[James R. Ferguson]

Hi (again) Mike:

Oh, you have _spaces_ in your filenames. I should have guessed that from your original post. Seeing actual input avoids ambiguous communication.

#!/usr/bin/sh
COPY_CMD="rsync -avz"
LIST="/home/allmerms/office_move/raid_mdb.txt"
DEST="/office"
while read FILE
do
[ -f "${FILE}" ] || { echo "missing file: '${FILE}'"; continue; }
DIR=$(dirname "${FILE}")
echo ${COPY_CMD} "'${FILE}'" "'${DEST}${DIR}'"
done < ${LIST}
exit 0

This also avoids forking a process to construct a list in memory. Instead we let the shell read each line of the file for us.

I would suggest that you read the 'rsync' manpages to make sure that the 'rsync' syntax gives you what you want.

Regards!

...JRF...

[Michael Allmer]

The last par of the directory path is being removed.

using the last iput file I provided, I ran the following.

#!/usr/bin/sh
#
LIST="/home/allmerms/office_move/raid_mdb.txt"
#
#
while read FILE
do
[ -f "${FILE}" ] || { echo "missing file: ${FILE}"; continue; }
DIR=$(dirname ${FILE})
echo "${DIR}"
done < ${LIST}
exit 0

The output file show that it is dropping everything in the path name after the first space.

[James R. Ferguson]

Hi (again) Mike:

> The output file show that it is dropping everything in the path name after the first space.

Yes, you dropped quotes. Compare:

...
while read FILE
do
#-> [ -f "${FILE}" ] || { echo "missing file: ${FILE}"; continue; }
DIR=$(dirname ${FILE})
echo "<<< ${DIR}"
DIR=$(dirname "${FILE}")
echo ">>> ${DIR}"
done < ${LIST}

Regards!

...JRF...

[Michael Allmer]

Thanks for all the assistance.

[Fredrik.eriksson]

This is closed and all... but why not just use the IFS?

#!/bin/nash
COPY_CMD="rsync -avz"
LIST=$(cat /home/allmerms/office_move/raid_mdb.txt)
DEST="/office"

IFS="
"
for i in $LIST; do
$COPY_CMD "${i}" $DEST
done

Best regards
Fredrik Eriksson