Script to replace backslashes with forward slashes

[apoth]

I need to recurse through some folders and replace any backslashes with forward slashes (Windows path names to Unix) in file contents.

It should just be a one liner but I'm useless with sed/awk and regexps in general. Can anyone offer a solution? Thanks.

[ghostdog74]

show a listing of how your directory and files look like

[apoth]

$ find . -type f -name "*.html"
./spacer.html
./subscriptionprofile/subscriptionprofile_associations.html
./subscriptionprofile/subscriptionprofile_01.html
./subscriptionprofile/subscriptionprofile_costassociation_cod.html
./subscriptionprofile/subscriptionprofile_costassociation_btv.html
./subscriptionprofile/subscriptionprofile_costassociation.html
./broadcast/channel_01.html
./broadcast/broadcast_associations.html
./broadcast/package_01.html
./broadcast/event_01.html
./broadcast/program_01.html
./broadcast/category_01.html
./Banner.html
./ondemand/catalogue_01.html
./ondemand/schedule_01.html
./ondemand/catalogue_schedule_scenario_01.html
./ondemand/asset_01.html
./ondemand/assetgroup_01.html
./ondemand/ondemand_associations.html
./[TUT-003] - QSP Overview Presentation.html
./rating/rating_01.html
./rating/rating_associations.html
./acl/acquiredcontentlist_01.html
./acl/acquiredcontentlist_02.html
./acl/acquiredcontentlist_03.html
./acl/acl_associations.html
./devices/feature_01.html
./devices/deviceprofile_01.html
./devices/applicationserver_01.html
./devices/devicegroup_01.html
./devices/deviceclassifier_01.html
./devices/settopbox_01.html
./devices/devices_associations.html
./devices/streamingserver_01.html
./general/acquireable_01.html
./general/chargeability_01.html
./general/subscribeable_01.html
./general/chargeability_associations.html
./general/audiovariants.html
./general/overview.html
./general/content_children_01.html
./general/languagevariants.html
./general/commonattributes.html
./general/content_parents_01.html
./Framework.html
./policygroup/policygroup_01.html
./policygroup/policyoftypebaseprice_01.html
./policygroup/policygroup_associations.html
./policygroup/policyoftypeother_01.html
./policygroup/policygroup_scenario_01.html
./policygroup/interval_01.html
./LinksTable.html
./blank.html
./accesspoint/accesspoint_associations.html
./accesspoint/accesspoint_01.html
./account/account_01.html
./account/account_associations.html
./account/user_01.html

[ghostdog74]

it pays to describe what you want to do properly, right.


# find . -name "*.jpg" | awk '{gsub("/","\\")}1'

[apoth]

it pays to describe what you want to do properly, right.

Not sure what you mean.





# find . -name "*.jpg" | awk '{gsub("/","\\")}1'


I couldn't get that working when I tested it, although this works:


echo "test\test" | awk '{ gsub(/\\/, "/"); print }'

I just need to get it to read in the contents of each of those files, run this against the contents and then spit it out to the same file now.

[ghostdog74]

how did you get the windows slash ?? are you running the command on windows?? why don't you show what's not working?

[apoth]

These are existing documents created on a Windows machine using Windows file paths for the links. Not that I'm sure that how the slashes got there is of any relevance.

No, I'm running the command on linux.

What do you mean by what's not working? I'm asking for a script to be created for this purpose, I don't have an existing script for the job that doesn't work?

[ghostdog74]

These are existing documents created on a Windows machine using Windows file paths for the links. Not that I'm sure that how the slashes got there is of any relevance.

No, I'm running the command on linux.

What do you mean by what's not working? I'm asking for a script to be created for this purpose, I don't have an existing script for the job that doesn't work?


first you said you need to recurse a directory and subdirectories in linux and replace "\" with "/" , correct ?
In *nix, the paths are all "/", as shown by your own "find" results a few posts back. so i ask you again, where is the windows file path with "\" you got from?describe your problem and what you want to do properly.

[WW]

ghostdog74: In the first line of the original post, apoth says he wants to change backslashes to slashes in the contents of the files.

apoth: This will change every occurrence of \ to / in all the .html files in the current directory and all subdirectories. It will also make backups of the originals with a tilde (~) added to the original file name.


$ find . -name "*.html" -type f -exec sed 's/\\/\//g' '{}' -i~ \;

It is similar to the find command that you used above, but it uses the -exec option to execute a command on each file found. The command is sed. The first argument of the sed command is 's/\\/\//g', which will substitute all occurrences of \ with /. The -i~ option tells sed to make the substitution "in place", and make a backup with the extension ~.

[murphlaw]

"an awk way"



find -name "*.html" -type f | while read file
do
awk '{ gsub( /\\/, "/" ); print }' $file > $file.$$
mv $file.$$ $file
done

[apoth]

Both great solutions, thanks guys.

[subverted]

"an awk way"



find -name "*.html" -type f | while read file
do
awk '{ gsub( /\\/, "/" ); print }' $file > $file.$$
mv $file.$$ $file
done


How about something completely different?

I'm using Vim to edit websites and like to test them in various browsers with:

:!firefox %

But using ies4linux it doesn't work because there are backslashes that vim can't interpret, so I need a script to reverse the forward slashes to backslashes as the file is called. Something like:


#!/bin/sh

for i in "$@"
do
a="$a 'echo "$i"|sed -e 's#/#\\#g<CR>'"
done

eval ie6 $a

Ha! Is that even remotely possible? or is there some regular expression for Vim that I am unaware of that would do the trick?

[silentrebel]

When I use this method:


$> hello="dkfafd\dklfafd\kdfasdfa"
$> goodbye=`echo $hello | awk '{ gsub( /\\/, "/" ); print }'`

I get:


awk: { gsub( /\/, "/" ); print }
awk: ^ unterminated string


I cannot figure out why!


------------------------------------------------------------
I was also wondering if this expression would equally work...:

awk '{ gsub( "\\", "/" ); print }'
Thanks

[ghostdog74]

When I use this method:


$> hello="dkfafd\dklfafd\kdfasdfa"
$> goodbye=`echo $hello | awk '{ gsub( /\\/, "/" ); print }'`

I get:


awk: { gsub( /\/, "/" ); print }
awk: ^ unterminated string


I cannot figure out why!


------------------------------------------------------------
I was also wondering if this expression would equally work...:

awk '{ gsub( "\\", "/" ); print }'
Thanks

change your backticks to $() syntax.

g=$(echo $hello | awk '{ gsub( /\\/, "/" ); print }')

[diesch]

find -type f -exec perl -i -pe's:\\:/:g' {} \;

[silentrebel]

This advice worked perfectly:

change your backticks to $() syntax.

g=$(echo $hello | awk '{ gsub( /\\/, "/" ); print }')


For the record,

goodbye=`echo $hello | awk '{ gsub("\\", "/"); print }'`
does not work, as I thought it might...

Could anyone tell me why the backticks do not work here?:

goodbye=`echo $hello | awk '{ gsub(/\\/, "/"); print }'`