| View previous topic :: View next topic |
| Author |
Message |
wswartzendruber
Veteran
Joined: 23 Mar 2004
Posts: 1261
Location: Idaho, USA
|
 Posted: Sat Oct 03, 2009 10:36 pm Post subject: Trying to Uppercase Text up to a Point |
 |
|
I'm guessing Perl's the best tool, but I need to turn this...
| Quote: | artist=ZZ Top
title=Gimme All Your Lovin'
album=Greatest Hits
Date=1992
Tracknumber=01
genre=Rock |
...into this...
| Quote: | ARTIST=ZZ Top
TITLE=Gimme All Your Lovin'
ALBUM=Greatest Hits
DATE=1992
TRACKNUMBER=01
GENRE=Rock |
_________________
Git has obsoleted SVN.
10mm Auto has obsoleted 45 ACP. |
|
| Back to top |
|
 |
yabbadabbadont
Advocate
Joined: 14 Mar 2003
Posts: 4791
Location: 2 exits past crazy
|
| Posted: Sat Oct 03, 2009 10:39 pm Post subject: |
|
|
| A shell script using cut (with = as the delimiter) and tr (to force uppercase) shouldn't be too hard to knock out. |
|
| Back to top |
|
|
wswartzendruber
Veteran
Joined: 23 Mar 2004
Posts: 1261
Location: Idaho, USA
|
| Posted: Sat Oct 03, 2009 10:52 pm Post subject: |
|
|
| yabbadabbadont wrote: | | A shell script using cut (with = as the delimiter) and tr (to force uppercase) shouldn't be too hard to knock out. |
That works except that it splits the tag names and tag values into two separate outputs. I don't know how to append two lines into one.
EDIT: Wait a minute, I think join is what I'm looking for.
_________________
Git has obsoleted SVN.
10mm Auto has obsoleted 45 ACP. |
|
| Back to top |
|
|
yabbadabbadont
Advocate
Joined: 14 Mar 2003
Posts: 4791
Location: 2 exits past crazy
|
| Posted: Sat Oct 03, 2009 10:55 pm Post subject: |
|
|
I was bored, so while you were replying, I was coding... 
| Code: | /home/daffy/temp $ cat DataFile
artist=ZZ Top
title=Gimme All Your Lovin'
album=Greatest Hits
Date=1992
Tracknumber=01
genre=Rock
/home/daffy/temp $ . ./go.sh
ARTIST=ZZ Top
TITLE=Gimme All Your Lovin'
ALBUM=Greatest Hits
DATE=1992
TRACKNUMBER=01
GENRE=Rock
/home/daffy/temp $ cat go.sh
while read TheLine
do
TheTag=$(echo "$TheLine" | cut -d = -f1)
TheRest=$(echo "$TheLine" | cut -d = -f1 --complement)
TheTag=$(echo "$TheTag" | tr [:lower:] [:upper:])
echo "$TheTag=$TheRest"
done < DataFile
|
You just need to redirect that last echo statement so that it appends to a new file.
| Code: | | echo "$TheTag=$TheRest" >> NewDataFile |
|
|
| Back to top |
|
|
wswartzendruber
Veteran
Joined: 23 Mar 2004
Posts: 1261
Location: Idaho, USA
|
| Posted: Sat Oct 03, 2009 11:37 pm Post subject: |
|
|
I came up with my own solution. Probably not as elegant:
| Code: | #!/bin/sh
#
# This little guy's going to fix metadata within FLAC files.
#
# $1 - Name of the FLAC file we're working on.
# First, dump the file's metadata somewhere.
metaflac --export-tags-to="$1.tags" "$1"
# Next, separate the tag names from their values, uppercasing the names.
cat "$1.tags" | cut --fields=1 --delimiter== | tr '[:lower:]' '[:upper:]' > "$1.tags.names"
cat "$1.tags" | cut --fields=2 --delimiter== > "$1.tags.values"
# Paste the two together.
paste --delimiters== "$1.tags.names" "$1.tags.values" > "$1.tags.almost"
# Sort the tags file.
cat "$1.tags.almost" | sort > "$1.tags"
# Remerge the sorted tags back into the file.
metaflac --remove-all-tags "$1"
metaflac --import-tags-from="$1.tags" "$1"
flac --best --force "$1"
# Clean up the garbage.
rm -rf "$1.tags" "$1.tags.names" "$1.tags.values" "$1.tags.almost" |
_________________
Git has obsoleted SVN.
10mm Auto has obsoleted 45 ACP. |
|
| Back to top |
|
|
John R. Graham
Administrator
Joined: 08 Mar 2005
Posts: 10589
Location: Somewhere over Atlanta, Georgia
|
| Posted: Sun Oct 04, 2009 1:39 am Post subject: |
|
|
| Code: | | awk -F '=' '{ print toupper($1) substr($0,length($1)+1) }' oldfile >newfile |
- John
_________________
I can confirm that I have received between 0 and 499 National Security Letters. |
|
| Back to top |
|
|
wswartzendruber
Veteran
Joined: 23 Mar 2004
Posts: 1261
Location: Idaho, USA
|
| Posted: Sun Oct 04, 2009 1:42 am Post subject: |
|
|
| john_r_graham wrote: | | Code: | | awk -F '=' '{ print toupper($1) substr($0,length($1)+1) }' oldfile >newfile |
- John |
Huh?! One line?
* Goes off to find rope and a tree. *
_________________
Git has obsoleted SVN.
10mm Auto has obsoleted 45 ACP. |
|
| Back to top |
|
|
John R. Graham
Administrator
Joined: 08 Mar 2005
Posts: 10589
Location: Somewhere over Atlanta, Georgia
|
| Posted: Sun Oct 04, 2009 3:36 pm Post subject: |
|
|
Yeah. I think that AWK code results in the most readable and understandable implementation of the standard *nix text processing languages. If you can guarantee that there's only one equals sign on each line, then the AWK gets even more simple and readable: | Code: | | awk -F '=' '{ print toupper($1) "=" $2 }' oldfile >newfile |
sed also results in a very short one liner based on regular expressions: | Code: | | sed -i -r 's/^([^=]*=)/\U\1\E/' filename |
However, it's pretty abstruse. At a glance you (or, at least, I) just can't tell what that does.
You're right that Perl can do the job but, as the best mechanism in Perl is also a regex-based one, it ends up looking almost identical to the sed implementation.
- John
_________________
I can confirm that I have received between 0 and 499 National Security Letters. |
|
| Back to top |
|
|
|
Portage & Programming
