Thanks for posting here.
Well I tested the script for the said scenario “fetching the last table”:

sh mysqldumpsplitter.sh SQL-FILE LAST-TABLE-NAME

and it worked fine for me.
Can you tell me your exact scenario?

Thanks,
Kedar.

/*!40101 SET SQL_MODE=@OLD_SQL_MODE */;
/*!40014 SET FOREIGN_KEY_CHECKS=@OLD_FOREIGN_KEY_CHECKS */;
/*!40014 SET UNIQUE_CHECKS=@OLD_UNIQUE_CHECKS */;
/*!40101 SET CHARACTER_SET_CLIENT=@OLD_CHARACTER_SET_CLIENT */;
/*!40101 SET CHARACTER_SET_RESULTS=@OLD_CHARACTER_SET_RESULTS */;
/*!40101 SET COLLATION_CONNECTION=@OLD_COLLATION_CONNECTION */;
/*!40111 SET SQL_NOTES=@OLD_SQL_NOTES */;

Thought I’d pass this helpful tip to those who’s *.sql files are stored/saved as *.sql.gz (compressed) files.

To avoid having to uncompress a full backup dump first (especially if its very large; and disk space is a premium).. I may the following minor code mods..

change all grep cmds to zgrep

change all lines
sed -n “/^– Table structure for table \`$tablename\`/,/^– Table structure for table/p” $1 > $TARGET_DIR/$tablename.sql
to
zcat $1 | sed -n “/^– Table structure for table \`$tablename\`/,/^– Table structure for table/p” >$TARGET_DIR/$tablename.sql

I hope you’ve downloaded the MyDumpSplitter.sh file from download link.
Now, “backup_feb15.log” seems a log file than a backup file! Are you sure that its a mysqldump file?

-Kedar

cdr_backup# ./MyDumpSplitter.sh backup_feb15.log billing
0 Table extracted from backup_feb15.log at .

Please help me on this
Thanks,
Sandeep

Could you please help? I get this when run without any arguments. The filname is “sp.sh”. The output is:

: command not found
: command not found
sp.sh: line 23: syntax error near unexpected token `elif’
‘p.sh: line 23: `elif [ $# = 1 ]; then

Please help.

<?php
$db = null;
$outfp = null;
$infp = fopen($argv[1], 'r');
while($line = fgets($infp)){
if(preg_match('/– Current Database: `([^`]+)`/',$line,$matches)){
if($outfp){
fclose($outfp);
}
$db = $matches[1];
$outfp = fopen($db.'.db.sql','w');
}
if($outfp){
fwrite($outfp, $line);
}
}

Now about your regexp, for sure if you want to extract “tasks” table you shouldn’t have used REGEXP mode… you should have directly specified the table name as “tasks_XYZ” will obviously a match to the regexp!
I hope I got what you’ve said?!!

Meanwhile, for extracting Databases from a fulldump following will help:

#Shell Script To Extract Database from Full Dump
for dbz in $(grep “\-\- Current Database” $1 | sed ‘s/`//g’ | awk -F “: ” {‘print $2′}|sort|uniq)
do
echo $dbz;
sed -n “/^– Current Database: \`$dbz\`/,/^– Current Database: \`/p” $1 > $dbz.db.sql
done;

One issue I had was that the script didn’t separate tables with the same name, but in different databases. So, I ended up with one .sql file with the drop, create and insert statements for two tables. I was confused for a while, because the resulting table was much too small. Finally, I realized that the smaller second table was replacing the first table when I loaded that .sql file.

Also, when specifying the table name, a regular expression is used that causes any suffix to also match. I had a table named “tasks” and a much larger table named “tasks_archive”, but I wanted only the statements for the tasks table. I assume it took a little extra time to also write out the tasks_archive table statements, but it wasn’t that big of a deal, since I could obviously ignore it.

PS, doing the reverse: loading the distinct tables back into mysql can be easily accomplished with Maatkit’s mk-parallel-restore (which also serves as a general-purpose parallelizing tool last I checked).