#Standard reference to interpreter (like for normal shell scripts):
#!/usr/local/bin/perl

#if you don't want to use STDIN, put an "open()" call here.

#<> reads a line from STDIN and puts it in the special $_ variable.
#The while loop simply iterates over all lines at STDIN.
while(<>)
{
  #The current line from STDIN is accessible via $_;

  #chop() is equal to chop($_) and chops off the last character from the string in $_
  #Hence, the "\n" (newline) at the end of each line get eliminated.
  chop();

  #This is short, regular expression syntax.
  #Basically, the if-test determines if $_ starts with (^) the string "insert into" and the "i" option makes the match
  #case insensitive.
  if (/^insert into/i)
  {
    #Here, we split the current line into pieces, based on the space as delimiter.
    #The result doesn't get stored in an array (or associative array/hash), but in separate variables via
    #the () list operator. 
    #So, the first word (everything up to the first space) gets assigned to "undef", the second as well, the third
    #to $table, etc.
    #You can think of undef as a special variable to which you can assign stuff you don't want to store.
    #As a result, we'll only store the 3rd and 5th words of each INSERT INTO statement in $table and $records.
    (undef,undef,$table,undef,$records)=split(/ /, $_);
    #$records contains a string like (val1,val2,val3),(val4,val5,val6),... (see an example mysqldump file)
    #So, the inserted records are separated by "," and I use split() once more to separate them.
    #This time, all records (since I don't know how many there are), are assigned to an array, called @recArray.
    #(just like '$' denotes a normal, scalar variable, '@' indicates an array and '%' a hash array).
    @recArray=split( /\,/ ,  $records);

    #prints table name
    print $table;     

    #Loop over each record in the @recArray array.
    #In each iteration, the current record in the array is stored in $rec.
    foreach $rec @recArray 
    {
      #$rec now has format (val1,val2,val3,...) => omit the '(' and ')' to get csv
      #If you know sed, the next syntax shouldn't be to awkward.
      #s/()// just substitutes '(' and ')' by nothing. The "g" option makes the substitution global,
      #so all occurrences of '(' and ')' get substituted.
      #=~ operator means "apply the substition to $rec and assign the result in $rec afterwards".
      $rec=~s/()//g;
      #This prints the result.
      print $rec;

      #In fact, coming to think of it,
      print "$rec\n";
      #is probably a better idea, since that will put the csv records on separate lines...
    }
  }
}