Add --chunk-index-columns to pt-osc.

bin/pt-online-schema-change

@@ -1893,6 +1893,7 @@ sub generate_asc_stmt {
       @asc_cols = $asc_cols[0];
    }
    elsif ( my $n = $args{n_index_cols} ) {
+      $n = scalar @asc_cols if $n > @asc_cols;
       PTDEBUG && _d('Ascending only first', $n, 'columns');
       @asc_cols = @asc_cols[0..($n-1)];
    }
@@ -5209,14 +5210,8 @@ sub main {
       }
    }
 
-   # Parse --chunk-index INDEX:N where N is the number of
-   # left-most columns of INDEX to use.
    # https://bugs.launchpad.net/percona-toolkit/+bug/1010232
-   my ($chunk_index, $n_chunk_index_cols)
+   my $n_chunk_index_cols = $o->get('chunk-index-columns');
-      = split(':', $o->get('chunk-index') || '');
-   if ( defined $chunk_index && $chunk_index eq '' ) {
-      $o->save_error('--chunk-index cannot be an empty string');
-   }
    if ( defined $n_chunk_index_cols
         && (!$n_chunk_index_cols
             || $n_chunk_index_cols =~ m/\D/
@@ -6108,8 +6103,8 @@ sub main {
       Cxn                => $cxn,
       tbl                => $orig_tbl,
       chunk_size         => $orig_tbl->{chunk_size},
-      chunk_index        => $chunk_index,
+      chunk_index        => $o->get('chunk-index'),
-      n_chunk_index_cols => $n_chunk_index_cols,
+      n_chunk_index_cols => $o->get('chunk-index-columns'),
       dml                => $dml,
       select             => $select,
       callbacks          => $callbacks,
@@ -7453,16 +7448,18 @@ behavior of choosing an index.  The tool adds the index to the SQL statements in
 a C<FORCE INDEX> clause.  Be careful when using this option; a poor choice of
 index could cause bad performance.
 
-This option supports a special syntax to select a prefix of the index instead of
+=item --chunk-index-columns
-the entire index. The syntax is NAME:N, where NAME is the name of the index, and
+
-N is the number of columns you wish to use. This works only for compound
+type: int
-indexes, and is useful in cases where a bug in the MySQL query optimizer
+
-(planner) causes it to scan a large range of rows instead of using the index to
+Use only this many left-most columns of a L<"--chunk-index">.  This works
-locate starting and ending points precisely. This problem sometimes occurs on
+only for compound indexes, and is useful in cases where a bug in the MySQL
-indexes with many columns, such as 4 or more. If this happens, the tool might
+query optimizer (planner) causes it to scan a large range of rows instead
-print a warning related to the L<"--[no]check-plan"> option. Instructing
+of using the index to locate starting and ending points precisely.  This
-the tool to use only the first N columns from the index is a workaround for
+problem sometimes occurs on indexes with many columns, such as 4 or more.
-the bug in some cases.
+If this happens, the tool might print a warning related to the
+L<"--[no]check-plan"> option.  Instructing the tool to use only the first
+N columns of the index is a workaround for the bug in some cases.
 
 =item --chunk-size
 

bin/pt-table-checksum

@@ -6171,8 +6171,6 @@ sub main {
       }
    }
 
-   # Parse --chunk-index INDEX:N where N is the number of
-   # left-most columns of INDEX to use.
    # https://bugs.launchpad.net/percona-toolkit/+bug/1010232
    my $n_chunk_index_cols = $o->get('chunk-index-columns');
    if ( defined $n_chunk_index_cols