In the location hash, the sort parameters are added as &tableID=column,direction, ... ,column,direction (no spaces). This option allows changing the column-direction separator, a comma by default, into the chosen separator.

*NOTE* Do not set this option to use a hash (#), ampersand (&) or equal sign (=) as it will interfere with how the hash parameters are set up.

In the location hash, the sort parameters are added as &tableID=column,direction, ... ,column,direction (no spaces). The tableID is set by this option.

This option setting is prioritized over the actual table ID attribute. If neither are set, the tableID will be set as the table's zero-based index on the page.

sort2Hash_tableID > table.id attribute > table index

This option replaces both sort2Hash_useHeaderText and sort2Hash_processHeaderText options.

The contents of this attribute will be used in the hash to identify a sorted column.

Use the header attribute as follows:

<th data-header="first">First Name</th>

Only the first two values will be used from this array. The first value is assigned to ascending sorts and the second is assigned to descending sorts.

Use the option as follows:

sort2Hash_directionText : [ 'asc', 'desc' ]

*NOTE* When converting the hash into a value, if the direction hash does not match the second value ('desc' in the example above), it will fallback to an ascending sort no matter what text in contained within the first value.

The following example is a duplicate of the internal code that encodes the hash. Adapt to fit your needs:

sort2Hash_encodeHash : function( config, tableId, component, value, rawValue ) {
  // config = table.config settings
  // tableId = processed table ID
  // component: 'sort', 'page', 'size' or 'filter'
  // value = string value that has encodeURIComponent applied
  // rawValue = value ( array or number )
  // return false to let the widget encode the string
  return '&' + component + '[' + tableId + ']=' + value;
}

Return a string to add to the window.location.hash directly.

If this function returns false, then the internal code will encode the hash for that component. This was done in case only one parameter needs special handling:

sort2Hash_encodeHash : function( config, tableId, component, value, rawValue ) {
  // only the filter gets special handling
  if ( component === 'filter' ) {
    // if working with rawValue, make sure to use
    // encodeURIComponent on the results
    var newValue = encodeURIComponent( rawValue.join( '--' ) );
    return tableId + 'filter=' + newValue;
  }
  return false;
}

The following example is a duplicate of the internal code that decodes the hash. Adapt to fit your needs:

sort2Hash_decodeHash : function( config, tableId, component ) {
  // config = table.config settings
  // tableId = processed table ID
  // component: 'sort', 'page', 'size' or 'filter'
  // return false to let the widget decode the hash

  // $.tablesorter.sort2Hash.getParam( parameter, url ); function added in v2.24.4
  // parameter = desired parameter to extract
  // url (optional) = hash or href string to extract the parameter value from
  return $.tablesorter.sort2Hash.getParam( component + '[' + tableId + ']' ) || '';
}

Return the component value or an empty string.

If this function returns false, then the internal code will attempt to decode the hash for that component. This was done in case only one parameter needs special handling:

sort2Hash_decodeHash : function( config, tableId, component ) {
  // only the filter gets special handling
  if ( component === 'filter' ) {
    return $.tablesorter.sort2Hash.getParam( tableId + 'filter' ) || '';
  }
  return false;
}

The following example is a duplicate of the internal code that removes unused values in the hash. Adapt to fit your needs:

sort2Hash_cleanHash : function( config, tableId, component, hash ) {
  // removeParam function added v2.24.4
  return $.tablesorter.sort2Hash.removeParam( component + '[' + tableId + ']' );
}

Return a string of the remaining components or an empty string.

If this function returns false, then the internal code will attempt to clean the hash for that component. This was done in case only one parameter needs special handling:

sort2Hash_cleanHash : function( config, tableId, component, hash ) {
  // config = table.config settings
  // tableId = processed table ID
  // component: 'sort', 'page', 'size' or 'filter'
  // return false to let the widget decode the hash

  if ( component === 'filter' ) {
    // removeParam function added v2.24.4
    return $.tablesorter.sort2Hash.removeParam( tableId + 'filter' );
    /* ORIGINAL METHOD, use if the "getParam" function doesn't work on your custom parameter
    var index,
      result = [],
      // regular expression to match the component & value
      regex = new RegExp( tableId + 'filter=([^&]*)' );
      // split hash into component parts
      parts = ( hash || '' ).slice(1).split( '&' ),
      len = parts.length;
    // cycle through each component part
    for ( index = 0; index < len; index++ ) {
      // if the component doesn't match the regular expression...
      if ( !regex.test( parts[ index ] ) ) {
        // then save it
        result.push( parts[ index ] );
      }
    }
    // if we still have something left, join the components back together
    // and return it so the next component can be processed
    // we don't update the window.location.hash, or the page jumps to the top!
    return result.length ? c.widgetOptions.sort2Hash_hash + result.join( '&' ) : '';
    */
  }
  return false;
}

If true, text from the header is used instead of a zero-based column index.

Please be aware that if the column text contains spaces or special characters, they will be encoded in the URL. So, "First £$€¤¥¢ Name" will become "First%20%C2%A3$%E2%82%AC%C2%A4%C2%A5%C2%A2%20Name". This would make the hash very difficult to read.

Further processing of this header cell text can be done using the sort2Hash_processHeaderText function.

If the sort2Hash_useHeaderText option is true, a function here will further process the header cell text.

Use this function to perform any processing on the header cell text, as desired.

At this point, the header cell text has not been encoded.

Here is one example:

sort2Hash_processHeaderText : function( text, config, columnIndex ) {
  // remove all non-alphanumeric characters (including spaces)
  return text.replace( /[^a-z0-9]/gi, '' );
}

Another example:

sort2Hash_processHeaderText : function( text, config, columnIndex ) {
  // completely custom text to use for the hash
  // this method assumes that the table layout is constant
  // (i.e. columns are not added, removed or rearranged)
  return [ 'first', 'last', 'age', 'total', 'disc', 'date' ][ columnIndex ];
}

This function will extract the value of the name passed to the function. Use it as follows:

// $.tablesorter.sort2Hash.getParam( name, hash );
// name = parameter name (key)
// hash = (optional) if undefined, it will get the string value from window.location.hash
var hash = '&sort[table-users][user_name]=asc&page[table-users]=1&size[table-users]=10&filter[table-users][user_name]=ad',
	result = $.tablesorter.sort2Hash.getParam( 'filter[table-users][user_name]', hash );
// result = "ad";

This function will remove the key & value from the passed hash. Use it as follows:

// $.tablesorter.sort2Hash.removeParam( name, hash );
// name = parameter name (key)
// hash = (optional) if undefined, it will get the string value from window.location.hash
var hash = '&sort[table-users][user_name]=asc&page[table-users]=1&size[table-users]=10&filter[table-users][user_name]=ad',
	result = $.tablesorter.sort2Hash.removeParam( 'filter[table-users][user_name]', hash );
// result = "&sort[table-users][user_name]=asc&page[table-users]=1&size[table-users]=10";

*NOTE* This function does not update window.location.hash.