The Perl Toolchain Summit needs more sponsors. If your company depends on Perl, please support this very important event.

NAME

Categories - Create and process categories within MySQL DB

VERSION

Categories.pm ver.1.0

DESCRIPTION

    Categories allows you to create and process categories (for products/directories/shops and etc...)

SYNOPSIS

 There is an example that you may use in your own CGI scripts:

 # --- Script begin here ---
 use Categories;

 # NOTE: new() method will create needed DB structure in MySQL (database & tables) if they not exist!
 #       Please create database before execute this script or DB USER must have privilege to create DBs!
 
 $obj = Categories->new(database => 'catsdb', user => 'db_user', pass => 'db_pass', host => 'localhost');
      # OR
      # $obj = Categories->new(dbh => $mysql_dbh_handler);
 
 if($obj)
  {
   my $comp_id = $obj->add(type=>'category',name=>'Computers',category=>0);
   my $film_id = $obj->add(type=>'category',name=>'Films',category=>0);
   my $matr_id = $obj->add(type=>'item',name=>'The Matrix',category=>$film_id,value=>'');
   my $one_id  = $obj->add(type=>'item',name=>'The One',category=>$film_id,value=>'');
   my $cpu_id  = $obj->add(type=>'category',name=>'CPU',category=>$comp_id);
   my $hdd_id  = $obj->add(type=>'category',name=>'HDD',category=>$comp_id);
   my $xp18_id = $obj->add(type=>'item',name=>'Athlon XP 1800+',category=>$cpu_id,value=>'');
   my $xp20_id = $obj->add(type=>'item',name=>'Athlon XP 2000+',category=>$cpu_id,value=>'');
   my $xp21_id = $obj->add(type=>'item',name=>'Athlon XP 2100+',category=>$cpu_id,value=>'');
   my $hdd1_id = $obj->add(type=>'item',name=>'Maxtor 80 GB',category=>$hdd_id,value=>'30 months warranty');
   my $hdd2_id = $obj->add(type=>'item',name=>'Maxtor 120 GB',category=>$hdd_id,value=>'36 months warranty');
   
   my @res = $obj->read(path=>'//Computers//HDD//*',sort=>'ID',preload=>YES,reverse=>NO,partial=>'YES');
   print "<HR>";
   if(scalar(@res))
     {
      foreach $l (@res)
       {
         my ($id,$parent_category,$name,$value,$route_path) = @$l;
         print "ID:     $id<BR>\n";
         print "PARENT: $parent_category<BR>\n";
         print "NAME:   $name<BR>\n";
         print "VALUE:  $value<BR>\n";
         $route_path =~ s~//~\\~sgi;
         $route_path =~ s~\\(.*?)\x0~\\~sgi;
         print "PATH:   $route_path<BR>\n";
         print "<HR>";
       }
     }
   
   # Find categories and items (filter=>ALL) that has NAME (by=>NAME) 'The Matrix' order by ID (sort=>ID)
   # and return multiple results (multiple=>'YES') if available, also return rout path to this element 
   # (route=>'YES') using category cache (preload=>'YES') to speed up searching. However 'preload' option may be
   # worse if categories table is too long, because script load whole table and may crush if not enough memmory!
   
   my @res = $obj->find('search'=>'The Matrix','sort'=>'ID','by'=>'NAME','filter'=>'ALL','multiple'=>'YES',
                        'route'=>'YES','preload'=>'YES','partial'=>'NO','reverse'=>'NO');
   if(scalar(@res))
     {
      foreach $l (@res)
       {
         my ($type,$id,$parent_category,$name,$value,$route_path) = @$l;
         print "Type:   $type<BR>\n";
         print "ID:     $id<BR>\n";
         print "PARENT: $parent_category<BR>\n";
         print "NAME:   $name<BR>\n";
         print "VALUE:  $value<BR>\n";
         $route_path =~ s~//~\\~sgi;
         $route_path =~ s~\\(.*?)\x0~\\~sgi;
         print "PATH:   $route_path<BR>\n";
       }
     }
    print "<HR>";
    
    # Modify: Change PARENT/CID and/or NAME
    $obj->modify(id=>$xp21_id,type=>'item',name=>'Duron 1300 MHz',value=>'6 months warranty');
    $obj->modify(id=>$comp_id,type=>'category',name=>'PC');
    $obj->modify(id=>$cpu_id,type=>'category',newcid=>0);
    
    $obj->deep_traverse('preload'=>'YES','id'=>0,'level'=>0,'path'=>'//','eval'=>\&Walk,'sort'=>'NAME');
    
    # Delete ROOT category, so all items/categories are deleted!
    $obj->del(type=>'category',id=>0);
   }
  else
   {
    print $Categories::error;
   }
   
 sub Walk
 {
  my $self = shift;
  my %inp  = @_;
 
  my $id              = $inp{'id'};
  my $level           = $inp{'level'};
  my $separator       = $inp{'separator'};
  my $path            = $inp{'path'};
  my $name            = $inp{'name'};   # ITEM only ($type='I')!
  my $value           = $inp{'value'};  # ITEM only ($type='I')!
  my $type            = $inp{'type'};
 
  $path =~ s~$separator~\\~sgi;
  $path =~ s~\\(.*?)\x0~\\~sgi;
  print $path."$name"."[$value]<BR>";
 }
 # --- Script ends here ---

SYNTAX

 That is simple function reference:
 
 # Create object of Categories type
 $object = Categories->new(database=>'catsdb', user=>'db_user', pass=>'db_pass', host=>'localhost', 
                           port=>'3306', create=>'Y', checkdb=>'Y', name=>'catdb', dbh=>$connect_db_handler);
    Where:
    database  - is your DB where categories (tables) will be placed. If database not exist module
                will try to create one for you, using supplied user and password. [REQUIRED if $dbh empty]
    user/pass - is your DB user and password [REQUIRED if $dbh empty]
    host      - MySQL DB host
    port      - your MySQL port
    create    - module will attempt to create DB and/or tables
    checkdb   - module will try to check DB structure
    name      - name of category object
    dbh       - you can supply already connected database handler instead of database/user/pass/host/port!
    
 
 
 # Test database structure (tables only)
 $state = $object->is_tables_exists(name=>'name_of_category_object');
    
    
 
 # Create database structure (tables only - database must exist!)
 $state = $object->create_tables(name=>'name_of_category_object');
    Create table structure (database should exist!)
 
 
 
 # Clear categories cache
 $state = $object->clear_cache();
 
 
 
 # Reload/create categories cache and return array of all categories. @cats is array of references to hashes;
 @cats = $object->preload_categories(name=>'name_of_category_object', sort=>'NAME', reverse=>'N');
    Where:
           sort      - is name of column (order by),
           reverse   - reverse results (DESC)
    HINT: $ref = $cats[0]; %hash = %$ref; $name = $hash{'NAME'}; $id = $hash{'ID'}; $parent = $hash{'PARENT'};
    
 
 
 # Find category/item by 'by' column, searching by 'search' keyword (may be on part of word 'partial'=>'Y');
 # also it may return full path to located item/category (route=>'Y') using categories cache (preload=>'Y')
 @res = $object->find(caseinsensitive=>'Y', filter=>'ITEMS', multiple=>'Y', by=>'ID', sort=>'NAME',
                      reverse=>'N', partial=>'N', search=>'keyword', check=>'N', route=>'N',
                      separator=>'//', preload=>'Y');
    Where:
          caseinsensitive  - search is caseinsensitive,
          filter           - define where sub must search (ITEMS,CATEGORIES,ALL),
          multiple         - allows muliple results,
          by               - search BY column,
          sort             - 'order by' all results,
          reverse          - reverse all results,
          partial          - allows partial search ( LIKE %something%),
          search           - search keyword,
          check            - test tables structure,
          route            - find path to root,
          separator        - use follow separator to separate categories in route path,
          preload          - allows categories cache.
    Method returns array of references to arrays results in follow syntax:
          @res = ($ref_1[,$ref_2,...]), where $ref_n is reference to array with follow structure:
          @$ref_n = ([0],[1],[2],[3],[4],[5]);
          where:
          [0] - 'I' or 'C' (Item or Category),
          [1] - ID of Item/Category,
          [2] - PARENT (category ID),
          [3] - NAME of Item/Category,
          [4] - VALUE of Item (ONLY) or empty string if Category,
          [5] - If route=>'Y' this will has follow syntax:
                $separator.$ID."\x0".$CATEGORY_NAME[$separator...]
 


 # Add ITEM/CATEGORY to categories tree. If proceed ITEM you can set 'value' to it.
 $id = $object->add(type=>'ITEM', category=>'0', name=>'Name_Of_Element', check=>'N', value=>'Items only');
     Where:
           type      - is type of element ('ITEM' and 'CATEGORY'),
           category  - is ID of parent (0 is root),
           name      - name of new item/category,
           value     - value of item (only)
           check     - test tables structure.
           $id is ID of created element.



 # Delete ITEM/CATEGORY (recursive) by it's own ID
 $cnt = $object->del(type=>'ITEM', id=>'0', check=>'N', preload=>'Y');
      Where:
            type      - is type of element ('ITEM' and 'CATEGORY'),
            id        - is ID of Item/Category (0 is root),
            check     - test tables structure,
            preload   - allows categories cache.
            $cnt is number of affected(deleted) rows.



 # Modify(update/rename/move) given ITEM/CATEGORY. If some parameter missed method will not change
 # current ITEM value (name,parent,value)
 $cnt = $object->modify(type=>'ITEM', id=>'id_of_element', newcid=>'id_of_new_parent', check=>'N',
                        name=>'new_name_of_element', value=>'Items only', preload=>'Y');
      Where:
            type      - is type of element ('ITEM' and 'CATEGORY'),
            id        - is ID of Item/Category,
            check     - test tables structure,
            name      - new name of item/category (if you dismiss filed, NAME will not be affected!),
            value     - new value of item (if you dismiss filed, VALUE will not be affected!),
            newcid    - new Parent category ID (if you dismiss filed, PARENT will not be affected!),
            preload   - allows categories cache.
            $cnt is number of affected(deleted) rows.
 
 
 
 # Traverse category tree (in width) and calling 'eval' callback sub for any found category!
 $cnt = $object->traverse(cid=>'id_of_category', eval=>\&callback_sub', check=>'N',
                          sort=>'NAME', reverse=>'N', preload=>'Y');
      This sub traverse in width.
      Where:
            cid      - ID of category that should be traversed,
            eval     - reference to sub that will be called for every category,
                       it will be called as: &$eval($self,'id'=>$current,'parent'=>$cid);



 # Traverse category tree (in deep) and calling 'eval' callback sub for any found CATEGORY/ITEM!
 $object->deep_traverse(id=>'id_of_category', level=>'0', separator=>'//', path=>'//',
                        eval=>\&callback_sub', sort=>'NAME', reverse=>'N', check=>'N',
                        preload=>'Y');
      deep_traverse is recursive sub and it traverse in deep. At fist step level should be '0' and
      path '//' (like separator); eval is also reference to callback sub and it will be called as:
      
      &$evala($self,'id'=>$id,'level'=>$level,'type'=>$whereis,'path'=>$path,
              'name'=>$item_name,'value'=>$item_value,'separator'=>$separator);
      where 'name'/'value' will be available only for Items (type=>'I'), but not for categories (type=>'C')
      
      
 
 # Load found categories/items for 'cid' category only! (no recurse)
 @res = $object->load_category(cid=>'id_of_category', sort=>'NAME', reverse=>'N', preload=>'Y');
      This method will load only Items/Categories of 'cid' category (without recurse)!


 
 # Read category/item[s] properties(data) searching by registry-like path (only by 'NAME')!
 @res = $object->read(path=>'//path//to//category//item', sort=>'ID', reverse=>'N', preload=>'Y',
                      partial=>'N', caseinsensitive=>'Y', separator=>'//', preload=>'Y', check=>'N');
      This method try to locate category or item[s] only by 'NAME' starting by 'root' category(PARENT=0),
      Where:
            path      - is 'registry'-like path to category/item[s]. If path ends with '//' (separator)
                        then searching is proceed for category, other else for item. If name of item is
                        '*' then all items of this 'path' will be fetched!
            partial   - This allows searhing for partial items (not for path or categories)! Moreover to
                        use '*' feature you must set partial=>'YES'
      Method returns array of references to arrays results in follow syntax:
            @res = ($ref_1[,$ref_2,...]), where $ref_n is reference to array with follow structure:
            @$ref_n = ($ID,$PARENT_ID,$NAME,$VALUE,$PATH);
            $PATH has follow syntax:  $separator.$ID."\x0".$CATEGORY_NAME[$separator...]

AUTHOR

 Julian Lishev - Bulgaria, Sofia, 
 e-mail: julian@proscriptum.com, 
 www.proscriptum.com