Transcript Content Management intern - THM

Content Management intern
„CMS“ definition
quotation:
Unter Content Management wird die systematische und
strukturierte Beschaffung, Erzeugung, Aufbereitung,
Verwaltung, Präsentation, Verarbeitung, Publikation und
Wiederverwendung von Inhalten (Content)
informationstechnisch vorgehaltener Informationen
verstanden.
(Rothfuss/Ried „Content Management mit XML “2001)
Overview
•
•
•
•
•
•
CMS classification
WebCMS classification
WebCMS components in detail
Operating and migration considerations
mid size vs. enterprise CMS comparison
CMS customization example
CMS classification
Typical cms groups are:
• Website-CMS (WebCMS, WCMS)
• Document Management Systems (DMS)
• MAM (Media Asset Management Systems)
• Portal Systems
WebCMS classification
by size / usage
• Small systems ( „small-business“, „lite“ )
• Midsize systems
• Huge systems („enterprise“)
CMS, technical view
environment, programming languages
•
•
•
•
Web Server
Programming Language
Database System
Content, Content Storage, Content
Objects
• Content Life Cycle
• Rights Management
• Content Output
Web Servers
• Common Web Servers (i.e. Apache, IIS)
are typically used as CMS environment
• enterprise level CMS sometimes use own
web servers which are dedicated to deliver
database-held content
• some CMS use application server
environments (ColdFusion, ...)
Programming Languages
• small and midsize CMS must integrate
easily into common web environments.
=> common scripting languages are used
• Use of new technologies such as J2EE
enables building of large-scale eCMS
Databases
• CMS must use common relational
databases for easy setup and integration
into existing environments
• However, CMS content is mostly
organized in object-oriented or XML-form
=> Content is mostly stored in RDBMS, but
some CMS use their own or application
server‘s mapping for (xml|oo)<->rdbms
Content
• To achieve Separation of content and presentation ,
documents have to be split into small pieces, like media
elements or text paragraphs.
• this results in a hierarchical structure of content
elements, with layout elements (i.e. html templates)
corresponding to sets of content elements
• The structure‘s depth varies among different CMS,
typical depths are:
- 1 (results in one big editor per page)
- 2 („container“-based editing )
- ∞ (object-oriented systems)
Content Life Cycle
The content lifecycle consists of these 5 steps:
•
•
•
•
•
Creation / import
review / editing
validation
publishing
archiving (includes deletion)
Every CMS implements these steps, although some
systems combine some of them into one workflow step.
Rights Management
• applies to multi-user CMS
• content elements and/or hierarchical
elements (i.e. folders) can be assigned
access rights
• typically rights are split into create / modify
/ publish / delete [ / rights manage ] rights.
• easy system setup can be achieved with
preconfigured roles.
Output
• Staging
the CMS‘s renderer ist not accessible from the
internet, web pages are therefore delivered to a
normal web server. The user accesses a static
export of the web site.
• Live Rendering
The requested web pages are rendered on the
fly and directly delivered from the CMS to the
user.
staging vs. live rendering I
Staging
+ Good performance, as only normal html-pages are
delivered. Especially no database bottlenecks
+ good scalability because the static web site exports can
easily be replicated to other servers
+ high security because the CMS can never be
compromised
- can only manage „normal“ web sites, cannot manage
dynamic pages (i.e. forums, communities, feedback
forms, personalization)
- content can be updated only in configured intervals
- CMS cannot help with web site statistics
staging vs. live rendering II
Live Rendering
+ Possibility to maintain full-scale Web
Applications / dynamic Web Sites
+ detailed statistics possible (i.e. click paths)
- low performance, especially for sripting
language CMS
- scalability hard to establish
- additional security measures must be
established in order to guard the CMS
Content Import, Export
• Content migration between most CMS is
nearly impossible because of fundamental
differences in the content hierarchy
• list-based im-/export of content fragments
(articles, container) is often possible
• xml-based formats are commonly used for
lossless export of DB-held content
Export considerations
these content groups can be exported and
may be imported in another System:
• Pages (page hierarchy)
• Page Contents (to be imported as
separate Containers in case of different
content storage hierarchies)
• Ressources (Media Assets)
• User / Group / Role Lists
Interfaces
external databases
There are two ways of accessing external
DBMS from cms:
1)
direct acces via template programming
Code to access the DB is directly placed in template, and is
executed during the rendering process (live cms) or on the
staging web server
2)
access through data source abstraction
data source is defined to the CMS‘s abstraction layer and
can afterwards be accessed with common CMS functions
Interfaces
Content Syndication
• Content Syndication means considering (own)
content (-elements) as assets which can be
used in different web sites
• CMS can assist in delivering content to other
web sites or integrating external content
• xml-based formats for article import and export
exist, for example NITF
CMS, technical view
example I
CMS, technical view
Example II
Customization Example
Adding external user authentication
Two strategies to use external user
databases for CMS:
• Periodically import user/group information
from external sources
• Drop internal user management, directly
use external source instead for all
requests
LDAP
Leightweight Directory Access Protocol
LDAP:
Access protocol for directory service databases (X.500)
defined in RFC1777
Directory Service:
• Database for attribute-based records (name-value pair
lists)
• Record groups are organized in trees
LDAP
Test Setup
This .ldif-File represents one x.500-entry for testing
purposes. The original object class InetOrgPerson has
been extended by cms-specific fields
hans_meiser.ldif:
dn: [email protected],o=test,c=de,dc=de
objectclass: MyInetOrgPerson
cn: Hans Meiser
mail: [email protected]
uid: hans
passwd: geheim
cmsgroup: user
sn: nix
telephoneNumber: 01234 / 5678 -90
WCMS user authentication
• using Web-based CMS means requesting independent web
pages
• web page context is a common Session, identified by a
session id
• When a user logs in with username and password, his user
data and rights info are added to the current session object.
 User authentication means:
session_id, username and password in,
userdata-added session or user_id from which data is to
add out.
WCMS User Authentication
Example Code Fragment
// Search for given username in CMS user DB
$all = db_query( "SELECT * FROM system_user
WHERE username=\"$username\"" );
if( $row = mysql_fetch_assoc( $all ) ) {
<?
/*
External user authentication API
in: $username byref
in: $password (plain) byref
if( $row['disabled'] ) {
// found but disabled? Login not allowed!
$valid = 0;
out: $valid = 0|1
out: $uid (system_user PK)
} else {
// user found, return user_id
$uid = $row['id'];
*/
}
} else {
// Username is valid but not known in local
DB:
// insert user record.
// Password remains empty => login without
external authentication will not be
possible
db_query( "INSERT INTO system_user " .
"SET username=\"$username\",
realname=\"$username\", created=NOW()" );
$uid = mysql_insert_id( );
db_query( "INSERT INTO system_user2group
SET user=$uid, in_group=1" );
}
// Example: Call external program to validate
user
$cmd = "(echo \"$username\"; echo \"$password\")
| /home/sayn/pwtest";
$p = popen($cmd, 'r');
while( !feof($p) )
$ret .= fread($p, 1024);
pclose($p);
$ret = explode( "\n", $ret );
if( $ret[0] != 'OK' ) {
$valid = 0;
} else {
$valid = 1;
}
WCMS User Authentication
using LDAP
// Example: lookup userdata in ldap directory
$ds=ldap_connect("localhost", 389);
ldap_set_option($ds, LDAP_OPT_PROTOCOL_VERSION, 3) // Set PRotocol to LDAPv3
$r=ldap_bind($ds); // anonymously bind for read-only access
// Lookup username in directory
// in this example, users should log in with their email address
$sr=ldap_search($ds, "o=test, c=de, dc=de", "mail=$username");
// Authentication is only possible if search returned one item
if( ldap_count_entries($ds, $sr) != 1 ) {
$valid = 0;
} else {
$info = ldap_get_entries($ds, $sr);
$username = $info[0]['uid'][0];
// plain text password comparison. Remember, this is just an example
if( ( $password=='' ) || ( $password != $info[0]['passwd'][0] ) ) {
$valid = 0;
} else {
...
other LDAP integration
• associate user groups to ldap subtrees
• use host system user authentication as
abstraction (see example 1)
• certificate-based authentication (user logs
on with ....)