Localization of Open Source Web Content Management System

Rufael Tadesse Fekade Getahun

HiLCoE, Computer Science Programme, Ethiopia HiLCoE, Ethiopia
Computer Revolution Africa, Ethiopia Department of Computer Science, Addis Ababa
[email protected] University, Ethiopia
[email protected]

Abstract
Content Management Systems (CMSs) are being applied by several firms for the purpose of facilitating
document control, auditing, editing and managing timeline. Most CMS provide Web-based publishing, format
management, revision control (version control), indexing, search, and retrieval. The open source CMSs are
becoming popular due to low cost, easy customization, easy to use, workflow management and built-in support
for foreign languages. However, users using local language are unable to understand the system and several
components related to it. Focusing on the localization aspect of CMS, several works have been done. The works
vary on the CMS selected to be localized and the approach followed during the localization process.
In this paper, the objective is to localize CMS into Ethiopic languages and enable non-expert users to be site
developers, content contributors and content creators. Various CMSs are assessed and Joomla Web Content
Management System (WCMS) is selected to be localized. Templates and modules based localization is adopted
and basic components of CMS such as entities, concepts, and ideas are localized into Ethiopic languages
(Amharic, Oromigna and Tigrigna) by creating a language pack. To empower developers and allow users in
writing using local language, a virtual keyboard for Amharic is also designed and embedded in to the CMS. The
localized Joomla has been given to end users and programmers to design their own website. They developed the
provided websites using the localized Joomla in ways required and made use of the designed virtual keyboard
with relative ease.
Keywords: Open Source Software; Content Management System; Localization; Joomla Web Content
Management System; Language Pack; Translation Component; Virtual Keyboard

implementation, and publication of content to a

1. Introduction
website and avoid the need for the user to possess
Recently, free and open source software have technical knowledge of editing hypertext mark-up
become common in many countries. Open-source language (HTML) files. Such systems allow content
software (OSS) is a software with its source code creators to directly add or update content on a
made available and licensed with free or affordable website without involving an Information
license fee in which the copyright holder provides the Technology (IT) department or external company.
rights to study, change, and distribute the software to WCMS, therefore, empowers members of each
anyone and for any purpose. Open-source software is department within an organization to update their
very often developed in a public, collaborative own sections on a website thereby reducing the
manner. Open-source software is the most prominent number of people involved in the end-to-end process
example of open-source development and often and ensuring that timely updates can be made [2].
compared to (technically defined) user-generated
In Ethiopia, the number of people using mostly
content or open-content movements [1].
native language as a form of communication tends to
The concept of open source has also been applied cover most of the country’s population. According to
in WCMS to facilitate the design, development, the 2007 Ethiopian census, most people speak
HiLCoE Journal of Computer Science and Technology, Vol. 2, No. 2 59

Oromigna, Amharic; Somaligna, Tigrigna; 2. Related Work

Sidamigna or Sidaamu Afoo, Wolaytta, Guragigna,
Several works have been done in the area of
Afar: 'Qafár af' and other languages [3]. Due to this,
localizing open source software. Each work differs
many people face challenges in using and adapting
from the other based on the set objective and the
technological solutions due to the required
selected content management system. Related to
knowledge on foreign languages. One of the
local context, there are two prominent localization
solutions to inculcate such population to the use of
attempts. In [5], the authors presented an approach to
technology is to localize systems into local context.
localize Open CMS for managing documents
Localization is the process of adapting, translating electronically, improving the internal operations of
and customizing a product (software) for a specific an organization, raising community satisfaction, and
target [4]. Localizing Open Source WCMS is simplifying document management and exchange.
performed to empower users into creating, managing The project base its domain on only Ethiopian
and editing content using local languages such as governmental organizations in the process of the
Amharic. WCMSs have different kinds of Open CMS alteration and puts more focus on trying
components embedded within a package to use the CMS for document management.
Considering a real world scenario, web content In [6], localizing Joomla content management
programmers prefer to work on open source CMS system 1.5 into Amharic to make professional and
than proprietary WCMSs due to cost, environment non-professional individuals to take part in web site
restriction, ownership and flexibility in modifying development has been conducted. The front and back
core functionality as required. However, content end interfaces are translated into Amharic and a
developers still face problems in understanding and virtual keyboard is designed to make Amharic text
making good use of predefined components of the entry easier. The developed language pack and
WCMS as local language is not handled. Most feel designed virtual keyboard is compatible to work with
comfortable if they have readymade components in Joomla version 1.5 and cannot be imported into the
local language that can be embedded into a website latest version of Joomla. The author points out
rather than using components built in foreign localizing extensions and templates as part of future
language. work.
The best option to facilitate such requirements is This research addresses localizing extensions and
to localize the WCMS including its components. templates and lays down an approach to extend the
Customizing Open Source WCMS together with its extension to other Ethiopic language as well.
components to support local languages enables users
(programmer or naïve user) with preference of using 3. The Proposed Solution
local language to operate and develop a web site The main objective of this research is to localize a
easily. CMS by identifying the popular CMS and assessing
The rest of this paper is organized as follows. users’ experience in using the CMS. In order to
Section two discusses related works. Section three identify the popular CMS, ten CMS products were
presents the proposed solution in which this research selected based on a study that uses WAP Analyser
aims to address and describes the main parts of the which unleashes WCMS technologies used. A total
solution. Section four focuses on discussion on the of eight criteria have been considered for evaluating
findings and comparison is made in relation to the CMSs where seven of the criteria are documented
similar works. Section five summarizes the paper and by different researchers (i.e., Usability, Performance,
provides areas for future works to be focused on. Security, Built in applications, Flexibility and Ease
of Customization Management and Interoperability
60 Localization of Open Source Web Content Management System

and Documentation and Support) and one has been the management provide the availability of Package
added to cover the requirements and the set creation, Front-end and Backend management after
objectives of the research (Modules/Plug-ins an end user has been authenticated. The translation
Extensibility and Template Usage). CMS matrix and database holds data translated by end users for local
Weight scoring method has been applied on the ten languages.
WCMSs to select the top three WCMS. After the top The translation component is a tool that is
three had been selected, a questionnaire was developed to support a translation mechanism for the
distributed to thirty website development companies newer version of Joomla to support Amharic,
in order to gather data related to use, experience and Oromigna and Tigrigna languages. The architecture
popularity of CMSs in the Ethiopian context. for the translation component is in alignment with the
The proposed solution focuses on providing a architecture of Joomla where the emerged component
localization means for the selected CMS, i.e., is embedded on the external layer of Joomla
Joomla. One part of localization that this research architecture. The component is designed to fill the
covers is minimizing the gap between end users and gap that exists within the language strings and
translation values with in Joomla. The general elements within Joomla architecture. The translation
architecture used to localize the CMS is shown in component has package creation and language
Figure 1. The architecture is a layer based management subsystems. The access control and
architecture and comprises an interface, database, security is dependent on the security structure for
authentication/access checking and management for backend administration of Joomla. Layered
front-end, backend and package creation. The architecture is used for the interfacing of subsystems
interface layer serves as a line between end users and and organises the component into a set of layers. The
the developed component whereas the component is developed in a way to convert file
authentication/access checking layer bases on a based structure into a database structure by storing all

Web Interface
Request
Access Checking and Authentication
List

Virtual Keyboard Backend Frontend Package

Interface Translation Translation Management
Management Management
Layout Creation and Translation Database
Input Management

security feature provided by Joomla to authenticate language related strings in a database.

an end user before using the component. Layers for
Figure 1: General Architecture

A translation database is designed to match the enable users access strings found with in a file on a
file based language files and categorized in a way to well-constructed and easy to manage interface. The
database works in collaboration with the component
HiLCoE Journal of Computer Science and Technology, Vol. 2, No. 2 61

developed. The component contains an interface that component allows translation for three Ethiopic
enables users to translate strings into the language languages. It enables users to translate, edit and
that they need. This capability is shown in Figure 2. create language packs for Joomla. The created
The component reads the default value from the language packs can then be imported into Joomla and
database and provides a space for localized values to be used. The translated values are stored based on the
be entered. The component is developed in a way to table structure that is created for each interface.
be easily installed and used in a friendly manner. The

Figure 2: Translation Component Main Menu Interface

This component allows a user to create a language KEY=”Value”
pack for a specific language selected. Each pack where “KEY” is a string to be translated and
includes ini files and localize files. The language “Value” is the translated string. For example if set of
package will be provided as an input to the content lines can be taken from the ini files,
management system and installed to apply a selected COM_BANNER=”Banner”
language. A translation pack consists of an xml file, a
ADDITIONAL_INFORMATION=”Additional
blank index file and a set of *.ini files containing the
Information”
language strings for a language. The easiest way to
The values contain strings that are displayed in
understand this structure is to look at the files in the
Joomla. The values can be changed to Unicode
language/en-GB folder of Joomla. The zip files
values and the modified values will be displayed in
contain xml files and set of ini files. Joomla language
Joomla.
INI files must be saved as UTF-8. The format of
COM_BANNER="ሰንደቅ"
language definition files is very basic. Blanks lines
and lines beginning with “#” are ignored and the ADDITIONAL_INFORMATION=”ተጨማሪ መረጃ”
latter may be used to add comments to the file. Each When Joomla reads the set of ini files, The
line consists of a pair of strings separated by an “KEY” that is the first word before the equal sign on
equals sign denoting key value pair represented in the every line must be in all capitals or Joomla will not
following pattern: be able to read the string for that particular item. The
62 Localization of Open Source Web Content Management System

“KEY” can include spaces and other punctuation A full language pack will contain translations for site
characters but there should not be any spaces in and administration parts. The translation files contain
either side of the equals sign as spaces are translation for components, plug-ins, modules and
significant. If more than one entry has the same left- templates where each of these files is attached with
hand side, the last one to be encountered is the one the interface and every functionality of Joomla. The
that will be used. The "VALUE" cannot include files are packed in to a zip file and can be provided to
double-quote characters ("). To get a double-quote Joomla as a language pack. Once the language pack
character, use the HTML special character sequence has been imported, a user can change the language
""" instead. Single-quote characters (') are valid that is required to be used for the overall operation
[7]. just by setting the language to Amharic. The
Localization for Joomla is maintained through translation of the every term is based on an ICT
basic steps to change the working environment into a glossary for Amharic, Oromigna and Tigrigna which
specified language selected by the end-user. Joomla was done by ICT Development Office and Ethiopian
uses three language packages, one for the frontend, Languages Research Center (ELRC) [8]. Figure 3,
one for the administrator and one for installation. shows the translation form that allows responsible
Ideally, all three must be created for a full package. persons to maintain the value of administrative units.

Figure 3: Translation Component Value Entry

Another part of localization for Joomla is the interface to text fields, password fields and text areas
concept of Virtual keyboard. Various JavaScript so that they can be filled with mouse only.
virtual keyboards have been created and are The script is developed for English language and
available, allowing users to enter their data in local other foreign languages to enable users enter text
languages. Converting these Javascript virtual using the virtual keyboard. The Java script interface
keyboards into Joomla module for making use of it is customized and embedded into a developed
on input fields is needed. A virtual keyboard module module to provide the necessary functionality for
is developed based on the layout created by Wyvern Joomla input fields. The customization flow for the
[9] which is a website providing Java script codes to emergence of the virtual keyboard is first to import
be customized for adding a graphical keyboard the script and style sheet that provide the
functionality and layout of the English virtual
HiLCoE Journal of Computer Science and Technology, Vol. 2, No. 2 63

keyboard including other foreign languages into html under the direction of technical committee for
file. The next step is to remove the support for other Information Technology, (QSAE/TC8), and
languages and add a support for Amharic to be published by the Quality and Standards Authority of
included in the language list that will be converted Ethiopia (QSAE) [10]. Figure 4 shows the virtual
into Joomla module so that it can be imported as a keyboard developed in this work and Figure 5 shows
package. a backend web page of the localized Joomla.
The layout is customized based on a Keyboard
standard for Ethiopia prepared by the working group

Figure 4: Amharic Virtual Keyboard Layout

Figure 5: Translated Interface

The outcome of this research is creating a
4. Discussion
language pack and virtual keyboard in addition lay
The findings of this research basically cover the down an approach based on the translation
development of a translation component that enables component to enable users to create language packs
language translation for several languages, a virtual for Amharic, Oromiffa and Tigrigna languages using
keyboard design which allows users to enter Amharic an easy to use interface which can be easily
characters and an Amharic language pack for Joomla embedded into the newer version of Joomla.
content management system.
64 Localization of Open Source Web Content Management System

5. Conclusion and Future Work [2] Gillian Klee, “Web Content Management and
SharePoint”, white paper, Waterstons, England,
In conclusion, the research covers areas of
2010.
localization for CMS. Different criteria are
[3] Central Statistical Agency, "Statistical Tables
considered for evaluating and selecting the top
for the 2007 Population and Housing Census of
performing CMS products. After evaluating the CMS
Ethiopia: Country Level", 2007, retrieved from
products, selected users are required to fill a
http://www.csa.gov.et/index.php, last accessed
questionnaire that assesses experience, popularity of
on Sept 12, 2013.
CMS in the local context. The questionnaires are
[4] Nepa Linux, Guide to Localization of Open
analysed and the CMS with the highest percentage,
Source Software, Centre for Research in Urdu
Joomla, is selected to be localized.
Language Processing (CRULP), National
As part of the localization, a language pack for
University of Computer and Emerging Sciences
the CMS was created to convert the frontend and
(NUCES), Pakistan, 2006.
backend interface of Joomla. The language pack is
[5] Zemen Adego, “Open Source Content
designed in a way to be imported into Joomla and
Management Software Customization for the
users can easily change the language. In addition,
Ethiopian Government”, Unpublished MSc
virtual keyboard module is designed for Joomla to
Thesis, Addis Ababa University, 2004.
enable users to encode Amharic text. The keyboard
[6] Lielet Getachew, “Localization of Open Source
can be imported into Joomla and users can input
Content Management Software into Amharic
characters on input fields.
Language”, Unpublished MSc Thesis, Addis
A Translation Component that enables users to
Ababa University, 2011.
translate to Ethiopic languages for Joomla CMS was
[7] “Making a language pack for Joomla” retrieved
developed. The translation component enables users
from http://docs.joomla.org/J3.1:Making_a_
to create Joomla language pack for Amharic,
Language_Pack_for_Joomla#How_to_create_lo
Oromiffa and Tigrigna. In order to validate and
calized_CORE_language_packs, last accessed
assess the usability of the developed approach, the
on Jan 13, 2014.
localized Joomla is provided to four selected users
[8] ICT Development Office and Ethiopian
for the purpose of Website development. The
Languages Research Center (ELRC),
selected users are able to develop the website with
“Translation of ICT Terms into three major
relative ease.
Ethiopian Languages Amharic, Oromo and
In the future, we plan to extend the proposed Tigrigna” retrieved from
approach to localize it to all Ethiopic languages http://www.scribd.com/doc/116396965/ICT-
based on users’ preference. The work focused on Glossary-Amharic-Tigrigna-Oromifa-
specific sets of languages and filling the gaps in Languages, last accessed on Jan 02, 2014.
between the strings available in a file structure and [9] Brian Huisman, “Java Script Graphical/Virtual
the interface displayed. The idea can be expanded for Keyboard Interface”, retrieved from
full support of other Ethiopic languages. http://www.greywyvern.com/code/javascript/ke
References yboard, last accessed on Mar 28, 2014.
[10] Technical Committee for Information
[1] St. Laurent and Andrew M, Understanding
Technology, Ethiopic Keyboard Layout 1st
Open Source and Free Software Licensing,
edition, 2007, Quality and Standards Authority
O'Reilly Media, 2008.
of Ethiopia (QSAE).