No more translation integration

If you don’t have inline translation, you will have to extract all the wordings in a file yourself, send it over to every translators, then integrate yourself all their translation. It is tedious, error prone and not agile at all: next release you will have to do the work again and again. Even for a small app, we are talking about hours of work.

With inline translation, like in Magento or Transl8, you let your users translate the text by browsing the app. No more file integration.

Ready to go live anytime

Translators can then translate step by step and even if the work is only 90% finished, you already have an integrated translation ready to be used.

Reduce the translation mistakes

Another bonus effect with online translation the text is presented in context –  by in context I mean be presented on the screen that pops up this message. You also reduce the errors of translation induced.

Reactive correction

You can change in production within seconds a mispelling in a language.

Check Transl8 for a solution of inline translation with Zend Framework.

Prepare the SQL datamart

Let’s say you have a social network. We want to track down the number of clicks on the “like” button, and be able to count by the type of media liked (image, video) and some characteristics of the user that liked the link (sex, age).
The model should look like:

CREATE TABLE tracking
(
"key" bigserial NOT NULL,
user_id integer,
media_type integer,
count_like_click integer
);

CREATE TABLE user
(
id bigserial NOT NULL,
name VARCHAR(50)
);

CREATE TABLE media_type
(
id bigserial NOT NULL,
name VARCHAR(50)
);

Virtual url in Google Analytics

Write this snippet of Javascript. Here is a sample corresponding to what we built:

<script type="text/javascript">
function like_click_listener(){
user_id = "42";
media_type = "1";
urchinTracker("/share_click/" + user_id + "/" + media_type);
}
</script>

Bring Urchin data to PHP Analytics

All you have to do is to query the API: https://secure.urchin.com/helpwiki/en/Data_API.html

The url follows this template:

http://server.com/api/urchin/v1/ReportService/data?login=$login&password=$password&ids=$PROFILE_ID&start-index=$start-index&max-results=$max-results&start-date=$start-date&end-date=$end-date&dimensions=u:request_stems&metrics=u:pages&table=1

Notice we query table 1, dimension u:request_stems and metrics u:pages. Then you can use your favorite tool to load this webservice data to the PHP Analytics database. It can be Talend Open Studio, Kettle or any ETL. We prefer a home made PHP script for such a simple example.

The metamodel

Now we will put analytical info over the database: there we turn our standard SQL into a full R-OLAP cube.

<?xml version="1.0"?>
<Schema name="MyWebsite">
<Cube name="Mega track">
<Table name="tracking" />
<Dimension name="User" foreignKey="user_id">
<Hierarchy hasAll="true" primaryKey="id">
<Table name="user" />
<Level name="User Name" column="name" type="String" uniqueMembers="true" />
</Hierarchy>
</Dimension>
<Dimension name="Media type" foreignKey="media_type">
<Hierarchy hasAll="true" primaryKey="id">
<Table name="media_type" />
<Level name="Media type" column="name" type="String" uniqueMembers="true" />
</Hierarchy>
</Dimension>
<Measure name="Like Clicks" dimension="Measures" column="count_like_click" aggregator="sum" formatString="#,##0" />
</Cube>
</Schema>

Now you can easily enrich your report with referential data, add a timeline …

Happy hacking!

Inovia releases Transl8, a Zend Framework-based library that will make internationalization of your Web applications very easy. Because managing translations files or tables is often a tedious process, Transl8 allows you to translate in-page your web applications, relying on the powerful Zend_Translate to store your data.

Discover Transl8, the aspirin for your internationalization headaches!

PHP Analytics packages tools such as pivots and cross tables out of the box. But besides that, you will also be able to build your own custom MDX queries for advances indicators.

Discover PHP Analytics, the first open-source and Zend-based library for analytics.