dainst / ariadne-portal Goto Github PK

License: MIT License

PHP 39.69% JavaScript 17.20% ApacheConf 0.11% CSS 5.94% HTML 37.07%

ariadne-portal's Introduction

#ARIADNE Portal

The ARIADNE Portal is a web application bases on Laravel. It's main purpose is to offer access to the ARIADNE catalog data provided through Elasticsearch.

##Setup for development

###Install composer https://getcomposer.org Follow install instructions for your operating system

###Setup Clone this repo from GitHub Create local config file Make a copy of .env.example and name it .env Edit .env in a text editor, update info about elastic search etc.

###Install vendor libraries via composer navigate to the root folder of the project (where composer.json is located) run:

composer install

Libraries used by the portal will now be downloaded, this could take a while The libraries will be downloaded into the directory called “vendor”, this directory is ignored in the file .gitignore If you have project files from your IDE in the same folder as the source code, add these to .gitignore

###Compile JS and CSS files

Gulp is used to compile SCSS into CSS files and to combine and minify all JavaScript files. Before deployment gulp has to be run:

npm install
bower install
gulp

For Windows:

Download node.js https://nodejs.org/en/download/
Download python https://www.python.org/getit/windows/
Run npm install -g node-gyp
Run npm install
Run npm install -g bower
Run npm install -g gulp

###Run the portal during development To run the portal in PHP:s built in webserver:

php artisan serve

The portal will now be accessible via http://localhost:8000

To automatically recompile js and css files after changes

gulp watch

###Development setup with apache

Easiest way to do development is to create a virtual host (vhost). Example config:

<VirtualHost ariadne.laravel.localhost:80>
    ServerAdmin [email protected]
    DocumentRoot "C:/Users/karl/ariadne-portal/public/"
	
	<Directory "C:/Users/karl/ariadne-portal/public/">
        Options Indexes FollowSymLinks Includes ExecCGI 
        Order allow,deny  
        Allow from all  
        AllowOverride All 
        AddDefaultCharset UTF-8
	</Directory>
	
    ServerName ariadne.laravel.localhost
    ErrorLog "logs/ariadne.laravel.localhost-error.log"
    CustomLog "logs/ariadne.laravel.localhost-access.log" common
</VirtualHost>

Observe the path of the directory, DocumentRoot should be the folder named public inside the project folder. Place this file in apaches vhost directory, be sure the line for loading the module vhost_alias_module and “Include conf/extra/httpd-vhosts.conf” is uncommented.

Next add an entry to your host file: (on windows C:\Windows\System32\Drivers\etc\hosts, mac: Open /Applications/Utilities/NetInfo Manager, linux: sudo nano /etc/hosts) ariadne.laravel.localhost 127.0.0.1

The portal should now be hosted via http://ariadne.laravel.localhost

###Distribution

In order to create a distribution package run:

gulp dist

##Deployment

The tar.gz package built with gulp dist - which can also be downloaded here - contains all files needed to run the application on a web server.

The contents of this package can be extracted and made available through any WebServer that supports PHP.

After extracting make sure that the variables in .env are set up for the particular environment.

Example configuration for Apache

    <VirtualHost *:80>
            ServerName portal.ariadne-infrastructure.eu
            ServerAlias ariadne-portal.dcu.gr
            DocumentRoot /var/www/ariadne-portal/public
            ServerAdmin webmaster@localhost
            Alias / "/var/www/ariadne-portal/public/"
            <Directory /var/www/ariadne-portal/public>
                    Options Indexes FollowSymLinks MultiViews
                    AllowOverride All
                    Order allow,deny
                    allow from all
                    AddDefaultCharset UTF-8
            </Directory>
            ErrorLog ${APACHE_LOG_DIR}/ariadne-portal_error.log
            LogLevel warn
            CustomLog ${APACHE_LOG_DIR}/access.log combined
    </VirtualHost>

ariadne-portal's People

Contributors

Watchers

Forkers

eighttrigrams

ariadne-portal's Issues

Query construction fails: It isn't possible to run a search with more than one filter.

Welcome page: Make boxes responsive

On my screen the images in the "browse" box stack. They should scale with the resolution.

Show link to children on a resource landing page

Do a query to a resource landing page for isPartOf (resources who is children of the current resource).
Show the number of resources and link to search results eg:

children (42)

Search bar: Dropdown list for field

As depicted in the mockups (v4) the search bar should feature a dropdown list for selecting the field the user wants to query.

Create a robots.txt

Until public announcement no page of the portal should be indexed by search engines.

Refined ranking

The ranking of search results should be refined.

One particular goal is to prevent resources from data providers with fewer items to get buried underneath other data.

Also it could make sense to boost certain fields, for example to rank resources higher whose titles match the query.

Use date_optional_time format for period and issued aggregations

Subject landing page

A separate view for displaying details about a single subject should be implemented.

Cool URIs & content negotiation

The portal should be able to resolve the URIs defined in the Gothenburg meeting.

It was decided that the following schema should be implemented:

http://registry.ariadne-infrastructure.eu/resource/{id}
http://registry.ariadne-infrastructure.eu/page/{id} (html)
http://registry.ariadne-infrastructure.eu/data/{id} (xml)
http://registry.ariadne-infrastructure.eu/resource/{id}.json

http://registry.ariadne-infrastructure.eu/provider/resource/{id} 
http://registry.ariadne-infrastructure.eu/provider/page/{id}
http://registry.ariadne-infrastructure.eu/provider/data/{id}
http://registry.ariadne-infrastructure.eu/provider/resource/{id}.json

http://registry.ariadne-infrastructure.eu/subject/resource/{id}
http://registry.ariadne-infrastructure.eu/subject/page/{id}
http://registry.ariadne-infrastructure.eu/subject/data/{id}
http://registry.ariadne-infrastructure.eu/subject/resource/{id}.json

Resource landing page: Shorten long descriptions

Long descriptions should be shortened to a few lines and be expandable with a "more" button.

Browse View: Timeline (When)

The timeline browse view should offer a way of navigating a search result or the whole catalog similar to the map browse view. It should display the distribution of resources in time and allow selecting a date range for starting or refining a search.

Similarly to the map view it should make use of aggregations that get refined when the user "zooms in".

Leaflet license infos are covered by footer

Hierarchical query expansion based on AAT subjects

Elasticsearch could be configured as described in Ceri Bindings paper "Implementing hierarchical semantic expansion in ARIADNE".

Another possibility would be enrich the data with ancestor concepts before ingest into elasticsearch in MoRe.

Search result filters: Map

A map that displays the distribution of the current search results and allows filtering based on a bounding box.

Trim search hits' meta description

Browse view: Subjects (What)

There should be a view that visualizes the distribution of different subjects in the catalog or a particular search result.

This view should include the AAT subjects, the languages and the data providers.

More flexible view for browsing similar resources

In addition to showing the context of the current in the sidebar on the resource landing page a separate full-page view could be implemented to allow fine-grained filtering of similar resources.

Search result filters: Timeline

A timeline that displays the distribution of the current search results and allows filtering based on a date range.

Auto-suggestions based on AAT subject terms

See page 2 in the mockups.

SEO: Include meta tags

Better page titles

The portal should display more meaningful page titles in order to improve search engine visibility and to allow a better user experience when navigating the browser history.

Navigation: "Catalog" should point to welcome page

Since we will have links to the different browse sections on the welcome page the link "Catalog" should point to the welcome page.

The Ariadne logo in the navigation bar should point to http://www.ariadne-infrastructure.eu/ .

Map on resource landing page contains marker for source resource

Auf Basis der location der angezeigten Resource wird eine geo distance gefilterte Query ausgeführt,
die die umliegenden Objekte aufliststet. Unter diesen befindet sich auch die location der originalen Resource, was derzeit nicht auffällt, da die originale
Resource später gezeichnet und Orange markiert wird.

IP ranges for elasticsearch

The access to the elasticsearch API should be restricted to localhost and certain IP address ranges.

If we find the time to set up local development instances of elasticsearch the access can be restricted to just localhost.

Resource landing page: Link to export

The link "export" on the current page currently has no functionality. It should point to a JSON representation of the resource.

Resource landing page: Link for citing the resource

The "cite" link should open a box that allows copying the URI into the clipboard.

Search result page: Display active filters on top

Hide types, make resource types more prominent

As discussed at the meeting in Vienna resource types should be the prominent category in the portal interface.

This means that the resource type should be the topmost facet and should also be the basis for the icons in the search result view.

The current types are technical details that should be hidden from the user.

Simplify distribution and deployment

Installing all the build tools (like npm, gulp, ...) on servers should be avoided and manually copying the files necessary for running the application can be a tedious task. Therefore a gulp task should be implemented that collects just the necessary files and packages them in an archive that can be copied and extracted for deployment.