aboutsummaryrefslogtreecommitdiff
path: root/documentation/deployment.rst
diff options
context:
space:
mode:
Diffstat (limited to 'documentation/deployment.rst')
-rw-r--r--documentation/deployment.rst20
1 files changed, 10 insertions, 10 deletions
diff --git a/documentation/deployment.rst b/documentation/deployment.rst
index 95f3bc23..d0639c45 100644
--- a/documentation/deployment.rst
+++ b/documentation/deployment.rst
@@ -10,7 +10,7 @@ Deploying the application with Vagrant
Installation of VirtualBox and Vagrant
-.. code:: bash
+.. code-block:: bash
$ sudo apt-get install virtualbox
$ wget https://dl.bintray.com/mitchellh/vagrant/vagrant_1.6.5_x86_64.deb
@@ -18,7 +18,7 @@ Installation of VirtualBox and Vagrant
Deployment of pyAggr3g470r
-.. code:: bash
+.. code-block:: bash
$ git clone https://bitbucket.org/cedricbonhomme/pyaggr3g470r.git
$ cd pyaggr3g470r/vagrant/
@@ -34,7 +34,7 @@ An instance of pyAggr3g470r is running `here <https://pyaggr3g470r.herokuapp.com
The geek way
''''''''''''
-.. code:: bash
+.. code-block:: bash
$ git clone https://bitbucket.org/cedricbonhomme/pyaggr3g470r.git
$ cd pyaggr3g470r
@@ -47,7 +47,7 @@ The geek way
To enable account creation for users, you have to set some environment variables:
-.. code:: bash
+.. code-block:: bash
$ heroku config:set SELF_REGISTRATION=1
$ heroku config:set PLATFORM_URL=<URL-of-your-platform>
@@ -77,7 +77,7 @@ And some other optional environment variables, as previously presented.
Deploying the application on a traditional server
-------------------------------------------------
-.. code:: bash
+.. code-block:: bash
$ sudo apt-get install python libpq-dev python-dev python-pip build-essential git
$ sudo apt-get install libxml2-dev libxslt1-dev # for lxml
@@ -89,7 +89,7 @@ Deploying the application on a traditional server
If you want to use PostgreSQL
'''''''''''''''''''''''''''''
-.. code:: bash
+.. code-block:: bash
$ sudo apt-get install postgresql postgresql-server-dev-9.3 postgresql-client
$ echo "127.0.0.1:5432:aggregator:pgsqluser:pgsqlpwd" > ~/.pgpass
@@ -101,7 +101,7 @@ If you want to use PostgreSQL
Edit the configuration file with the line:
-.. code:: cfg
+.. code-block:: cfg
[database]
uri = postgres://pgsqluser:pgsqlpwd@127.0.0.1:5433/aggregator
@@ -111,7 +111,7 @@ If you want to use SQLite
Just edit the configuration file with the line:
-.. code:: cfg
+.. code-block:: cfg
[database]
uri = sqlite+pysqlite:///pyAggr3g470r.db
@@ -119,7 +119,7 @@ Just edit the configuration file with the line:
Finally:
-.. code:: bash
+.. code-block:: bash
$ python db_create.py
$ python runserver.py
@@ -142,7 +142,7 @@ Automatic updates
You can fetch new articles with `cron <https://en.wikipedia.org/wiki/Cron>`_ and the script *fetch.py*.
For example if you want to check for updates every 30 minutes, add this line to your cron rules (*crontab -e*):
-.. code:: bash
+.. code-block:: bash
*/30 * * * * cd ~/.pyaggr3g470r/ ; python fetch.py firstname.lastname@mail.com
bgstack15