fluffy-engine

A fluffy engine for XDU 2017 fall SS SPM final exam.

Install

For cabal:

cabal update
cabal install fluffy
cabal install fluffy-parser

Usage

Get Quick Start

You can run the following command to deploy it directly, but you might need read Env and Dependence first:

curl -sSL http://raw.githubusercontent.com/Qinka/fluffy-engine/master/scripts/simple-run.sh | sh

Now you can go to have a tea, and take a rest, because that will take a lots of time. If you setting back-end port to 3000, you can then visit http://localhost:3000 on your browser.

Dependence

You need install:

• Docker

• md5sum

• curl

• wget

• bash or bash-like shell

Env

FLUFFY_SPM_PATH

When you try to run that script, you need set environment variables. One you must set is FLUFFY_SPM_PATH. This env var sets up the path of question banks of course(SPM). For example, if your bank located at /home/qinka/2017spm/XDExamReview2017, or say one file of bank is /home/qinka/2017spm/XDExamReview2017/TrueFalse/Chapter_2_The_Project_Management_and_Information_Technology_Context.docx, you should set FLUFFY_SPM_PATH up as /home/qinka/2017spm/XDExamReview2017/TrueFalse.

FLUFFY_SKIP_CHECK_MD5

Moreover, the script will check the MD5 of those files to make sure that process of parsing will work. If you want to skip this step, you need to set environment variable FLUFFY_SKIP_CHECK_MD5 up as any non-empty string.

FLUFFY_DB_PASSWORD

This env var will set up your database's password, and the default is fluffypassword

FLUFFY_PREFIX

This env var will set up a prefix for each container's name, and the default is fluffy

FLUFFY_DB_PORT_E

This env var will set up a export port for database, and the default is "not export", or say empty.

FLUFFY_PORT

This env var will set up the port of back-end, and 3000 is default.

Tips

You might need to use a docker mirror and hackage mirror site. For the former, you need search it on the Internet. For the latter, you need edit the cabal configure file in the container to change to mirror site, (you can search it on the Internet).

Run

First step is pull the images.

The images include:

• qinka/fluffy:fluffy-latest
• postgtrs:latest
• haskell:latest

You can pull them from Docker Hub or other mirror sites with commoands:

docker pull qinka/fluffy:fluffy-latest
docker pull postgres:latest
docker pull haskell:latest

The image fluffy:fluffy-latest is the back-end. It need a PostgreSQL database(image postgres). The image haskell is used to run the scripts to upload data from *.docxs.

Second step is run the images.

You can run it via docker-compose, or just run the command one-by-one. The key idea is fluffy need to be linked with database.

The following is the example running with shell command one-by-one.

Create and run the PostgreSQL image: postgres:latest :

docker run -d --name fluffy-db -e POSTGRES_PASSWORD=fluffypassword postgres:latest

The detail can be found on Docker Hub - postgres.

Create and run the Fluffy image: qinka/fluffy:fluffy-latest:

docker run -d --name fluffy-be --link fluffy-db:db -p 3000:3000 fluffy:fluffy-latest -c "fluffy 3000 host=db port=5432 user=postgres password=fluffypassword"

Then you can visit host-of-docker:3000, and the home page will be shown.

(optional for haskell image)

To update the data to database with script, you also need a container with haskell's image. So you can run:

docker run -d --name fluffy-haskell --link fluffy-db:db haskell:latest

Third step is import the data into database with script.

You need firstly initialize the sql table to PostgreSQL database. Copy the sql file to postgres image:

wget https://raw.githubusercontent.com/Qinka/fluffy-engine/master/database/fluffy.sql 
docker cp fluffy.sql fluffy-db:/fluffy.sql
rm fluffy.sql

After sql file copied, run the psql to initialize database with sql script:

docker exec fluffy-db psql -d postgres -U postgres -f /fluffy.sql

Till now, the database is initialized. Next thing to do is import the data into database with the script. Copy the script into database:

wget -sSL https://raw.githubusercontent.com/Qinka/fluffy-engine/master/scripts/update.sh
docker cp - fluffy-haskell:/update.hs
rm update.hs
docker exec fluffy-haskell chmod a+x /update.hs

Copy the SPM's question bank into container fluffy-haskell:

docker cp /path/of/spm/s/question/bank fluffy-haskell:/

Than run the script:

docker cp /path/to/question/bank/Fillintheblanks fluffy-haskell:/
docker cp /path/to/question/bank/MultipleChoice  fluffy-haskell:/
docker cp /path/to/question/bank/TrueFalse       fluffy-haskell:/
docker cp /path/to/question/bank/WritingPart     fluffy-haskell:/
docker exec fluffy-haskell git clone https://github.com/Qinka/fluffy-engine.git
docker exec fluffy-haskell apt update
docker exec fluffy-haskell apt install -y libpq5 libpg-dev
docker exec fluffy-haskell cabal update
docker exec fluffy-haskell cabal install --only-dependencies fluffy-engine/fluffy-parser
docker exec fluffy-haskell cabal install postgresql-simple pandoc bytestring monad-logger fast-logger text
docker exec fluffy-haskell cabal install fluffy-engine/fluffy-parser
docker exec fluffy-haskell runhaskell /update.hs / host=db port=5432 user=postgres password=fluffypassword

Final

Then you can visit the site to check the data.

Customization

The fluffy has a way to customize: there is a javascript file, named prelude.js, which will be loaded. You can change this javascript file to customize this site. The all the static file can be placed or added into the data directory of fluffy(in docker image is /usr/share/fluffy). So when customizing fluffy, you can add the files into data directory of fluffy and those file can be found under route /static.

With Dockerfile

You can add the file you want to add into fluffy via creating a new docker with dockerfile, for example:

FROM qinka/fluffy:fluffy-latest
MAINTAINER Johann Lee <[email protected]>
ADD /your/host/files /usr/share/fluffy

With docker cp

You can also copy the files by commoand:

docker cp /path/to/your/files fluffy:/usr/share/fluffy