CSA

Crowdsourcing System for Association Data

URL: https://csa.cbdb.fas.harvard.edu

Supports:

• Suggest new value or revise any current field value for CBDB
• Workflow to review / adopt crowd suggestoins
• Keyword based auto-suggest per CBDB data (currently only person field)
• Import new task from a CSV file
• Export to CSV / TSV (to be merged to CBDB)

Running the app

Option 1: Docker (recommended)

Pull and run the pre-built container:

docker pull quay.io/oopus/csa:latest

Run with a Docker named volume:

docker run -d \
  -v data:/usr/src/cbdbapp/data \
  -p 3000:3000 \
  -p 5001:5001 \
  -e CI=true \
  --restart=always \
  --name csa \
  quay.io/oopus/csa:latest

Or bind-mount a local data/ directory:

docker run -d \
  -v $(pwd)/data:/usr/src/cbdbapp/data \
  -p 3000:3000 \
  -p 5001:5001 \
  -e CI=true \
  --restart=always \
  --name csa \
  quay.io/oopus/csa:latest

The data/ directory must contain the required database files (cbdb.db, tasks.db, users.db). Contact hongsuwang#fas.harvard.edu to obtain them.

• Frontend: http://localhost:3000
• Backend API: http://localhost:5001

To stop / restart:

docker stop csa
docker start csa

To rebuild from source:

docker build -t quay.io/oopus/csa:latest .

Option 2: Build from GitHub

Build and run directly from the GitHub repository (no need to clone):

docker build -t csa https://github.com/cbdb-project/crowdsource-webapp.git

docker run -d \
  -v $(pwd)/data:/usr/src/cbdbapp/data \
  -p 3000:3000 \
  -p 5001:5001 \
  -e CI=true \
  --restart=always \
  --name csa \
  csa

Replace /path/to/data with the absolute path to your data/ directory.

Option 3: Local development

1. Replace ./src/config.js with ./src/config_local.js

2. Install dependencies:

    npm install
   

3. Launch:

    node src/server.js &
    npm start
   

Production deployment (with TLS)

In production, use a reverse proxy (e.g. Caddy, Nginx) for TLS termination:

Frontend(port=3000) >>> Reverse Proxy(TLS, port=443/5000) >>> Backend(port=5001)

File / Directory structure

• src: source JS / CSS files.
• data: database files (cbdb.db, tasks.db, users.db) required for the app to function.
• Dockerfile: build your own Docker container.
• public: public HTML files.

Task CSV Format

First line: column specification (mandatory)

The first line of the CSV should be (column_name) or (column_name=column_type), where:

• column_type should be one of the following: blank (which defaults to "string"), "key" (i.e. primary key), "person" (a cbdb person), "string" (any arbitrary string), "int" (integer)
• One (and only one) of the columns should be of "key" type

Here's a valid example:

line=key,writer,sequence,person_id,title,assoc_type=string,assoc_name_chn=string,assoc_personid=person

And an invalid one:

line=key,writer=key,sequence,person_id,title,assoc_type=,assoc_name_chn=string,assoc_personid=person

Problems: multiple primary keys ("line" and "writer"), and empty column type ("assoc_type")

Second line: user friendly column name (mandatory)

Those are the column names presented to end user.

Examples

  line=key,writer,sequence,person_id,title,assoc_type=string,assoc_name_chn=string,assoc_personid=person,assoc_count=int,assoc_year=int,nianhao=string,year=int,range=string,collection,c_source,volume
  行号,作者,序列,人物id,标题,关联类型,关联类型描述,关联人物,关联次数,关联年份,年号,公元年,范围,文集,出处,卷
  51,羅倫,1,34531,在告與三閤老,致書Y,,,,,,,,一峰先生文集,7204,8
  52,羅倫,2,34531,在告與崔冡宰,致書Y,,,,,,,,一峰先生文集,7204,8
  53,羅倫,3,34531,與李賔之,致書Y,李東陽,28691,,,,,,一峰先生文集,7204,8
  54,羅倫,4,34531,與劉用光,致書Y,劉煊,199087,,,,,,一峰先生文集,7204,8
  55,羅倫,5,34531,復胡提學書,答Y書,,,,,,,,一峰先生文集,7204,8

Technical Contributors

Lei DING, Hongsu WANG, Zihan ZHOU, Jimmy WEI, Xuan ZHANG

Academic Advisors/Contributors

Peter BOL, Song CHEN, Kwok-leong TANG, Katherine ENRIGHT

License

Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International license