bigquery

bigquery

This is a simple sdk for using google bigquery. Still working... if any problem, please let me know. Want to know what is BigQuery? Look my slide: http://www.slideshare.net/peihsinsu/google-bigquery-introduction

Installation

npm install bigquery

Apply service account

Follow the doc: http://gappsnews.blogspot.tw/2013/10/connect-cloud-platform-bigquery-using.html

Convert p12 key (If use old version service account...)

From admin console, create a service account, save the client_secrets.json and it's key ex: Translate p12 to pem

openssl pkcs12 -in privatekey.p12 -out privatekey.pem -nocerts
openssl rsa -in privatekey.pem -out key.pem

or

openssl pkcs12 -in privatekey.p12 -nodes -nocerts > key.pem

Initial using Service Account json_file

If you are create the service account after 2015Q2, you will find the service account provide a json file formate download. The json file is like:

{
  "private_key_id": "e7************************************e8",
  "private_key": "-----BEGIN PRIVATE KEY-----\nMIIC***....skip...****AFW1Y\n-----END PRIVATE KEY-----\n",
  "client_email": "86*********8-if***************************[e-mail]",
  "client_id": "86*********8-if***************************pq2.apps.googleusercontent.com",
  "type": "service_account"
}

The file is already include the secret file and the account information. We support to use the file for easy auth.

var bq = require('bigquery')

bq.init({
  json_file: '/path/to/your-service-account-json-file.json'
});

Initial using client_email directly

About 2015Q1 end, google start to deprecate the client_secret file download. You can change the init method like this:

bq.init({
  client_email: '[e-mail]',
  key_pem: '/path-to-key.pem'
});

Initial using old client_secret.json (will deprecated)

Load bigquery lib, specify your project id then setup the service account and the client_secret.json file path, pem key file path for auth use.

var bq = require('bigquery')
  , fs = require('fs')
  , prjId = 'your-bigquery-project-id'; //you need to modify this

bq.init({
  client_secret: '/path-to-client_secret.json',
  key_pem: '/path-to-key.pem'
});

Do Query

Query the dataset that you have. The sample bellow is to query the public dataset of wikipedia.

bq.job.query(prjId, 'select count(*) from publicdata:samples.wikipedia', function(e,r,d){
  if(e) console.log(e);
  console.log(JSON.stringify(d));
});

Get Query Results

Retrieve results from a previously run query. Larger queries will timeout and respond with no data before the results are collected. Use this to retrieve the results. Below is an example of querying and waiting for results.

bq.job.query(prjId, 'select count(*) from publicdata:samples.wikipedia', function(e,r,d){
  if(e) console.log(e);
  if(d.jobIsComplete){
      console.log(JSON.stringify(d));
  } else {
       bq.job.getQueryResults(prjId, d.jobReference.jobId, function(e,r,d){
          console.log(JSON.stringify(d));
       }
  }
});

Load data

Before load data, you must create your dataset and table first. After that, you can use the following code to load data. (In this case test is the dataset id, testtb1 is the table name.)

var data = [
 {
   "insertId": "201403221228", //option
   "json": {
     "name": "simon",
     "sex": "M",
     "age": 35
   }
 }
];

bq.job.load(prjId, 'test', 'testtb1', data, function(e,r,d){
  if(e) console.log(e);
  console.log(JSON.stringify(d));
})

PS: The insertId use for prevent the duplicate insert of data.

Create Dataset & Table

Create dataset and table, the schema please ref: https://developers.google.com/bigquery/preparing-data-for-bigquery

bq.dataset.create(prjId, 'dataset_name', function(e,r,d){
  if(e) console.log(e);
  console.log(d);
});

var schema = {
  "fields": [
   {
    "name": "field1",
    "type": "string",
    "description": "test"
   },
   {
    "name": "field2",
    "type": "integer",
    "description": "test for int"
   }
  ]
 };
bq.table.create(prjId, 'dataset_name', 'table_name', schema, function(e,r,d){
  if(e) console.log(e);
  console.log(d);
});

Get Table

bq.table.get(prjId, 'test123', 'table_name', function(e,r,d){
  if(e) console.log(e);
  console.log(d);
});

Delete Table

bq.table.delete(prjId, 'dataset_name', 'table_name', function(e,r,d){
  if(e) console.log(e);
  console.log(d);
});

Get Dataset

bq.dataset.get(prjId, 'dataset_name', function(e,r,d){
  if(e) console.log(e);
  console.log(d);
});

Delete Dataset

bq.dataset.delete(prjId, 'dataset_name', 'table_name2', function(e,r,d){
  if(e) console.log(e);
  console.log(d);
});

Other request timeout parameters

If you want to add timeout parameter to restrict to your bigquery api request time. You can add timeout parameter to the init() like:

bq.init({
  client_secret: __dirname + '/client_secret.json',
  key_pem: __dirname + '/key.pem', 
  timeout:1000
});