Talk Title - LabKey Software

Extending
LabKey Server Part 2:

Creating Applications

Adam Rauch

Partner, LabKey
Software

adam@labkey.com

Key extension mechanisms:



Custom queries


External schemas / data sources


Script reports


Server API / client libraries


File
-
based modules


Java modules






Retrieve & present data





Create applications


Extension Mechanisms

2

Terminology


Server API


An HTTP/JSON
-
based web services API


Provides direct, programmatic access to LabKey data and services from
most programming languages


Client Libraries


Provide Server API access from many popular languages: JavaScript,
Java, R, SAS, Perl, Python


Sometimes referred to as the “Client API”

3

SQL Server
Database

LabKey Database

Web Site

Query

Authentication

and Security

Data
Visualization

Study
Management

Collaboration

Assays

Flow
Cytometry

Proteomics

Server API

Client API

MySQL

Database

HTTP/HTML

HTTP/JSON

SAS Repository

4

LabKey API


Develop scripts and standalone applications that interact with
LabKey via the HTTP server API


Applications send and receive data over HTTP


Subject to security checks just like browser requests


Access the LabKey server API from any language that can handle
HTTP requests and JSON


Client libraries developed for popular languages:


JavaScript


R


Java


SAS


Perl


Python


LabKey API is stable and well documented

5

Authentication Using Credentials File


Standard CURL mechanism that keeps user names and
passwords (relatively) secure, out of code


R, Perl, and SAS libraries use
netrc

file for credentials


.
netrc

(Linux/Unix) or
_
netrc

(Windows) file in home directory


Python library uses file named
labkeycredentials.txt


Contents determine credentials to send with each request


Example
netrc

contents:


machine labkey.org

login adam@labkey.com

password
mypassword

6

JavaScript API



Most extensive client library; all other libraries are subsets,
patterned after JavaScript API


Designed for building interactive, data
-
driven web pages


Built on Ext JS application framework


Cross
-
browser JavaScript framework for rich web apps


Provides many UI widgets used throughout the product


LabKey API is free, but developing closed
-
source code to the Ext JS API
requires a
Sencha

license


JavaScript API understands current context:


Current folder


Currently logged in user


Server properties


Primarily an
asynchronous

API; uses callbacks extensively

7

JavaScript API Highlights


Query


Retrieve and modify data from any table or query (
selectRows
,
executeSql
,
insertRows
,
updateRows
,
deleteRows
)


Retrieve meta data about queries, views, schemas


Security


Retrieve information about users, groups, permissions


WebPart
: Render any
webpart

into a page


Ext.store
: Ext JS integration with UI elements


ActionURL
: Build standard LabKey URLs


Pipeline: Retrieve job status and start analyses


Utils
: Other useful methods


And much, much more…

8

JavaScript API Tips & Tricks


Results format: JavaScript objects and arrays


Authentication: automatically handled by browser


Documentation: LabKey Admin/Developer menu




Think
async
!


JavaScript debugging


Firebug


Chrome JavaScript Console


Use “Export to JavaScript” from any data grid


Wiki pages: great way to test and deploy code

9

R & Perl Client Libraries


Rlabkey

R Package


Functions: Query, Assay, context information


Results format: R data frames


Authentication:
netrc

file


Distribution: CRAN (standard R distribution network)


Documentation: CRAN, vignette included in package


Tip: Use “Export to R script” from any data grid


LabKey Perl Module


Functions: Query


Results format: hash reference


Authentication:
netrc

file


Distribution: CPAN (standard Perl distribution network)


Documentation: CPAN

10

Java & SAS Client Libraries


Java Client Library


Functions: Query, Assay, Security


Results format: Java maps and lists


Authentication: Credentials passed into methods


Distribution: LabKey download page


Documentation:
JavaDoc

in client library zip file


SAS Client Library


Functions: Query


Results format: SAS data sets


Authentication:
netrc

(or passed into macros)


Distribution: LabKey download page


Documentation: labkey.org


Tip: Use “Export to SAS script” from any data grid

11

Python Client Library


Brand new… so new, I haven’t tried it yet!


Functions: Query


Results format: Python dictionary


Authentication: labkeycredentials.txt file


Distribution: LabKey download page


Documentation: labkey.org

12

13

File
-
Based (aka “Simple”) Modules


Convenient mechanism for packaging and distributing code


Great for collaborative development (source code control)
and multi
-
server deployments (test, staging, production)


Can include:


HTML/JavaScript
pages and static resources


SQL scripts, custom queries, custom views,
s
chema
meta data


Assay definitions


Validation and transformation scripts


Reports


Folder types


No Java, no compile


Easy
dev



test


staging


production cycle


Tip: labkey.org tutorial and “Map
of Module
Content” topic

14

Java Modules


Can also extend LabKey via custom modules written in Java


Nearly all core modules are Java modules


Use MVC model based on Spring


Leverage massive internal API and large set
of third
-
party libraries


Pros


Most flexible and powerful way to extend LabKey


Cons


Require strong Java skills


Difficult to maintain when outside of the core team


Java server API is minimally documented and less stable


Your module(s) will likely require editing and recompiling with each new
LabKey release


Tip:
create_module

ant task builds a new module

15

Any questions?

Adam Rauch

adam@labkey.com