Tableau Conference Sample Application

This is sample code, and is not supported by Tableau. Please be kind and don't call Technical Support for help - they don't know what this is!

What does it do?

Demonstrates the use of the Tableau JavaScript, TDE and REST APIs, along with Trusted Authentication
Allows you to create users w/o TabCmd or the Tableau Portal
Queries / displays data sources on a site
Allows you to upload an arbitrary CSV file directly to Tableau and then build a report against same with Web Authoring

Version

1.0

Installation

To begin, you must download and install Node.Js for your operating system (Windows, Linux)
You must also install and configure the Python-based Extract API.
And what good is the API without Python itself? Install Python 2.7.X for your OS.
Make sure that the Python install folder is in your PATH
Next, download the code using the Download ZIP button to your upper-right
Unzip to c:\node, and navigate there on the command-line.
Execute these commands to prepare the app

npm install

Note:

This sample application has only been "tested" (note the quotes?!) on Windows. It should work on Linux too. If it doesn't...well, you're a Linux person..which means you are smart can probably troubleshoot this yourself. Sorry for the OS bigotry.

To run the application, type:

node app

App Configuration

We're a bit lazy. We admit it. As a result you need to update the code a bit.

In c:\node\app.js modify:

Line 44: Update the username & password for an admin on your server
Line 47: Update the Tableau Server name we'll be connecting to
Line 185: Update the Tableau Server name again

In c:\node\views\aboutcompany.ejs, modify:

Line 8: Update the server name to your server

In c:\node\views\analyze.ejs, modify:

Line 68: Update the server name to your server

Note: This app hard codes the Site ID & Name 'rest' when we request trusted tickets. If you want to create users and/or read data sources from a site other than 'rest', you'll need to find/replace this value.

And another note: The app demonstrates creating a user in Tableau via the "register" functionality, but it doesn't save said user in local storage so you can use it again after you stop / restart the application. Consider modifying line 39 of app.js if you want to add a "permanent" user to the application.

Tableau Server Configuration

This application leverages Tableau Trusted Authentication to enable logged on user to view the viz on /about. Please review the documentation on same, and configure Tableau to trust the IP address or Machine Name of the box which runs this application.
This app assumes a site called "rest" exists on your server. Create a new site using 'rest' for both the name and id of the site.

License

MIT

FAQs / Notes

I see an error that sort of looks like this (data type may vary) whenever your code tries to turn a csv into an extract. What's going on?

Example:

Error: TypeError: an integer is required
    at PythonShell.parseError (C:\node\node_modules\python-shell\index.js:131:17)
    at ChildProcess.<anonymous> (C:\node\node_modules\python-shell\index.js:67:28)
    at ChildProcess.emit (events.js:98:17)
    at Process.ChildProcess._handle.onexit (child_process.js:809:12)
    ----- Python Traceback -----
    File "C:\node\public\js\csv_2_tde.py", line 216, in <module>
      newrow.setInteger(columnposition, datatyper(row[fieldname]))
    File "c:\python27\lib\site-packages\dataextract\Base.py", line 248, in setInteger
      , c_int(value)

Your CSV file most likely has mixed data types in the same column. For example, the code was expecting an INT based on the majority of the other values in the column, but this row just handed us a STRING.

What can you do? In /public/js/csv_2_tde.py, you can temporarily set the rowoutput flag (line 14) to True. Then navigate to /public/js and run the script manually to see debugging information. This extra information will tell you the row on which the script failed so you can fix your data. Don't forget to change rowoutput back to False.

#####I used to be able to process files with no problem. Now I see an error like the one above when I attempt to upload a file I KNOW is good. Why? The csv_2_tde.py script attempts to process ANY csv file it sees in the /public/uploads folder. You probably have an older, "broken" CSV file still sitting there and the script is trying to process it (and failing) BEFORE it gets to your "known good file". Clean up the /uploads folder and try again.

#####Why is my (int, datetime, double, whatever) column in the CSV being created as a String in the extract? When in doubt, the python script which generates your extract via the Tableau Extract API goes with "string". The script probaby sees data it can't quite figure out (like a date that looks like this: "12/31/1999 0:00")

#####Help! I can't this running! I see an error or something! I need technical support! Sorry about that, but this is a proof-of-concept sample application - not something that is production ready. It isn't supported by Tableau. In a pinch you might get someone in the Tableau Developer Community to give you a push in the right direction, however.

Special thanks to Ryan Robitialle for writing the Python script years ago so that we don't have to!

NNagpal / TCC_APIs