twitter connector

[Inactive User]

I have a few problems relating to the output of the Alteryx Twitter Connector, hopefully someone more familiar with the workings of the connector can help explain why.

1. I'm getting what appears to be a copy of the TweetBody coming through into the TweetID column more often than an actual TweetID (see image) should this happen?

2.There is also multiple rows for the same TweetID

2.  When I attempt to parse the TweetBody column it causes the Alteryx JSON Parser to fail with "Error: Designer x64: The Designer x64 reported: InboundNamedPipe::ReadFile:  Not enough bytes read. The pipe has been ended".  Without the JSON Parse tool the workflow will run through to completion.

3.  The description for the connector says that it only returns tweets from the last 7 days, however I'm seeing Tweet posted timestamps stretching back to 2009.

any assistance would be appreciated.

[Aguisande]

Hi, 

I'm experiencing the same thing. It looks that Twitter API is returning different things that it used to.

[NeilR · Accepted solution]

@Aguisande is correct the Twitter API response changed. @Inactive User Look for a fixed version of the Twitter Search tool on the Alteryx Analytics Gallery in the coming weeks.

[Inactive User]

Thanks @NeilR, look forward to seeing the updated tool on the Gallery soon.

[NeilR · Accepted solution]

The updated Twitter Search tool is now available in the Data Sources District. The previous version that shipped with Alteryx will be deprecated in the upcoming release (and therefore will be removed from the tool palette), but in the meantime once you install this tool you will see both versions side by side in the tool palette. To remove the previous version click the + (plus) symbol on the far right-hand side of the tool palette, navigate to the Connectors category, and uncheck the first Twitter Search tool.

[Aguisande]

Hi Neil,

It appears that the Tweet Source is not being parsed (to remove the html tags).

What is awesome to see is that the extended characters (accents and so) are being pulled like they appear on the original tweets.

Thanks for the update

[NeilR]

@Aguisande We decided to leave the source unparsed since the href attribute could be valuable in addition to the link name.

[NeilR]

And by the way, the Cleanse macro can take care of those HTML tags for you.

[Aguisande]

Neil,

Makes sense to keep it that way (now that I know it).

Best,

_AG_ 

[Aguisande]

Yes, I love the "old" Cleanse macro. 

Thanks