This processor collects various objects (eg. tasks, comments, etc...) from Asana via the specified
AsanaClientService. When the processor started for the first time with a given configuration
it collects each of the objects matching the user specified criteria, and emits FlowFiles
of each on the NEW relationship. Then, it polls Asana in the frequency of the configured Run Schedule
and detects changes by comparing the object fingerprints. When there are updates, it emits them through
the UPDATED and REMOVED relationships, respectively.
Each emitted FlowFile contains the Json representation of the fetched Asana object. These can be
processed further via the respective processors, that accept text data in this format. The FlowFiles
emitted from the REMOVED relationship have no content, because the actual data is not stored in the
processor, and so there is no way to retrieve the deleted content.
Each FlowFile, regardless to which relationship they were emitted from, have an asana.gid
attribute set, which contain the ID of the object in Asana. These IDs are globally unique within the Asana instance,
regardless of what type of object they were assigned to. In case of Events, these IDs are generated by the
client, because Asana does not keep track of these objects.
These are used only for content change detection.
Fingerprints are generally calculated by applying an SHA-512 algorithm on the retrieved object. In case
of immutable objects, like Attachments, these fingerprints are static, so updates (which is impossible
anyway) are not detected. In case of Projects and Tasks, where the last modification time is available,
these timestamps are stored as fingerprints.
By default, this processor emits each fetched object from Asana in a separate FlowFile. This is usually OK
for a workspace having low traffic, and thus generating data in low rate. For workspaces with high volume of traffic,
it is advisable to set the batch size to a reasonably high value, to have better performance. With this value set to
something other than the default (1), the processor will emit FlowFiles that have multiple items batched
together in a Json array, but in exchange, without having the asana.gid attribute set.
In case of collecting some objects, like Project Events, Tasks, and Team Members, the processor
requires/allows defining filters. In example: if you would like to collect Tasks, then you need to define the project
from where the tasks you would like to collect.
In these cases, when the filters refer to some parent object, you need to provide its name in the configuration, in
case-sensitive manner. Another important note to keep in mind, Asana lets the users create multiple objects with the
same name. In example: you can create two projects with name 'My project'. But when you need to refer to this project
by its name, it is impossible to figure out which 'My project' you intended to refer to, therefore these situations
should be avoided. In such cases, this processor picks the first one returned by Asana when listing them. This is not
random, but the ordering is not guaranteed.