How tracardi process event context

This documentation describes in detail how Tracardi retrieves event context and profile geo location, utm parameters, etc.

Description

The code for retrieving the profile's geo location in Tracardi follows the following steps:

It first checks if the session is new by evaluating the session.operation.new condition. If the session is new, it proceeds with the following steps. If not, it skips the subsequent steps and moves to the next part of the code.
When a new session is detected, the code adds a "session-opened" event to the list of registered events. This event signifies the creation of a new session.
The code then retrieves the user agent data from the session's context. The user agent data provides information about the user's browser, device, and operating system. It extracts the user agent string from the session's context and parses it using the parse function.
From the parsed user agent data, the code extracts various device-related information such as the operating system version, operating system name, and device type. The device type can be "mobile," "pc," "tablet," or "email" based on the user agent data.
Next, the code checks if the session's context contains additional device information under the "device" key. If present, it uses that information to populate the session's device properties such as device name, brand, model, touch capability, and device type. If the session's context does not have device information, the code uses the corresponding values from the parsed user agent data.
The code then retrieves the spoken languages from the request headers and the geo country code from the device's geo information. It looks for the "accept-language" header in the tracker payload's request headers and parses the languages using the parse_accept_language function. It filters out languages with a length of 2 characters and maps them to the corresponding spoken languages using a predefined dictionary. The code also checks if the device's geo information includes a country code and maps it to spoken languages using the same dictionary. The resulting spoken languages are set as the session's language context and stored in the profile's "pii.language.spoken" field.
If the profile's auxiliary data does not contain a "geo" field, the code initializes it as an empty dictionary.
The code determines the continent based on the timezone information present in the tracker payload's context. It extracts the timezone value and checks if it is in UTC or a specific region. If it is in a specific region, the code splits the timezone string and retrieves the continent. The continent is then stored in the profile's auxiliary data under the "geo.continent" key.
The code identifies the markets based on the language codes obtained earlier. It checks if the language codes exist in a predefined dictionary that maps language codes to market values. If there are matching market values, they are added to the profile's auxiliary data under the "geo.markets" key.
The code tries to retrieve additional screen-related information from the tracker payload's context. It retrieves the screen resolution, color depth, and orientation if available, and assigns them to the session's device properties.
The code determines if the session's device is a bot by checking the user agent data. If the user agent indicates that it is a bot, the session's "app.bot" property is set to True.
The code extracts the browser name from the user agent data and assigns it to the session's "app.name" property. It also retrieves the browser version and assigns it to the session's "app.version" property. The session's "app.type" property is set to "browser" to indicate that it is a browser application.
If the tracker payload's context contains UTM parameters, the code creates a UTM object from the context data and assigns it to the session's "utm" property. The UTM parameters are used for tracking the source of the incoming traffic.
The code attempts to retrieve the device's IP address from the request headers. It looks for the "x-forwarded-for" header and assigns its value to the session's device IP property.
Lastly, the code tries to retrieve the browser's language from the session's context and assigns it to the session's "app.language" property.

After completing the above steps, the code checks if the location information is available in the tracker payload's context. If it is present, the code proceeds with the following steps. Otherwise, it skips to the next part of the code.

The code validates and parses the location information using the Geo class. If the validation succeeds, the code removes the location data from the tracker payload's context.
If the session's device geo information is empty, indicating that no geo information has been assigned yet, the code updates it with the new location data. This ensures that the session's device geo information is populated with the latest available location.
The code updates the profile's last device geo information if the new location data is different from the existing one. This ensures that the profile's last device geo information reflects the most recent geo location.

Finally, the code handles the scenario where there is no geo location available, but the session's device IP address is present. It checks if the necessary environment variables, namely MAXMIND_LICENSE_KEY and MAXMIND_ACCOUNT_ID, are set. If these variables are available, the code proceeds with the following steps. Otherwise, it skips the remaining part.

If the profile's last device geo information is empty, indicating that no geo location has been assigned yet, the code fetches the geo location using the MaxMind API. It creates a GeoLiteCredentials object with the MaxMind license key and account ID and calls the get_geo_maxmind_location function with the session's device IP address.
The fetched geo location is assigned to the profile's last device geo information, ensuring that it reflects the newly obtained geo location.
If the session is new, indicating that the user has just initiated the session, the code fetches the geo location using the MaxMind API. It assigns the obtained geo location to the session's device geo information and updates the profile's last device geo information if it is empty or different from the new location.
If there is an error while fetching the geo location using the MaxMind API, the code logs the error message.

The code for retrieving the profile's geo location in Tracardi follows the logic described above, enabling the system to determine and update the geo location information associated with the profile based on the available data from sessions, user agents, headers, and external APIs.