identity-finder

Find possible identity fields per data source type and columns values (optional)

identities.yaml

This file describes how to find possible identities for a given data source type.

It has two sections:

• datasources - describes how to find identity data in a given data source type.
• default - default list to fall-back when a given data source configuration cannot be found.

Default section

This contains a list of all possible identity field candidates across all data source types.

If the identity is named, say user you can possibly find identity data in fields like: user_id, user_uuid, user_bot, user_name, user_org_name, user_multi_org_names etc.

If the identity name contains a suffix _[2-N] - it means that a given identity is a so-called multi identity, for example recipient_[2-50].

It means that there can actually be up to 50 idenity fields with that name, for example recipient, recipient_2, ..., recipient_50 (up to 50 can also mean 10, 2, 1 or none).

Actual possible identity data fields would be: recipient_uuid, recipient_2_id, recipient_50_org_name etc.

Datasources section

Property name is the actual data source name, so properies can be: git, gerrit, github, confluence etc.

Each property can have 4 arrays - all are optional, they are:

• identities: keeps a list of possible identity fields, like [author, reporter], if that array is empty [] or not specified - it means that a given data source has no identities data (for example github/repository).
• columns: for some datasources (like gerrit, github) list of possible identity columns depends on which kind of document we're looking at, here we can specify columns to distinguish different document types, most often this contains type and/or category column(s), example: [category, type]. If columns is empty ([]) or not specified - it means all documents have the same structure in this data source type.
• values: must be used with columns to specify expected columns values, so it must have the same number of elements as columns, example combination to specify GitHub PR reviews: columns:[category, type], values:[pull_request, pull_request_review] or gerrit approval: columns:[type], values:[approval].
• max: specifies if some of identities fields are so called multi-value fields or not, that array must eiter have the same number as identities array or be empty (we are assuming as if it has value 1 for all identities array entries then). For each identity this value can be: 1 (single item), 0 (means the same as 1), -1: unlimited, N. Specifying for example for grups.io: identities:[author, recipient] and max:[1,50] means that actual identity candidates are: author (single) and recipient, recipient_2, ..., recipient_50 (up to 50 recipients).