Problem
Currently, Attribouter fetches its information from two different locations: the configuration file that you write, and the repository API. However, there will occasionally be some information that is provided by both the file and the repository metadata. Because Attribouter gets the information from the API shortly after processing the configuration file, the obvious behavior would be to simply replace the information from the configuration file with its GitHub counterpart. However, that isn't a perfect solution, as shown in the following situations:
- one
ContributorWedgein aContributorsWedgedoes not have a git counterpart, but all of the others do; it would disappear - a
ContributorWedgehas anemailattribute, but its git counterpart does not provide one; it would disappear - the
AppWedge'sdescriptionattribute is a string resource that is translated into several languages, but the git repositories description is in English; it would always display in English
Solution
In order to work around these situations, rather than simply replacing duplicate information, Attribouter "merges" them together. Using this method, if an attribute is in the configuration file that doesn't have a git counterpart, it is kept, but all others are overwritten. Likewise, ContributorWedges inside of a ContributorsWedge are merged if their login attributes match, and are left alone if they do not (ex: contributors in the config file not present on git are left in the list, and contributors not in the config file but present on git are added as well).
However, you may notice that this solution does not account for the last situation, as it isn't possible to detect this automatically. As a workaround, Attribouter looks for a "^" character at the start of an attribute before merging it. If present, it will ignore its git counterpart. In other words, if the translated string was @string/app_desc, the description attribute can be set to ^@string/app_desc to prevent it from being overwritten by the API data.