René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Support analyzing multiple builds (by incorporating directory into the table indexes)","articleBody":"The current analyze schema assumes that only a single build is being analyzed, making it harder to compare builds.\n\nCurrently if you try to analyze multiple builds there are SQL exceptions thrown.\n\nIn particular it expects that the SerializedFiles names are unique (e.g. only one \"level0\" or only one version of a particular asset bundle).\n\nAnother repro: It will fail to analyze multiple BuildReport files if they have the same filename.  The schema overall supports multiple BuildReports, but if they have the same filename it breaks and we currently use the filename as a way to distinguish them from each other\n\nI think we could adjust the schema so that the directory where the archive or serialized file is incorporated into the tables as a way to distinguish builds.  E.g. the key to the directory would be added into the serialized file table and index would be on that key plus the Serialized File name.\n\nOne challenge is dealing with references between serialized files.  The reference records the name of the serialized file.  To resolve that reference properly we have to track the context of which directory is being scanned, e.g. assume that references only point to files inside the same directory.\n\nOnce working additional examples should be added to the documentation showing how to query just a single build and how to do some basic comparison of builds when they are inside the same database.","author":{"url":"https://github.com/SkowronskiAndrew","@type":"Person","name":"SkowronskiAndrew"},"datePublished":"2025-10-17T13:53:44.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":1},"url":"https://github.com/36/UnityDataTools/issues/36"}