gems-uff / mestre Goto Github PK

The chunkRelSize attribute in the dataset should be a value between 0 and 1, representing the size of a conflicting chunk proportional to the file where it is located.
For example, for a conflicting chunk with 10 lines located in a file with 500 lines, the attribute value should be 0.02.

Currently, the script collect_attributes_db.py is using the size of the file after the conflict resolution. However, this should not happen, since it would generate inconsistencies. For example, the current version of the dataset includes rows with a chunkRelSize of 2864. The file to be considered must be the conflicted file, after the merge replay.

Since this attribute is going to be extracted again after the fix, we should also add the attributes extracted with the collect_chunk_size.py script into the collect_attributes_db.py script.

We should also add a mechanism to collect information about chunks that could not be extracted by the script.