Bugzilla

This metabug encompasses the work required to cache accessible table information (row/column coordinates, headers, etc.) from content processes in the parent process. This is part of the "Cache the World" project.

As part of the Cache the World project, we need to be able to cache table information from content processes in the parent process.

Trying to calculate counts, indexes, headers, etc. in the content process and send them up to the parent process would be very expensive and difficult to keep updated after mutations. Instead, we will send only what is absolutely necessary (spans and explicitly associated headers). We will then lazily create a cache data structure only when table information is requested by a client, looping through the entire table and calculating all the information we need. Whenever there is a mutation of the table structure, we throw away the entire cache, rebuilding it next time the client requests information. While building the cache is expensive, this ends up being a lot cheaper than continually trying to calculate the information we need on every client request. In addition, there are certain things that are very difficult (if not impossible) to implement without walking the entire table; e.g. handling row spans.

Aside from supporting CTW, this allows us to finally eliminate our dependency on layout for table info (bug 1686391), resolving a lot of dependent bugs, and make ARIA tables faster (bug 1591326).