API for cached tables #23
Conversation
pcmoritz
force-pushed
the
cache
branch
2 times, most recently
from
914a2d0 to
ab766a3
Compare
|
|
||
| /* Update task in the task log cache. This will append the new status to the | ||
| * task log cache. */ | ||
| void cached_task_log_update_task(db_cache *db, |
There was a problem hiding this comment.
This is just updating status. We should make clear about this in the name of the function as well.
|
|
||
| /* Return an array of plasma managers that have the object (the last parameter | ||
| * is an out parameter, which is assumed to be an array we append to). */ | ||
| int cached_object_table_get(db_cache *db, |
There was a problem hiding this comment.
Maybe change cached_* to cache_* ?
While for get() "cached" makes sense, it doesn't make that much sense for add().
In general the prefix of the function should be a noun as it is used to name the object on which the function operate on. Just emulate a poor man object oriented design ;-)
object_name_method_name()
or maybe
object_name__method_name()
to better separate object_name and method_name.
If we follow the above rule, maybe the best prefixes are object_cache, task_cache, etc.
|
|
||
| /* Get task from the task log cache. The last two parameters are out parameters, | ||
| * which are both assumed to be allocated. */ | ||
| int cached_task_log_get_task(db_cache *db, |
There was a problem hiding this comment.
By the above rule the name for this would be
task_log_cache_get_task()
or
task_log_cache__get_task()
These are going to be used in the global scheduler and potentially also other components if performance requires it.