| Age | Commit message (Collapse) | Author |
|---|
|
Some of these unwraps are fine to stay, mostly those that deal with
locks - in this case, crashing the program if something goes wrong is
probably fine.
However, we also had a lot of other places where we panic'd on errors,
even though we really shouldn't have. For example, an invalid zip file
would bring down the whole scanner. In this case, we now use proper
Result<>s and we log the error.
Some places stay with unwrap() for now, mainly the code that is rare and
obvious when it goes wrong - such as an overflow in input values. It
could be made nicer, but it is not a priority for now.
Some unwraps() have been changed to expect() to signal why they
shouldn't fail.
|
|
|
|
This enables filters such as
-time > 2020-01-01
-time < 2020-02-03
...
for time and duration, and later possibly also for more things (such as
a COUNT(...) construct).
This work tries to integrate them into the existing filter system as
seamless as possible, by providing a Comparator which implements
LogFilter.
The "type checking" is done at parse time, so nonsensical comparisons
like -time > 12s flat out give a parse error. This however might be
changed to a more dynamic system with run-time type checking, in which
case we could do away with the type parameter on Producer and simply
work with a generic Value. The comparator would then return an error if
two non-identical types would be compared.
Note that the system does not support arithmetic expressions, only
simple comparisons to constant values.
|
|
Having a ::new on each of the filter types was a bit weird, especially
because we returned Box<dyn ...> instead of Self (and clippy rightfully
complained). With this patch, we now have a bunch of normal functions,
and we don't show to the outside how a filter is actually implemented
(or what struct is behind it).
|
|
|
|
It's nice if you can print out the filter tree for debugging, so we're
requireing filters to be Debug now.
|
|
As it turns out, we can easily re-use the existing Filter machinery to
generalize over LogFilters (which operate on LogResults) and
PlayerFilters (which operate on Players).
The feature trait_aliases is not strictly needed but makes the function
signatures a bit nicer and easier to read, and it reduces the chances of
an error (e.g. by using Filter<&PartialEvtc, ...>).
|
