Summary:
This diff improves LogBox performance by not storing and updating logs that are ignored.
Previous we stored all logs, including ignored, as a set. This was so that later, when we show a list of all logs, we would be able to show the ignored logs as well if toggled on. We stored the logs as:
```
const logs = new Set([
{
message: "Not ignored",
ignored: false,
},
{
message: "Ignored",
ignored: true,
},
// 100s more ignored logs
]);
```
But it turns out, we can have hundreds of ignored logs within seconds in some parts of the app. This means we we're re-rendering the LogBoxContainer hundreds of times with a filter on this set to filter out the ignored logs, just to change none of the content.
Now we store as:
```
const logs = new Set([
{
message: "Not ignored",
},
]);
```
Later, when we want to show ignored logs, we'll store these separately.
Changelog: [Internal]
Reviewed By: cpojer
Differential Revision: D18055799
fbshipit-source-id: f5e21f66bb4ab6137d5d3908e8c03e119e3805d5
Summary:
In YellowBox we were parsing the component stack trace in a narrow set of use cases, which are commonly missed (I see this working ~50% of the time). That seems to be overly cautious. In LogBox, since there's a nice UI for component stack traces, let's be more aggressive about finding the component stack and we can address it with feedback if we find that there are real issues with this strategy in practice
Changelog: [Internal]
Reviewed By: cpojer
Differential Revision: D18053389
fbshipit-source-id: 48f116e2bd3f8cc43d53d3668fd6d5b8d7cba2a4