Don't use a slice of pointers to opml items, when we can simply use a slice of
items instead. This should reduce the amount of memory allocations and the
number of indirections the GC has to process, speedup up the import process.
Note that this doesn't introduce any additional copies, as the only time a
slice of subscription is created, the items are created and inserted inline.
For small fixed-size structures, it's better to use a slice of values, instead
of a slice of pointers to values: they're stored contiguously and thus can be
iterated on quickly by the CPU, and it does remove an indirection per object
every time the GC kicks in.
- Use `user` everywhere, instead of sometimes `loggedUser`
- Delay the instantiation of some variables: no need to perform SQL queries for
nothing.
- Remove a SQL query getting the whole user struct when only user.ID is used.
Since tdewolff/minify supports SVG minimization, let's make use of it. As we
need to keep the license in the SVG because we're nice netizens, we can at
least use SPDX identifiers instead of using it verbatim.
This does save a couple of kB.
Each batch of feeds sent to the worker pool is now guaranteed to contain unique feed URLs.
When `POLLING_LIMIT_PER_HOST` is set, an additional limit is applied to the number of concurrent requests per hostname, helping to prevent overloading a single server.
Note: Additional requests may still be made during feed refresh. For example, to fetch feed icons or when the web scraper is enabled for a particular feed.
The unread page may show outdated entries when navigating back from an article, due to Chrome's back/forward cache (bfcache) restoring the page from memory.
Reference: https://web.dev/articles/bfcache
There is no need to have templates used only used in a single file be part of
every single other ones. This should reduce a bit the resident memory
consumption of miniflux.
- Use a simple struct instead of two slices to store the data and the checksums
of resources
- Remove a superfluous call to Sprintf
- Factorise presence check and data retrieval in some maps
- Size the maps when possible
There is no need to perform a heavy-weight SQL query gathering all the
information available on a feed when we're only interested in its last check
timestamp.
- Replace a call to fmt.Sprintf with a concatenation
- Explicit declaration of return values in FetchJobs
- Initialize the size of FetchJobs return value to b.limit: when b.limit is
used, which is most of the time, this avoid resizing the slice, and when it
isn't, the size of the map is set to 0, which is equivalent to the previous
situation anyway.
- Move a call to `request.UserID(r)` to a lower scope.
- Move the seeking inside of DetectFeedFormat instead of having it everywhere
in ParseFeed
- Provide a hint for the compiler to eliminate a useless bound check in
DetectJSONFormat, otherwise it'll check that buffer[i] is valid on every
iteration of the loop. This shouldn't make a big difference, but oh well.
Apparently, postgresql isn't smart enough to realize that once a true value
value is found as part of a `SELECT true`, there is no need to scan the rest of
the table, so we have to make this explicit. We could also have used the
`SELECT EXISTS(…)` construct, but it's more verbose and I don't like it.
