| Age | Commit message (Collapse) | Author |
|---|
|
This is meant to spare clients from having to percent-encode
room aliases, v3 event ids etc. that happen to hit the endpoint path.
It is unfair to expect clients to do that since they are not supposed
to care about the shape of CS API, which parameter should be encoded
in which way. The trick (together with the slightly updated GTAD
configuration) is to percent-encode parts that happen to be QStrings
and not `const char[]`'s while passing all constant parts as plain
C character literals. This also allows to make it more certain that
the path is correctly encoded by passing and storing QByteArray's
wherever the path is already encoded, and only use QStrings (next to
const char arrays) before that.
Since the change alters the API contract (even if that contract was
crappy), some crude detection of percent-encoded stuff on input is
inserted; if input is already percent-encoded, a warning is put to
the logs, alerting developers about the change.
|
|
|
|
Qt 6 has Qt::SingleShotConnection; connectSingleShot remains just for
the sake of compatibility across Qt 5 and Qt 6. If you target Qt 6 only,
feel free to use the Qt facility directly.
|
|
libQuotient 0.7 really requires Qt 5.12, nothing earlier will work.
|
|
This has always been merely a workaround to enable
connectUntil/connectSingleShot and was never intended to be used
elsewhere, let alone in clients.
|
|
|
|
|
|
|
|
It was too permissive on characters before the identifier and also
allowed the domain name to start on dash, which should not occur.
Closes #512.
|
|
|
|
Fixes #510.
|
|
|
|
|
|
|
|
Add room types to RoomCreateEvent
|
|
|
|
Add convenience function for activating encryption
|
|
Implement the mxc protocol in the NetworkAccessManager
|
|
|
|
|
|
Clients don't need to do it themselves.
|
|
|
|
|
|
|
|
EncryptionEvent constructor
|
|
To provide more room for internal changes in BaseJob.
|
|
- QNetworkReply::isSequential() already returns `true`, there's no need
to overload it again.
- Use `Q_SLOTS` instead of `slots` because it's an external library
interface and clients may use other libraries using `slots` identifier;
- Use `emit` instead of `Q_EMIT` because this is a part of internal
implementation and if we ever use a library that has an `emit`
identifier, a massive search-replace will be in order anyway.
- Use `QMetaObject::invokeMethod()` with a queued connection as
a clearer way to achieve the same goal as `QTimer::singleShot(0, ...)`.
|
|
QThreadStorage accepts ownership over stored objects.
|
|
The query is easier to manipulate; and the original mxc URL is not used
for the real network request anyway.
|
|
|
|
|
|
|
|
Co-authored-by: Alexey Rusakov <Kitsune-Ral@users.sf.net>
|
|
Co-authored-by: Alexey Rusakov <Kitsune-Ral@users.sf.net>
|
|
There was a mess with fileTransferCancelled(); it was only emitted when
a download (but not an upload) was cancelled; besides, in case of
downloads a file transfer info structure was getting deleted whereas
uploads left a file transfer in Cancelled status. This all now converges
on:
- fileTransferFailed() for both failures and cancellations (to simplify
slot connection, and also to follow the practice in, e.g., Qt Network).
- the file transfer info structure is kept around in Cancelled status,
following the logic used for failures. There's no particular cleanup
which may become a problem if one uploads and cancels many times
(download file transfers are keyed to event ids, mitigating
the problem); this will be fixed in another commit.
Closes #503. Closes #504.
|
|
|
|
|
|
Allows images to be loaded using the NetworkAccessManager instead of an
ImageProvider
|
|
Require CMake 3.16, kindly ask for C++20; drop qmake
|
|
Also: drop olm from the LGTM build environment, it's of no use there
for now.
|
|
|
|
|
|
Also, refresh the documentation a bit.
|
|
CI: add Windows, CodeQL
|
|
|
|
This supersedes passing clang-format options in a separate
CLANG_FORMAT_ARGS CMake variable.
|
|
|
|
`${{ runner.workspace }}` is, unfortunately, not portable to Windows.
|
|
|
|
|
