* Re-added the ability to set a single sorter, this time by ID.
* Multi-sorter setting now uses the magic sort_N IDs.
* Cleaned up internal handling to remove reliance on header grid columns.
Sorters will still be looked for in the header, but rely on the specified ID.
This did (I'm pretty sure) work before, but now it's explicit.
* The order_pair typedef has been removed
* set_active_sorting_option has been removed set_active_sorter and now takes its arguments separately instead of as a pair.
* set_active_sorter will use the bound sorter header on-modified handler
instead of calling order_by_column directly
* set_sorting_options has been renamed to set_sorters
* get_active_sorting_option has been renamed get_sorter
This unifies handling of the scrollbar modes without having to manually set them in the builders. At one point, this wasn't too common, but it's become common enough that having a separate builder is cleaner
This makes attribute_value and lexical_cast use the "new" to/from_chars api.
Its main advantages are:
- It's guaranteed to be locale independent, hopefully fixing all cases of #3945 and similar
- It fixes some cases config serialization, in particular the test
```
cfg["x"] = "9.87654321";
BOOST_CHECK_EQUAL(cfg["x"], 9.87654321);
```
- Previously the lexical_cast implementation used exception
handling for invalid formats (catching std::invalid_argument)
which made noise during debugging (and is also slower if it
is not optimized out).
- It's faster
So far afaik the only compiler which has a complete and proper to/from_chars implementation is msvc, gccs implementation of from_chars sometimes uses strtod under the hood and clang simply hasn't implemented from_chars for floating point numbers yet at all (actually the upcomig clang 20 will have it). Luckily for us, there is now also boost::charconv that can be used. So this raises to minimum build requirement to have at least one of:
- msvc 2019 update 5
- gcc 11
- clang 14 (i have added a fallback implementation of from_chars for this case, that doesn't support all of its features, and is probably certainly not as fast as the boost version, but supports the features that we use from it)
- boost 1.85
Since in particular the gcc implementation isn't that good (at least it on gcc11), boost charconv is the preferred implementation that is used if available.
This also removes a strange overload for
pointers to integers in lexical_cast while changing lexical_cast to use the new api.
Fixed map editor crashing when creating or opening scenario after played local scenario before opening map editor. Resolves#9563. The cause of the bug was that the ai manager singleton pointer was not set to nullptr after it was destructed. Fixed this by making ai manager destructor set singleton to nullptr. Before this the ai_map_ map member has to be cleared in destructor because it might try to access the singleton when destructed.
---------
Co-authored-by: SomeName42 <>
These were added by clang-tidy's autofix, which both tried to convert
the parameters to const references and std::move them. The correct
behaviour is to convert only to a const reference.
Instead of two functions for translatable and non-translatable options, comparison will be determined by the return value of the sorter. Existing translatable sorting options have been adjusted to return t_string rather than string.
For whatever reason, callback_save_value had type
std::function<void(int)> in the .cpp file but
std::function<void(const int)> in the header file. This reconciles them
to both take in an integer as a parameter.
This was the last use of lg::format_timespan, which was only ever used in server code anyway. The server probably shouldn't be formatting this anyway, but as long as it is, we don't need a dedicated function or fancy formatting.
Bad design to return mutable reference to static object. This method has the possibility of creating a new default entry in the by_name map, but the remove call will be a no-op in that case (as it was before), so no harm done.
previously the game would return to titlescreen here if mp_info_
but was null but
`game_config::debug && state_.classification().is_multiplayer()`
returned true which happend when a multiplayer game was reloaded
via the sp loadgame dialog.
This fixes the issue by simply removing the "always show mp staging
between scenarios in debug mode" feature, which imo never was that
useful anyways.
This not only removes duplicate information which can always cause bugs if it's out of sync,
this can also be used an an easy way to detect the old undo stack format
this way even when enter/exit_hex events with [on_undo] are involved, undoing happens
in reverse order of the gamestate changes during the original action
This commit splits undo actions into multiple smaller steps. The main advantages:
- Its allows us always undo the parts of an action in the reverse order of which
they originally happen, preventing bugs that could be caused if the parts
interact with each other in a way where the order matters (like for example
an [on_undo] resetting a units moves, when the event happend in a move action
which would then also reset the units moves on undoing).
- It's easier to add (c++) undo steps for specific steps even if they are used
from wml, like spending gold.
- [do_command] no longer confused the undo stack (in fact it's by default now undoable).
- All actions are put onto the undo stack in the same way, previously it was
often unclear whether the undo stack is empty during a specific step during
the execution of an action, in particular when people tried to use
`undo_stack->empty()` to check whether an action could still be undone.
This also adjusts the periods for years and months. Previously, we were using values of 30 days for a month and 12 months for year. Now, we use the chrono values of a month as 1/12 of a year and a year as 365.2425 days.
Instead of simply returning a string, `is_ip_banned` now returns a struct with an error code, ban reason, and ban time remaining. This avoids doing time duration formatting on the server and allows the error message to be localized on the client. It also makes the ban handling interface more generic in server_base, which should hopefully allow forum bans to be handled this way as well.
