We notice that masks can be described by 2 parameters:
1. Whether they have 0 or 1 as their high bits.
2. Whether they shift to the left or to the right on the next line.
Describing masks this way allows us to lift them to template variables and simplify the code.
We also avoid handling the mask in the `RenderLine` loop entirely.
Also fixes a foliage rendering bug: Transparent foliage pixels were previously blended but they should have been simply skipped.
1. Fixes the return value (bytes rendered).
2. Fixes line wrapping / end-of-rendering based on the given rectangle:
1. Accounts for `BaseLineOffset`.
2. Fixes an off-by-one error for the y coordinate.
3. Wraps the cursor when needed.
3. Fix chat input box dimensions (height is 3 * line height).
4. Set the hint that indicates that we do not render the current
IME suggestion (SDL_TEXTEDITING). This indicates to IME
that it should render the suggestion instead.
1. Unifies SDL1 and SDL2 text input handling.
2. Moves game-specific text input handling out of misc_msg.
3. Disables Unicode processing when not inputting text in SDL1.
This fixes gold withdrawal in SDL1.
When using `UNPACKED_MPQS`, avoid all the SDL machinery for reading
files.
This is beneficial not only due to reduced indirection but also because
we can test for the file's existence and get the file size without
opening it, which is much faster.
When rendering directly to the output buffer, we need to maintain the
state of what has been drawn and what needs redrawing per-buffer.
We previously tried to do it implicitly by checking `SDL_DOUBLEBUF` and
other flags. The previous implementation was broken in several
ways, resulting in rendering issues on devices that support 8-bit output
directly.
Changes this mechanism to explicitly maintain buffer state per output
buffer. The new mechanism doesn't require knowledge of the number of
buffers, and thus also works correctly with triple-buffering.
Fixes#5447
https://github.com/diasurgical/devilutionx-mpq-tools produces an unpacked MPQ
with all the graphics converted to CLX and the unused files removed.
This is primarily useful on RAM-constrained platforms, such as PS2,
because it eliminates the MPQ overhead.
Adds a build option to load from such unpacked directories instead of the MPQ.
These directories are searched for in the same locations
where the MPQs would be searched for otherwise.
Example directory layout:
* /usr/local/share/diasurgical/devilutionx/diabdat/ -- unpacked and converted diabdat.mpq
* /usr/local/share/diasurgical/devilutionx/hellfire/ -- unpacked and converted hellfire MPQs (all of them merged into 1 directory)
* /usr/local/share/diasurgical/devilutionx/fonts/ -- unpacked fonts.mpq
* /usr/local/share/diasurgical/devilutionx/pl/ -- unpacked pl.mpq
These directory structure is produced by calling `unpack_and_minify_mpq`
We want to be able to use unpacked MPQs on low-end platforms
(PS2/rg99/etc).
This is tricky on case-sensitive filesystems. Avoids case issues by
lowercasing all paths in the code (then we'll just need lowercased
listfiles).
The format is almost identical to CL2, except it uses the frame header
to store frame width and height instead of 5 32-line offsets.
This means we always have access to frame dimensions, so we can use it
as an on-disk format for our graphics as well.
Additionally, we may be able to optimize the rendering even more
in the future now that we have guaranteed knowledge of frame dimensions.
Convert CEL files to CL2 at load time. CL2 format is more efficient and is about as fast to render.
CEL vs CL2 sizes, on dLvl 5: https://gist.github.com/glebm/9bbdd76962abcd4fd2405ecd3379af97
Memory:
* Peak memory (while loading): -300 KiB
* Memory in-game (dLvl5): -700 KiB
* RG99 binary size: -15 KiB (1333096 -> 1317192)
Performance on rg99:
* On average, -1 FPS in town.
* Same FPS in dungeon (20 FPS on dLvl 1).
Rather than using the Windows-like `MAX_PATH`, what we really want is 2
things:
1. Maximum path of a file in an MPQ. This is now `MaxMpqPathSize`.
2. Max size for the known monster paths. This is now 64 for sounds and
TRNs, which is more than enough (picked arbitrarily).
Adds simple string / integer concatenation functions.
Many of the uses of `fmt::format` are simply concatenation
of a few strings and integers.
`StrCat` is an easier-to-read alternative to such uses of `fmt`.
1. Store all key data and all values data as 2 char arrays.
2. Change map keys to char pointers.
3. Change map values to offsets into the values array.
Example savings for Russian: 460 KiB -> 374.2 KiB (-85.8 KiB)
* Map: 68.5 KiB (with `string_view` keys it would be 101.5 KiB)
* Keys array: 108.7 KiB
* Values array: 197.0 KiB
* Non-int versions of `Point` and `Displacement`
This will allow us to make some structs, such as `ActorPosition`, much
smaller.
* ActorPosition: Use smaller types
`Monsters`: 56K -> 46K
* player.cpp: Reduce size of `DirectionSettings`
* CrawlTable: Displacement -> DisplacementOf<int8_t>
* CrawlTable: vector<vector> -> array<vector>
Also only allocate one vector during construction instead of two.
A bit less indirection.
* Monster#enemyPosition: Point -> WorldTilePosition
sizeof(Monster): 240 -> 232
* Monster: Further optimize field layout and sizes
sizeof(Monster): 232 -> 208
`Monsters` is down to 40,000 bytes
* DMonsterStr: _mx/_my -> position