magnum

Commit Graph

Author	SHA1	Message	Date
Vladimír Vondruš	440ece15fb	Audio,Text,Trade: accept void ArrayViews in importer plugins. This makes it much less annoying to pass arbitrarily typed data, such as std::uint8_t or char8_t and what not. It was already done like this for the new shader converter plugins, where the input is often 32-bit ints for SPIR-V. OTOH the internal virtual API is kept with ArrayView<const char>, as that makes it easier to operate on by the implementations.	5 years ago
Vladimír Vondruš	cbf227eb5c	Trade: bump the AbstractImporter plugin interface string. This is one of those nasty ABI incompatibilites that shuffles functions around, so better force a recompile this way.	5 years ago
Vladimír Vondruš	116d327a64	Trade: AbstractImporter::doOpenData() now takes a rvalue Array. This allows to better describe memory ownership and transfer it instead of forcing the plugins to allocate their own local copy if the import happens in-place on the imported data. Right now that's mainly for the openFile() use case, which implicitly allocated an Array with file contents only to pass it to openData() which then made a copy because it could not make any assumption about data scope. In other words, certain plugins (TgaImporter, KtxImporter, DdsImporter, CgltfImporter and possibly others) will now have their peak memory usage halved.	5 years ago
Vladimír Vondruš	70009347bc	Trade: use a better-debuggable way to test passed parameters. The original tests were implemented back when it was only possible to use the verification macros from within test cases.	5 years ago
Vladimír Vondruš	56cc2d6f91	Trade: doc++	5 years ago
Vladimír Vondruš	78cf81bfca	Trade: mutable access to MaterialData attribute values. Hah, so many overloads. Not providing mutable access to keys or layer offsets as that would break the invariant of the internal array always being sorted.	5 years ago
Vladimír Vondruš	1fb707497f	Trade: not sure why these would need to be in that ifdef.	5 years ago
Vladimír Vondruš	145d41dcff	Trade: doc++	5 years ago
Vladimír Vondruš	602f08881b	Trade: ensure AbstractImporter::fooForName() returns indices in bounds. I just accidentally returned a wrong index from a SceneData backwards compatibility wrapper and realized it'd be good to have that checked always.	5 years ago
Vladimír Vondruš	5ef7bd26e7	Trade: improve MaterialData type mismatch assertions. Followup to `26a0bf94ff` where it was done for MeshData, forgot that this class suffers from the same issue.	5 years ago
Vladimír Vondruš	ebfbfcbb73	Trade: doc++	5 years ago
Vladimír Vondruš	781c5c3e08	Trade: doc++	5 years ago
Vladimír Vondruš	9b41947dbc	Trade: doc++	5 years ago
Vladimír Vondruš	813d82717f	Trade: no need to export this helper.	5 years ago
Vladimír Vondruš	25320af545	Trade: pass strided views by const&. Those have 3 pointers at least, my limit for passing by value is trivial copyability and two pointers. I hope that reflects the actual HW at least vaguely, heh.	5 years ago
Vladimír Vondruš	2b2b7d5f2f	Trade: provide a clear doc snippet for AbstractImporter::openData(). The snippet with callbacks was too heavyweight for this kind of operation and workflow with in-memory data may be non-obvious to some users.	5 years ago
Vladimír Vondruš	5c06eecde8	Trade: actually, do it like this.	5 years ago
Vladimír Vondruš	bf6caece50	Trade: fix MeshData range checks for array attributes. And add a comment explaining why we don't check the pointer for empty meshes -- otherwise empty interleaved meshes would fail with stuff like Trade::MeshData: attribute 0 [0xc:0xc] is not contained in passed vertexData array [0x0:0x0] which ... helps nobody.	5 years ago
Vladimír Vondruš	b4666a2194	Trade: improve tests for MeshData index/attribute view range checks. Found a bug with attribute array size not being taken into account. It triggers now.	5 years ago
Vladimír Vondruš	2059b2b164	Trade: doc++	5 years ago
Vladimír Vondruš	093957a6ed	Trade: no need to zero-init memory in MeshData::*AsArray() helpers. It's all going to get overwritten by the data anyway, so this just unnecessarily hurts perf.	5 years ago
Vladimír Vondruš	3eb5f887d0	Trade: move a helper definition above where it's used first.	5 years ago
Vladimír Vondruš	cc8d88cbb1	Trade: move a test case to where it fits better.	5 years ago
Vladimír Vondruš	206c4d7a9e	Trade: no need to type that much.	5 years ago
Vladimír Vondruš	26a0bf94ff	Trade: improve MeshData type mismatch assertions. We do know what type the user requested, so why not print it as well. And for the array access the message was especially lazy.	5 years ago
Vladimír Vondruš	7a91b56aa1	Trade: test all MeshData mutable/array attribute assertion variants.	5 years ago
Vladimír Vondruš	c904cd0599	Trade: properly propagate array size from MeshData::attributeData(). The test now passes again.	5 years ago
Vladimír Vondruš	f4da3d78cc	Trade: merge MeshData array attribute tests with the main test. Because otherwise we don't properly test all cases. Case in point -- attributeData(UnsignedInt) wasn't correctly propagating the array size, causing the new tests to fail. Fix in the next commit.	5 years ago
Vladimír Vondruš	0ba5f33f73	Trade: consistent terminology in an assert message.	5 years ago
Vladimír Vondruš	343fc8794d	Trade: don't need to delegate trivial things.	5 years ago
Vladimír Vondruš	c24a762e15	Trade: not sure what's up with this silly temporary. Some places have it, some don't. It's shorter without.	5 years ago
Vladimír Vondruš	2b8d9c7e74	Trade: internal API was named in an outdated way. Long ago, during one of the few initial MeshData iterations. there was no VertexFormat, but rather Trade::MeshAttributeType. I managed to get rid of it mostly, except this place.	5 years ago
Vladimír Vondruš	fe57f23e12	Trade: const++	5 years ago
Vladimír Vondruš	e1d62fb154	Trade: typos.	5 years ago
Vladimír Vondruš	46a6724198	Trade: doc++	5 years ago
Vladimír Vondruš	03b7adf71b	imageconverter: some examples of how to use the new functionality. Originally I wanted to show how to convert a JPEG to an EXR directly, however after trying and miserably failing to implement that inside OpenExrImageConverter I realized the plugin is definitely not the place where to perform such conversion. So this will have to wait until there's some proper API in TextureTools or somewhere.	5 years ago
Vladimír Vondruš	77b979041b	imageconverter: ability to extract layers again.	5 years ago
Vladimír Vondruš	5d4342f1c1	imageconverter: implicitly import and convert all image levels. In case of --layers and --levels this only works if the input images have a single level, otherwise --level has to be set. The internal implementation would be too complex otherwise. As a consequence, combining a set of 2D mipmapped images into a 3D mipmapped image means one first has to combine particular 2D image levels to 3D levels and then combine all 3D levels to a 3D mipmapped image, it can't be done in a single step and it also can't be done by first combining levels and then layers.	5 years ago
Vladimír Vondruš	e017ba135c	imageconverter: ability to combine multiple levels together. I still have to figure out the interactions between --layers and --levels and how to implicitly handle conversion of multi-level files.	5 years ago
Vladimír Vondruš	b83099a8c6	imageconverter: ability to combine multiple image layers together. Hey, this was fun, after all!	5 years ago
Vladimír Vondruš	c3213cff9b	imageconverter: print the input filename in each error message. In preparation for support of multiple input files.	5 years ago
Vladimír Vondruš	7a072c59c2	imageconverter: ability to convert also 1D and 3D images. As well as a proper, graceful and helpful diagnostic for images and levels out of bounds. Before the thing just asserted and died.	5 years ago
Vladimír Vondruš	7702c579a5	imageconverter: shorten the argument help a bit.	5 years ago
Vladimír Vondruš	36b7446a6a	Trade: add docs about multi-level image conversion.	5 years ago
Vladimír Vondruš	eea9d32959	Trade: tighten AbstractImageConverter input requirements. The multi-level APIs still don't check anything regarding level sizes and order since for example *.ico has no restrictions at all. But the rest (like non-zero size) is a restriction for all file formats I'm aware of. TgaImageConverter test had to be adapted, I expect a lot of breakages in plugins tests as well. But user code should be fine I think. Also reduced the rather excessive dimensions in AbstractImageConverterTest, since the allocation requirement now made the default Emscripten heap OOM.	5 years ago
Vladimír Vondruš	696e9fb105	imageconverter: note how many dimensions the image has in --info.	5 years ago
Vladimír Vondruš	8ce842117e	imageconverter: am I stupid? yes!	5 years ago
Vladimír Vondruš	008583d2b2	Trade: support specifying mip levels in AbstractImageConverter. To avoid combinatorial explosion of functions that have to be implemented on the plugin side, the single-image variants proxy to the multi-level variants, if available. THe plugin interface string had to be bumped because otherwise there was a really nasty silent ABI break where instead of doSetFlags() it was suddenly calling doConvert().	5 years ago
Vladimír Vondruš	4a7b15cc95	Trade: simplify AbstractImageConverter tests a bit. No need to fill in the images with data if these won't get used at all.	5 years ago
Vladimír Vondruš	b0dc496cba	Trade: doc++	5 years ago

1 2 3 4 5 ...

483 Commits (440ece15fb93304da9a5451e7bcda543b9dde569)