diff options
Diffstat (limited to 'indra/newview/llmeshrepository.cpp')
-rwxr-xr-x[-rw-r--r--] | indra/newview/llmeshrepository.cpp | 304 |
1 files changed, 304 insertions, 0 deletions
diff --git a/indra/newview/llmeshrepository.cpp b/indra/newview/llmeshrepository.cpp index ff4b812cb0..f2f4586d6c 100644..100755 --- a/indra/newview/llmeshrepository.cpp +++ b/indra/newview/llmeshrepository.cpp @@ -35,6 +35,7 @@ #include "llappviewer.h" #include "llbufferstream.h" #include "llcurl.h" +#include "lldatapacker.h" #include "llfasttimer.h" #include "llfloatermodelpreview.h" #include "llfloaterperms.h" @@ -58,6 +59,8 @@ #include "llworld.h" #include "material_codes.h" #include "pipeline.h" +#include "llinventorymodel.h" +#include "llfoldertype.h" #ifndef LL_WINDOWS #include "netdb.h" @@ -82,6 +85,8 @@ U32 LLMeshRepository::sPeakKbps = 0; const U32 MAX_TEXTURE_UPLOAD_RETRIES = 5; +void dumpLLSDToFile(const LLSD& content, std::string filename); + std::string header_lod[] = { "lowest_lod", @@ -478,6 +483,46 @@ public: } }; +class LLWholeModelFeeResponder: public LLCurl::Responder +{ + LLMeshUploadThread* mThread; +public: + LLWholeModelFeeResponder(LLMeshUploadThread* thread): + mThread(thread) + { + } + virtual void completed(U32 status, + const std::string& reason, + const LLSD& content) + { + //assert_main_thread(); + llinfos << "completed" << llendl; + mThread->mPendingUploads--; + dumpLLSDToFile(content,"whole_model_response.xml"); + + mThread->mWholeModelUploadURL = content["uploader"].asString(); + } +}; + +class LLWholeModelUploadResponder: public LLCurl::Responder +{ + LLMeshUploadThread* mThread; +public: + LLWholeModelUploadResponder(LLMeshUploadThread* thread): + mThread(thread) + { + } + virtual void completed(U32 status, + const std::string& reason, + const LLSD& content) + { + //assert_main_thread(); + llinfos << "upload completed" << llendl; + mThread->mPendingUploads--; + dumpLLSDToFile(content,"whole_model_upload_response.xml"); + } +}; + LLMeshRepoThread::LLMeshRepoThread() : LLThread("mesh repo", NULL) { @@ -1239,6 +1284,7 @@ LLMeshUploadThread::LLMeshUploadThread(LLMeshUploadThread::instance_list& data, mUploadObjectAssetCapability = gAgent.getRegion()->getCapability("UploadObjectAsset"); mNewInventoryCapability = gAgent.getRegion()->getCapability("NewFileAgentInventoryVariablePrice"); + mWholeModelFeeCapability = gAgent.getRegion()->getCapability("NewFileAgentInventory"); mOrigin += gAgent.getAtAxis() * scale.magVec(); } @@ -1330,6 +1376,223 @@ BOOL LLMeshUploadThread::isDiscarded() void LLMeshUploadThread::run() { + if (gSavedSettings.getBOOL("MeshUseWholeModelUpload")) + { + doWholeModelUpload(); + } + else + { + doIterativeUpload(); + } +} + +#if 1 +void dumpLLSDToFile(const LLSD& content, std::string filename) +{ + std::ofstream of(filename.c_str()); + LLSDSerialize::toPrettyXML(content,of); +} +#endif + +void LLMeshUploadThread::wholeModelToLLSD(LLSD& dest, bool include_textures) +{ + // TODO where do textures go? + + LLSD result; + + LLSD res; + result["folder_id"] = gInventory.findCategoryUUIDForType(LLFolderType::FT_OBJECT); + result["asset_type"] = "mesh"; + result["inventory_type"] = "object"; + result["name"] = "your name here"; + result["description"] = "your description here"; + + // TODO "optional" fields from the spec + + res["mesh_list"] = LLSD::emptyArray(); +// TODO Textures + //res["texture_list"] = LLSD::emptyArray(); + S32 mesh_num = 0; + S32 texture_num = 0; + + std::set<LLViewerTexture* > textures; + + for (instance_map::iterator iter = mInstance.begin(); iter != mInstance.end(); ++iter) + { + LLMeshUploadData data; + data.mBaseModel = iter->first; + + LLModelInstance& instance = *(iter->second.begin()); + + for (S32 i = 0; i < 5; i++) + { + data.mModel[i] = instance.mLOD[i]; + } + + std::stringstream ostr; + + LLModel::Decomposition& decomp = + data.mModel[LLModel::LOD_PHYSICS].notNull() ? + data.mModel[LLModel::LOD_PHYSICS]->mPhysics : + data.mBaseModel->mPhysics; + + decomp.mBaseHull = mHullMap[data.mBaseModel]; + + LLSD mesh_header = LLModel::writeModel( + ostr, + data.mModel[LLModel::LOD_PHYSICS], + data.mModel[LLModel::LOD_HIGH], + data.mModel[LLModel::LOD_MEDIUM], + data.mModel[LLModel::LOD_LOW], + data.mModel[LLModel::LOD_IMPOSTOR], + decomp, + mUploadSkin, + mUploadJoints); + + data.mAssetData = ostr.str(); + + LLSD mesh_entry; + + LLVector3 pos, scale; + LLQuaternion rot; + LLMatrix4 transformation = instance.mTransform; + decomposeMeshMatrix(transformation,pos,rot,scale); + +#if 0 + mesh_entry["childpos"] = ll_sd_from_vector3(pos); + mesh_entry["childrot"] = ll_sd_from_quaternion(rot); + mesh_entry["scale"] = ll_sd_from_vector3(scale); +#endif + mesh_entry["position"] = ll_sd_from_vector3(LLVector3()); + mesh_entry["rotation"] = ll_sd_from_quaternion(rot); + mesh_entry["scale"] = ll_sd_from_vector3(scale); + + // TODO should be binary. + std::string str = ostr.str(); + mesh_entry["mesh_data"] = LLSD::Binary(str.begin(),str.end()); + + res["mesh_list"][mesh_num] = mesh_entry; + + // TODO how do textures in the list map to textures in the meshes? + if (mUploadTextures) + { + for (std::vector<LLImportMaterial>::iterator material_iter = instance.mMaterial.begin(); + material_iter != instance.mMaterial.end(); ++material_iter) + { + + if (textures.find(material_iter->mDiffuseMap.get()) == textures.end()) + { + textures.insert(material_iter->mDiffuseMap.get()); + + std::stringstream ostr; + if (include_textures) // otherwise data is blank. + { + LLTextureUploadData data(material_iter->mDiffuseMap.get(), material_iter->mDiffuseMapLabel); + if (!data.mTexture->isRawImageValid()) + { + data.mTexture->reloadRawImage(data.mTexture->getDiscardLevel()); + } + + LLPointer<LLImageJ2C> upload_file = + LLViewerTextureList::convertToUploadFile(data.mTexture->getRawImage()); + ostr.write((const char*) upload_file->getData(), upload_file->getDataSize()); + } + LLSD texture_entry; + texture_entry["texture_data"] = ostr.str(); + res["texture_list"][texture_num] = texture_entry; + texture_num++; + } + } + } + + mesh_num++; + } + + result["asset_resources"] = res; +#if 1 + dumpLLSDToFile(result,"whole_model.xml"); +#endif + + dest = result; +} + +void LLMeshUploadThread::doWholeModelUpload() +{ + mCurlRequest = new LLCurlRequest(); + + // Queue up models for hull generation (viewer-side) + for (instance_map::iterator iter = mInstance.begin(); iter != mInstance.end(); ++iter) + { + LLMeshUploadData data; + data.mBaseModel = iter->first; + + LLModelInstance& instance = *(iter->second.begin()); + + for (S32 i = 0; i < 5; i++) + { + data.mModel[i] = instance.mLOD[i]; + } + + //queue up models for hull generation + LLModel* physics = NULL; + + if (data.mModel[LLModel::LOD_PHYSICS].notNull()) + { + physics = data.mModel[LLModel::LOD_PHYSICS]; + } + else if (data.mModel[LLModel::LOD_MEDIUM].notNull()) + { + physics = data.mModel[LLModel::LOD_MEDIUM]; + } + else + { + physics = data.mModel[LLModel::LOD_HIGH]; + } + + if (!physics) + { + llerrs << "WTF?" << llendl; + } + + DecompRequest* request = new DecompRequest(physics, data.mBaseModel, this); + gMeshRepo.mDecompThread->submitRequest(request); + } + + while (!mPhysicsComplete) + { + apr_sleep(100); + } + + bool do_include_textures = false; // not needed for initial cost/validation check. + LLSD model_data; + wholeModelToLLSD(model_data, do_include_textures); + + mPendingUploads++; + LLCurlRequest::headers_t headers; + mCurlRequest->post(mWholeModelFeeCapability, headers, model_data, + new LLWholeModelFeeResponder(this)); + + do + { + mCurlRequest->process(); + } while (mCurlRequest->getQueued() > 0); + + mCurlRequest->post(mWholeModelUploadURL, headers, model_data["asset_resources"], new LLWholeModelUploadResponder(this)); + + do + { + mCurlRequest->process(); + } while (mCurlRequest->getQueued() > 0); + + delete mCurlRequest; + mCurlRequest = NULL; + + // Currently a no-op. + mFinished = true; +} + +void LLMeshUploadThread::doIterativeUpload() +{ if(isDiscarded()) { mFinished = true; @@ -2767,6 +3030,47 @@ void LLMeshUploadThread::createObjects(LLMeshUploadData& data) } } +void LLMeshUploadThread::decomposeMeshMatrix(LLMatrix4& transformation, + LLVector3& result_pos, + LLQuaternion& result_rot, + LLVector3& result_scale) +{ + // check for reflection + BOOL reflected = (transformation.determinant() < 0); + + // compute position + LLVector3 position = LLVector3(0, 0, 0) * transformation; + + // compute scale + LLVector3 x_transformed = LLVector3(1, 0, 0) * transformation - position; + LLVector3 y_transformed = LLVector3(0, 1, 0) * transformation - position; + LLVector3 z_transformed = LLVector3(0, 0, 1) * transformation - position; + F32 x_length = x_transformed.normalize(); + F32 y_length = y_transformed.normalize(); + F32 z_length = z_transformed.normalize(); + LLVector3 scale = LLVector3(x_length, y_length, z_length); + + // adjust for "reflected" geometry + LLVector3 x_transformed_reflected = x_transformed; + if (reflected) + { + x_transformed_reflected *= -1.0; + } + + // compute rotation + LLMatrix3 rotation_matrix; + rotation_matrix.setRows(x_transformed_reflected, y_transformed, z_transformed); + LLQuaternion quat_rotation = rotation_matrix.quaternion(); + quat_rotation.normalize(); // the rotation_matrix might not have been orthoginal. make it so here. + LLVector3 euler_rotation; + quat_rotation.getEulerAngles(&euler_rotation.mV[VX], &euler_rotation.mV[VY], &euler_rotation.mV[VZ]); + + result_pos = position + mOrigin; + result_scale = scale; + result_rot = quat_rotation; +} + + LLSD LLMeshUploadThread::createObject(LLModelInstance& instance) { LLMatrix4 transformation = instance.mTransform; |