// Copyright 2008-2016 David Robillard // SPDX-License-Identifier: ISC /** @file forge.h An API for constructing LV2 atoms. This file provides an API for constructing Atoms which makes it relatively simple to build nested atoms of arbitrary complexity without requiring dynamic memory allocation. The API is based on successively appending the appropriate pieces to build a complete Atom. The size of containers is automatically updated. Functions that begin a container return (via their frame argument) a stack frame which must be popped when the container is finished. All output is written to a user-provided buffer or sink function. This makes it possible to create atoms on the stack, on the heap, in LV2 port buffers, in a ringbuffer, or elsewhere, all using the same API. This entire API is realtime safe if used with a buffer or a realtime safe sink, except lv2_atom_forge_init() which is only realtime safe if the URI map function is. Note these functions are all static inline, do not take their address. This header is non-normative, it is provided for convenience. */ #ifndef LV2_ATOM_FORGE_H #define LV2_ATOM_FORGE_H /** @defgroup forge Forge @ingroup atom An API for constructing LV2 atoms. @{ */ #include #include #include #include #include #include #include #include #ifdef __cplusplus extern "C" { #endif // Disable deprecation warnings for Blank and Resource LV2_DISABLE_DEPRECATION_WARNINGS /** Handle for LV2_Atom_Forge_Sink. */ typedef void* LV2_Atom_Forge_Sink_Handle; /** A reference to a chunk of written output. */ typedef intptr_t LV2_Atom_Forge_Ref; /** Sink function for writing output. See lv2_atom_forge_set_sink(). */ typedef LV2_Atom_Forge_Ref (*LV2_Atom_Forge_Sink)( LV2_Atom_Forge_Sink_Handle handle, const void* buf, uint32_t size); /** Function for resolving a reference. See lv2_atom_forge_set_sink(). */ typedef LV2_Atom* (*LV2_Atom_Forge_Deref_Func)( LV2_Atom_Forge_Sink_Handle handle, LV2_Atom_Forge_Ref ref); /** A stack frame used for keeping track of nested Atom containers. */ typedef struct LV2_Atom_Forge_Frame { struct LV2_Atom_Forge_Frame* parent; LV2_Atom_Forge_Ref ref; } LV2_Atom_Forge_Frame; /** A "forge" for creating atoms by appending to a buffer. */ typedef struct { uint8_t* buf; uint32_t offset; uint32_t size; LV2_Atom_Forge_Sink sink; LV2_Atom_Forge_Deref_Func deref; LV2_Atom_Forge_Sink_Handle handle; LV2_Atom_Forge_Frame* stack; LV2_URID Blank LV2_DEPRECATED; LV2_URID Bool; LV2_URID Chunk; LV2_URID Double; LV2_URID Float; LV2_URID Int; LV2_URID Long; LV2_URID Literal; LV2_URID Object; LV2_URID Path; LV2_URID Property; LV2_URID Resource LV2_DEPRECATED; LV2_URID Sequence; LV2_URID String; LV2_URID Tuple; LV2_URID URI; LV2_URID URID; LV2_URID Vector; } LV2_Atom_Forge; static inline void lv2_atom_forge_set_buffer(LV2_Atom_Forge* forge, uint8_t* buf, size_t size); /** Initialise `forge`. URIs will be mapped using `map` and stored, a reference to `map` itself is not held. */ static inline void lv2_atom_forge_init(LV2_Atom_Forge* forge, LV2_URID_Map* map) { lv2_atom_forge_set_buffer(forge, NULL, 0); forge->Blank = map->map(map->handle, LV2_ATOM__Blank); forge->Bool = map->map(map->handle, LV2_ATOM__Bool); forge->Chunk = map->map(map->handle, LV2_ATOM__Chunk); forge->Double = map->map(map->handle, LV2_ATOM__Double); forge->Float = map->map(map->handle, LV2_ATOM__Float); forge->Int = map->map(map->handle, LV2_ATOM__Int); forge->Long = map->map(map->handle, LV2_ATOM__Long); forge->Literal = map->map(map->handle, LV2_ATOM__Literal); forge->Object = map->map(map->handle, LV2_ATOM__Object); forge->Path = map->map(map->handle, LV2_ATOM__Path); forge->Property = map->map(map->handle, LV2_ATOM__Property); forge->Resource = map->map(map->handle, LV2_ATOM__Resource); forge->Sequence = map->map(map->handle, LV2_ATOM__Sequence); forge->String = map->map(map->handle, LV2_ATOM__String); forge->Tuple = map->map(map->handle, LV2_ATOM__Tuple); forge->URI = map->map(map->handle, LV2_ATOM__URI); forge->URID = map->map(map->handle, LV2_ATOM__URID); forge->Vector = map->map(map->handle, LV2_ATOM__Vector); } /** Access the Atom pointed to by a reference. */ static inline LV2_Atom* lv2_atom_forge_deref(LV2_Atom_Forge* forge, LV2_Atom_Forge_Ref ref) { // NOLINTNEXTLINE(performance-no-int-to-ptr) return forge->buf ? (LV2_Atom*)ref : forge->deref(forge->handle, ref); } /** @name Object Stack @{ */ /** Push a stack frame. This is done automatically by container functions (which take a stack frame pointer), but may be called by the user to push the top level container when writing to an existing Atom. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_push(LV2_Atom_Forge* forge, LV2_Atom_Forge_Frame* frame, LV2_Atom_Forge_Ref ref) { frame->parent = forge->stack; frame->ref = ref; if (ref) { forge->stack = frame; // Don't push, so walking the stack is always safe } return ref; } /** Pop a stack frame. This must be called when a container is finished. */ static inline void lv2_atom_forge_pop(LV2_Atom_Forge* forge, LV2_Atom_Forge_Frame* frame) { if (frame->ref) { // If frame has a valid ref, it must be the top of the stack assert(frame == forge->stack); forge->stack = frame->parent; } // Otherwise, frame was not pushed because of overflow, do nothing } /** Return true iff the top of the stack has the given type. */ static inline bool lv2_atom_forge_top_is(LV2_Atom_Forge* forge, uint32_t type) { return forge->stack && forge->stack->ref && (lv2_atom_forge_deref(forge, forge->stack->ref)->type == type); } /** Return true iff `type` is an atom:Object. */ static inline bool lv2_atom_forge_is_object_type(const LV2_Atom_Forge* forge, uint32_t type) { return (type == forge->Object || type == forge->Blank || type == forge->Resource); } /** Return true iff `type` is an atom:Object with a blank ID. */ static inline bool lv2_atom_forge_is_blank(const LV2_Atom_Forge* forge, uint32_t type, const LV2_Atom_Object_Body* body) { return (type == forge->Blank || (type == forge->Object && body->id == 0)); } /** @} @name Output Configuration @{ */ /** Set the output buffer where `forge` will write atoms. */ static inline void lv2_atom_forge_set_buffer(LV2_Atom_Forge* forge, uint8_t* buf, size_t size) { forge->buf = buf; forge->size = (uint32_t)size; forge->offset = 0; forge->deref = NULL; forge->sink = NULL; forge->handle = NULL; forge->stack = NULL; } /** Set the sink function where `forge` will write output. The return value of forge functions is an LV2_Atom_Forge_Ref which is an integer type safe to use as a pointer but is otherwise opaque. The sink function must return a ref that can be dereferenced to access as least sizeof(LV2_Atom) bytes of the written data, so sizes can be updated. For ringbuffers, this should be possible as long as the size of the buffer is a multiple of sizeof(LV2_Atom), since atoms are always aligned. Note that 0 is an invalid reference, so if you are using a buffer offset be sure to offset it such that 0 is never a valid reference. You will get confusing errors otherwise. */ static inline void lv2_atom_forge_set_sink(LV2_Atom_Forge* forge, LV2_Atom_Forge_Sink sink, LV2_Atom_Forge_Deref_Func deref, LV2_Atom_Forge_Sink_Handle handle) { forge->buf = NULL; forge->size = forge->offset = 0; forge->deref = deref; forge->sink = sink; forge->handle = handle; forge->stack = NULL; } /** @} @name Low Level Output @{ */ /** Write raw output. This is used internally, but is also useful for writing atom types not explicitly supported by the forge API. Note the caller is responsible for ensuring the output is appropriately padded. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_raw(LV2_Atom_Forge* forge, const void* data, uint32_t size) { LV2_Atom_Forge_Ref out = 0; if (forge->sink) { out = forge->sink(forge->handle, data, size); } else { out = (LV2_Atom_Forge_Ref)forge->buf + forge->offset; uint8_t* mem = forge->buf + forge->offset; if (forge->offset + size > forge->size) { return 0; } forge->offset += size; memcpy(mem, data, size); } for (LV2_Atom_Forge_Frame* f = forge->stack; f; f = f->parent) { lv2_atom_forge_deref(forge, f->ref)->size += size; } return out; } /** Pad output accordingly so next write is 64-bit aligned. */ static inline void lv2_atom_forge_pad(LV2_Atom_Forge* forge, uint32_t written) { const uint64_t pad = 0; const uint32_t pad_size = lv2_atom_pad_size(written) - written; lv2_atom_forge_raw(forge, &pad, pad_size); } /** Write raw output, padding to 64-bits as necessary. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_write(LV2_Atom_Forge* forge, const void* data, uint32_t size) { const LV2_Atom_Forge_Ref out = lv2_atom_forge_raw(forge, data, size); if (out) { lv2_atom_forge_pad(forge, size); } return out; } /** Write a null-terminated string body. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_string_body(LV2_Atom_Forge* forge, const char* str, uint32_t len) { const LV2_Atom_Forge_Ref s = lv2_atom_forge_raw(forge, str, len); if (!s) { return s; } const LV2_Atom_Forge_Ref t = lv2_atom_forge_raw(forge, "", 1); if (!t) { return t; } lv2_atom_forge_pad(forge, len + 1U); return t; } /** @} @name Atom Output @{ */ /** Write an atom:Atom header. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_atom(LV2_Atom_Forge* forge, uint32_t size, uint32_t type) { const LV2_Atom a = {size, type}; return lv2_atom_forge_raw(forge, &a, sizeof(a)); } /** Write a primitive (fixed-size) atom. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_primitive(LV2_Atom_Forge* forge, const LV2_Atom* a) { return ( lv2_atom_forge_top_is(forge, forge->Vector) ? lv2_atom_forge_raw(forge, LV2_ATOM_BODY_CONST(a), a->size) : lv2_atom_forge_write(forge, a, (uint32_t)sizeof(LV2_Atom) + a->size)); } /** Write an atom:Int. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_int(LV2_Atom_Forge* forge, int32_t val) { const LV2_Atom_Int a = {{sizeof(val), forge->Int}, val}; return lv2_atom_forge_primitive(forge, (const LV2_Atom*)&a); } /** Write an atom:Long. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_long(LV2_Atom_Forge* forge, int64_t val) { const LV2_Atom_Long a = {{sizeof(val), forge->Long}, val}; return lv2_atom_forge_primitive(forge, (const LV2_Atom*)&a); } /** Write an atom:Float. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_float(LV2_Atom_Forge* forge, float val) { const LV2_Atom_Float a = {{sizeof(val), forge->Float}, val}; return lv2_atom_forge_primitive(forge, (const LV2_Atom*)&a); } /** Write an atom:Double. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_double(LV2_Atom_Forge* forge, double val) { const LV2_Atom_Double a = {{sizeof(val), forge->Double}, val}; return lv2_atom_forge_primitive(forge, (const LV2_Atom*)&a); } /** Write an atom:Bool. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_bool(LV2_Atom_Forge* forge, bool val) { const LV2_Atom_Bool a = {{sizeof(int32_t), forge->Bool}, val ? 1 : 0}; return lv2_atom_forge_primitive(forge, (const LV2_Atom*)&a); } /** Write an atom:URID. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_urid(LV2_Atom_Forge* forge, LV2_URID id) { const LV2_Atom_URID a = {{sizeof(id), forge->URID}, id}; return lv2_atom_forge_primitive(forge, (const LV2_Atom*)&a); } /** Write an atom compatible with atom:String. Used internally. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_typed_string(LV2_Atom_Forge* forge, uint32_t type, const char* str, uint32_t len) { const LV2_Atom_String a = {{len + 1, type}}; LV2_Atom_Forge_Ref out = lv2_atom_forge_raw(forge, &a, sizeof(a)); if (out) { if (!lv2_atom_forge_string_body(forge, str, len)) { LV2_Atom* atom = lv2_atom_forge_deref(forge, out); atom->size = atom->type = 0; out = 0; } } return out; } /** Write an atom:String. Note that `str` need not be NULL terminated. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_string(LV2_Atom_Forge* forge, const char* str, uint32_t len) { return lv2_atom_forge_typed_string(forge, forge->String, str, len); } /** Write an atom:URI. Note that `uri` need not be NULL terminated. This does not map the URI, but writes the complete URI string. To write a mapped URI, use lv2_atom_forge_urid(). */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_uri(LV2_Atom_Forge* forge, const char* uri, uint32_t len) { return lv2_atom_forge_typed_string(forge, forge->URI, uri, len); } /** Write an atom:Path. Note that `path` need not be NULL terminated. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_path(LV2_Atom_Forge* forge, const char* path, uint32_t len) { return lv2_atom_forge_typed_string(forge, forge->Path, path, len); } /** Write an atom:Literal. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_literal(LV2_Atom_Forge* forge, const char* str, uint32_t len, uint32_t datatype, uint32_t lang) { const LV2_Atom_Literal a = { {(uint32_t)(sizeof(LV2_Atom_Literal) - sizeof(LV2_Atom) + len + 1), forge->Literal}, {datatype, lang}}; LV2_Atom_Forge_Ref out = lv2_atom_forge_raw(forge, &a, sizeof(a)); if (out) { if (!lv2_atom_forge_string_body(forge, str, len)) { LV2_Atom* atom = lv2_atom_forge_deref(forge, out); atom->size = atom->type = 0; out = 0; } } return out; } /** Start an atom:Vector. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_vector_head(LV2_Atom_Forge* forge, LV2_Atom_Forge_Frame* frame, uint32_t child_size, uint32_t child_type) { const LV2_Atom_Vector a = {{sizeof(LV2_Atom_Vector_Body), forge->Vector}, {child_size, child_type}}; return lv2_atom_forge_push( forge, frame, lv2_atom_forge_write(forge, &a, sizeof(a))); } /** Write a complete atom:Vector. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_vector(LV2_Atom_Forge* forge, uint32_t child_size, uint32_t child_type, uint32_t n_elems, const void* elems) { const LV2_Atom_Vector a = { {(uint32_t)sizeof(LV2_Atom_Vector_Body) + n_elems * child_size, forge->Vector}, {child_size, child_type}}; const LV2_Atom_Forge_Ref out = lv2_atom_forge_write(forge, &a, sizeof(a)); if (out) { lv2_atom_forge_write(forge, elems, child_size * n_elems); } return out; } /** Write the header of an atom:Tuple. The passed frame will be initialised to represent this tuple. To complete the tuple, write a sequence of atoms, then pop the frame with lv2_atom_forge_pop(). For example: @code // Write tuple (1, 2.0) LV2_Atom_Forge_Frame frame; LV2_Atom* tup = (LV2_Atom*)lv2_atom_forge_tuple(forge, &frame); lv2_atom_forge_int(forge, 1); lv2_atom_forge_float(forge, 2.0); lv2_atom_forge_pop(forge, &frame); @endcode */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_tuple(LV2_Atom_Forge* forge, LV2_Atom_Forge_Frame* frame) { const LV2_Atom_Tuple a = {{0, forge->Tuple}}; return lv2_atom_forge_push( forge, frame, lv2_atom_forge_write(forge, &a, sizeof(a))); } /** Write the header of an atom:Object. The passed frame will be initialised to represent this object. To complete the object, write a sequence of properties, then pop the frame with lv2_atom_forge_pop(). For example: @code LV2_URID eg_Cat = map("http://example.org/Cat"); LV2_URID eg_name = map("http://example.org/name"); // Start object with type eg_Cat and blank ID LV2_Atom_Forge_Frame frame; lv2_atom_forge_object(forge, &frame, 0, eg_Cat); // Append property eg:name = "Hobbes" lv2_atom_forge_key(forge, eg_name); lv2_atom_forge_string(forge, "Hobbes", strlen("Hobbes")); // Finish object lv2_atom_forge_pop(forge, &frame); @endcode */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_object(LV2_Atom_Forge* forge, LV2_Atom_Forge_Frame* frame, LV2_URID id, LV2_URID otype) { const LV2_Atom_Object a = { {(uint32_t)sizeof(LV2_Atom_Object_Body), forge->Object}, {id, otype}}; return lv2_atom_forge_push( forge, frame, lv2_atom_forge_write(forge, &a, sizeof(a))); } /** The same as lv2_atom_forge_object(), but for object:Resource. This function is deprecated and should not be used in new code. Use lv2_atom_forge_object() directly instead. */ LV2_DEPRECATED static inline LV2_Atom_Forge_Ref lv2_atom_forge_resource(LV2_Atom_Forge* forge, LV2_Atom_Forge_Frame* frame, LV2_URID id, LV2_URID otype) { const LV2_Atom_Object a = { {(uint32_t)sizeof(LV2_Atom_Object_Body), forge->Resource}, {id, otype}}; return lv2_atom_forge_push( forge, frame, lv2_atom_forge_write(forge, &a, sizeof(a))); } /** The same as lv2_atom_forge_object(), but for object:Blank. This function is deprecated and should not be used in new code. Use lv2_atom_forge_object() directly instead. */ LV2_DEPRECATED static inline LV2_Atom_Forge_Ref lv2_atom_forge_blank(LV2_Atom_Forge* forge, LV2_Atom_Forge_Frame* frame, uint32_t id, LV2_URID otype) { const LV2_Atom_Object a = { {(uint32_t)sizeof(LV2_Atom_Object_Body), forge->Blank}, {id, otype}}; return lv2_atom_forge_push( forge, frame, lv2_atom_forge_write(forge, &a, sizeof(a))); } /** Write a property key in an Object, to be followed by the value. See lv2_atom_forge_object() documentation for an example. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_key(LV2_Atom_Forge* forge, LV2_URID key) { const LV2_Atom_Property_Body a = {key, 0, {0, 0}}; return lv2_atom_forge_write(forge, &a, 2 * (uint32_t)sizeof(uint32_t)); } /** Write the header for a property body in an object, with context. If you do not need the context, which is almost certainly the case, use the simpler lv2_atom_forge_key() instead. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_property_head(LV2_Atom_Forge* forge, LV2_URID key, LV2_URID context) { const LV2_Atom_Property_Body a = {key, context, {0, 0}}; return lv2_atom_forge_write(forge, &a, 2 * (uint32_t)sizeof(uint32_t)); } /** Write the header for a Sequence. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_sequence_head(LV2_Atom_Forge* forge, LV2_Atom_Forge_Frame* frame, uint32_t unit) { const LV2_Atom_Sequence a = { {(uint32_t)sizeof(LV2_Atom_Sequence_Body), forge->Sequence}, {unit, 0}}; return lv2_atom_forge_push( forge, frame, lv2_atom_forge_write(forge, &a, sizeof(a))); } /** Write the time stamp header of an Event (in a Sequence) in audio frames. After this, call the appropriate forge method(s) to write the body. Note the returned reference is to an LV2_Event which is NOT an Atom. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_frame_time(LV2_Atom_Forge* forge, int64_t frames) { return lv2_atom_forge_write(forge, &frames, sizeof(frames)); } /** Write the time stamp header of an Event (in a Sequence) in beats. After this, call the appropriate forge method(s) to write the body. Note the returned reference is to an LV2_Event which is NOT an Atom. */ static inline LV2_Atom_Forge_Ref lv2_atom_forge_beat_time(LV2_Atom_Forge* forge, double beats) { return lv2_atom_forge_write(forge, &beats, sizeof(beats)); } LV2_RESTORE_WARNINGS #ifdef __cplusplus } /* extern "C" */ #endif /** @} @} */ #endif // LV2_ATOM_FORGE_H