Class Parser

Nested Relationships

Nested Types

Class Documentation

class Parser

Public Types

enum MaxArrayPolicy

Values:

enumerator DISCARD_LARGE_ARRAYS
enumerator KEEP_LARGE_ARRAYS
enum BlobPolicy

Values:

enumerator STORE_BLOB_AS_COPY
enumerator STORE_BLOB_AS_REFERENCE
typedef std::function<void(const ROSType&, Span<uint8_t>&)> VisitingCallback

Public Functions

Parser(const std::string &topic_name, const ROSType &msg_type, const std::string &definition, SchemaFormat format = ROS_MSG)
Parameters:

  • topic_name – name of the topic to be used as node of the StringTree

  • msg_type – message type of the topic.

  • definition – schema text (ROS .msg or DDS IDL)

  • format – schema format: ROS_MSG (default) or DDS_IDL

inline void setMaxArrayPolicy(MaxArrayPolicy discard_entire_array, size_t max_array_size)

Default values are DISCARD_LARGE_ARRAYS and 100. The maximum permissible value of max_array_size is 10.000 (but don’t)

inline MaxArrayPolicy maxArrayPolicy() const
inline size_t maxArraySize() const
inline void setBlobPolicy(BlobPolicy policy)
inline BlobPolicy blobPolicy() const
const std::shared_ptr<MessageSchema> &getSchema() const

getSchema provides some metadata amount a registered ROSMessage.

ROSMessage::Ptr getMessageByType(const ROSType &type) const
bool deserialize(Span<const uint8_t> buffer, FlatMessage *flat_output, Deserializer *deserializer) const

deserializeIntoFlatContainer takes a raw buffer of memory and extract information from it. This data is stored in two key/value vectors, FlatMessage::value and FlatMessage::name. It must be noted that the key type is FieldsVector. This type is not particularly user-friendly, but allows a much faster post-processing.

IMPORTANT: this approach is not meant to be used with use arrays such as maps, point clouds and images.For this reason the argument max_array_size is used.

This function is almost always followed by CreateRenamedValues, which provide a more human-readable key-value representation.

Parameters:

  • buffer – raw memory to be parsed.

  • flat_output – output to store the result. It is recommended to reuse the same object multiple times to avoid memory allocations and speed up the parsing.

Returns:

true if the entire message was parsed or false if parts of the message were skipped because an array has (size > max_array_size)

bool deserializeIntoJson(Span<const uint8_t> buffer, std::string *json_txt, Deserializer *deserializer, int indent = 0, bool ignore_constants = false) const
bool serializeFromJson(const std::string_view json_string, Serializer *serializer) const
void applyVisitorToBuffer(const ROSType &msg_type, Span<uint8_t> &buffer, VisitingCallback callback) const

applyVisitorToBuffer is used to pass a callback that is invoked every time a chunk of memory storing an instance of ROSType = monitored_type is reached. Note that the VisitingCallback can modify the original message, but can NOT change its size. This means that strings and vectors can not be change their length.

Parameters:

  • msg_identifier – String ID to identify the registered message (use registerMessageDefinition first).

  • monitored_type – ROSType that triggers the invocation to the callback

  • buffer – Original buffer, passed as mutable since it might be modified.

  • callback – The callback.

bool walkSchema(Span<const uint8_t> buffer, Deserializer *deserializer, MessageWriter *writer) const

Walk the schema and write deserialized values to a MessageWriter. This is the unified deserialization path used by deserialize() and deserializeIntoJson().

inline void setWarningsStream(std::ostream *output)

Change where the warning messages are displayed.