Abstract classes

Abstract classes describe the music representation at logical level and are used to handle musical tag. Each tag has a corresponding abstract class, which name starts with AR and which is implemented in a file that carries the name of the class. For example, the \meter tag is described by the ARMeter class, which is implemented in the ARMeter.cpp and ARMeter.h files.

The main purpose of the AR classes is to store and possibly pre-process the tag parameters.

---

Main musical concepts

Almost all the AR classes represent a Guido tag. However, the main musical concepts (like notes or voices) are not based on tags but are also represented using AR classes:

• ARMusic: represents the score, i.e. a set of voices,
• ARMusicalVoice: represents a musical voice. Note that AR to AR operations are implemented by the ARMusicalVoice class,
• ARNote: represents notes.

---

Utilities

• ARFactory: this class is in charge of the creation of all the musical elements (voices, notes, tags,...). It is used either by the parser of by the GUIDOFactory API.
• ARMusicalVoiceState: used to maintain the current voice state at a given location when traversing the voice. For example, a voice state contains the list of opened tags at a given voice position.

---

AR to AR operations

AR to AR operations enrich the abstract representation with elements that can be automatically computed from the current descrption (e.g. like bars). They consists also in a set of logical layout operations (e.g. compute the beam direction when not set). Theses operations are implemented in ARMusicalVoice by a set of methods named doAuto...:

• doAutoKeys(): traverses the voice and look for key changes. Whenever a key-change occurs, the previous key needs to be 'naturalized',
• doAutoDispatchLyrics(): converts lyrics tags into individual text-tags for all events within the range,
• doAutoTies(): traverses the voice and split ties so that there are always TWO events within a tie-Range,
• doAutoCheckStaffStateTags(): makes sure that clef/key/meter-information is set,
• doAutoDisplayCheck(): checks whether the events in the voice can be displayed as single graphical objects,
• doAutoBarlines(): add automatic barlines, breaks notes if needed,
• doAutoMeasuresNumbering(): add mesaures numbering,
• doAutoEndBar(): add end bar (unless prevented with an auto tag),
• doAutoBeaming(): add automatic beaming tags,
• doAutoGlissando(): post-processing for the glissando tags,
• doAutoFeatheredBeam(): post-processing for the fbeam tags.