kernpy 0.0.1__py3-none-any.whl → 1.0.0__py3-none-any.whl

kernpy/core/document.py

@@ -0,0 +1,965 @@
+from __future__ import annotations
+
+from copy import copy, deepcopy
+from collections import deque, defaultdict
+from abc import ABC, abstractmethod
+from enum import Enum
+from typing import List, Optional, Dict, Union
+from collections.abc import Sequence
+from queue import Queue
+
+from kernpy.core import TokenCategory, CORE_HEADERS, TERMINATOR
+from kernpy.core import MetacommentToken, AbstractToken, HeaderToken
+from .transposer import transpose, Direction, NotationEncoding, AVAILABLE_INTERVALS
+from .tokens import NoteRestToken, Subtoken
+from .transposer import IntervalsByName
+
+
+class SignatureNodes:
+    """
+    SignatureNodes class.
+
+    This class is used to store the last signature nodes of a tree.
+    It is used to keep track of the last signature nodes.
+
+    Attributes: nodes (dict): A dictionary that stores the last signature nodes. This way, we can add several tokens
+    without repetitions. - The key is the signature descendant token class (KeyToken, MeterSymbolToken, etc...) - The
+    value = node
+
+    """
+
+    def __init__(self):
+        """
+        Create an instance of SignatureNodes. Initialize the nodes as an empty dictionary.
+
+        Examples:
+            >>> signature_nodes = SignatureNodes()
+            >>> signature_nodes.nodes
+            {}
+        """
+        self.nodes = {}
+
+    def clone(self):
+        """
+        Create a deep copy of the SignatureNodes instance.
+        Returns: A new instance of SignatureNodes with nodes copied.
+
+        # TODO: This method is equivalent to the following code:
+        # from copy import deepcopy
+        # signature_nodes_to_copy = SignatureNodes()
+        # ...
+        # result = deepcopy(signature_nodes_to_copy)
+        # It should be tested.
+        """
+        result = SignatureNodes()
+        result.nodes = copy(self.nodes)
+        return result
+
+    def update(self, node):
+        self.nodes[node.token.__class__.__name__] = node
+
+
+class TreeTraversalInterface(ABC):
+    """
+    TreeTraversalInterface class.
+
+    This class is used to traverse the tree. The `TreeTraversalInterface` class is responsible for implementing
+    the `visit` method.
+    """
+
+    @abstractmethod
+    def visit(self, node):
+        pass
+
+
+class Node:
+    """
+    Node class.
+
+    This class represents a node in a tree.
+    The `Node` class is responsible for storing the main information of the **kern file.
+
+    Attributes:
+        id(int): The unique id of the node.
+        token(Optional[AbstractToken]): The specific token of the node. The token can be a `KeyToken`, `MeterSymbolToken`, etc...
+        parent(Optional['Node']): A reference to the parent `Node`. If the parent is the root, the parent is None.
+        children(List['Node']): A list of the children `Node`.
+        stage(int): The stage of the node in the tree. The stage is similar to a row in the **kern file.
+        last_spine_operator_node(Optional['Node']): The last spine operator node.
+        last_signature_nodes(Optional[SignatureNodes]): A reference to the last `SignatureNodes` instance.
+        header_node(Optional['Node']): The header node.
+    """
+    NextID = 1  # static counter
+
+    def __init__(self,
+                 stage: int,
+                 token: Optional[AbstractToken],
+                 parent: Optional['Node'],
+                 last_spine_operator_node: Optional['Node'],
+                 last_signature_nodes: Optional[SignatureNodes],
+                 header_node: Optional['Node']
+                 ):
+        """
+        Create an instance of Node.
+
+        Args:
+            stage (int): The stage of the node in the tree. The stage is similar to a row in the **kern file.
+            token (Optional[AbstractToken]): The specific token of the node. The token can be a `KeyToken`, `MeterSymbolToken`, etc...
+            parent (Optional['Node']): A reference to the parent `Node`. If the parent is the root, the parent is None.
+            last_spine_operator_node (Optional['Node']): The last spine operator node.
+            last_signature_nodes (Optional[SignatureNodes]): A reference to the last `SignatureNodes` instance.
+            header_node (Optional['Node']): The header node.
+        """
+        self.id = Node.NextID
+        Node.NextID += 1
+        self.token = token
+        self.parent = parent
+        self.children = []
+        self.stage = stage
+        self.header_node = header_node
+        if last_signature_nodes is not None:
+            self.last_signature_nodes = last_signature_nodes.clone()  #TODO Documentar todo esto - composición
+            # self.last_signature_nodes = copy.deepcopy(last_signature_nodes) # TODO: Ver en SignatureNodes.clone
+        else:
+            self.last_signature_nodes = SignatureNodes()
+        self.last_spine_operator_node = last_spine_operator_node
+
+    def count_nodes_by_stage(self) -> List[int]:
+        """
+        Count the number of nodes in each stage of the tree.
+
+        Examples:
+            >>> node = Node(0, None, None, None, None, None)
+            >>> ...
+            >>> node.count_nodes_by_stage()
+            [2, 2, 2, 2, 3, 3, 3, 2]
+
+        Returns:
+            List[int]: A list with the number of nodes in each stage of the tree.
+        """
+        level_counts = defaultdict(int)
+        queue = deque([(self, 0)])  # (node, level)
+        # breadth-first search (BFS)
+        while queue:
+            node, level = queue.popleft()
+            level_counts[level] += 1
+            for child in node.children:
+                queue.append((child, level + 1))
+
+        # Convert the level_counts dictionary to a list of counts
+        max_level = max(level_counts.keys())
+        counts_by_level = [level_counts[level] for level in range(max_level + 1)]
+
+        return counts_by_level
+
+    def dfs(self, tree_traversal: TreeTraversalInterface):
+        """
+        Depth-first search (DFS)
+
+        Args:
+            tree_traversal (TreeTraversalInterface): The tree traversal interface. Object used to visit the nodes of the tree.
+        """
+        node = self
+        tree_traversal.visit(node)
+        for child in self.children:
+            child.dfs(tree_traversal)
+
+    def dfs_iterative(self, tree_traversal: TreeTraversalInterface):
+        """
+        Depth-first search (DFS). Iterative version.
+
+        Args:
+            tree_traversal (TreeTraversalInterface): The tree traversal interface. Object used to visit the nodes of the tree.
+
+        Returns: None
+        """
+        stack = [self]
+        while stack:
+            node = stack.pop()
+            tree_traversal.visit(node)
+            stack.extend(reversed(node.children))  # Add children in reverse order to maintain DFS order
+
+    def __eq__(self, other):
+        """
+        Compare two nodes.
+
+        Args:
+            other: The other node to compare.
+
+        Returns: True if the nodes are equal, False otherwise.
+        """
+        if other is None or not isinstance(other, Node):
+            return False
+
+        return self.id == other.id
+
+    def __ne__(self, other):
+        """
+        Compare two nodes.
+
+        Args:
+            other: The other node to compare.
+
+        Returns: True if the nodes are not equal, False otherwise.
+        """
+        return not self.__eq__(other)
+
+    def __hash__(self):
+        """
+        Get the hash of the node.
+
+        Returns: The hash of the node.
+        """
+        return hash(self.id)
+
+    def __str__(self):
+        """
+        Get the string representation of the node.
+
+        Returns: The string representation of the node.
+        """
+        return f"{{{self.stage}: {self.token}}}"
+
+
+class BoundingBoxMeasures:
+    """
+    BoundingBoxMeasures class.
+    """
+
+    def __init__(
+            self,
+            bounding_box,
+            from_measure: int,
+            to_measure: int
+    ):
+        """
+        Create an instance of BoundingBoxMeasures.
+
+        Args:
+            bounding_box: The bounding box object of the node.
+            from_measure (int): The first measure of the score in the BoundingBoxMeasures object.
+            to_measure (int): The last measure of the score in the BoundingBoxMeasures object.
+        """
+        self.from_measure = from_measure
+        self.to_measure = to_measure
+        self.bounding_box = bounding_box
+
+
+class MultistageTree:
+    """
+    MultistageTree class.
+    """
+
+    def __init__(self):
+        """
+        Constructor for MultistageTree class.
+
+        Create an empty Node object to serve as the root, \
+        and start the stages list by placing this root node inside a new list.
+
+        """
+        self.root = Node(0, None, None, None, None, None)
+        self.stages = []  # First stage (0-index) is the root (Node with None token and header_node). The core header is in stage 1.
+        self.stages.append([self.root])
+
+    def add_node(
+            self,
+            stage: int,
+            parent: Node,
+            token: Optional[AbstractToken],
+            last_spine_operator_node: Optional[Node],
+            previous_signature_nodes: Optional[SignatureNodes],
+            header_node: Optional[Node] = None
+    ) -> Node:
+        """
+        Add a new node to the tree.
+        Args:
+            stage (int):
+            parent (Node):
+            token (Optional[AbstractToken]):
+            last_spine_operator_node (Optional[Node]):
+            previous_signature_nodes (Optional[SignatureNodes]):
+            header_node (Optional[Node]):
+
+        Returns: Node - The added node object.
+
+        """
+        node = Node(stage, token, parent, last_spine_operator_node, previous_signature_nodes, header_node)
+        if stage == len(self.stages):
+            self.stages.append([node])
+        elif stage > len(self.stages):
+            raise ValueError(f'Cannot add node in stage {stage} when there are only {len(self.stages)} stages')
+        else:
+            self.stages[stage].append(node)
+
+        parent.children.append(node)
+        return node
+
+    def dfs(self, visit_method) -> None:
+        """
+        Depth-first search (DFS)
+
+        Args:
+            visit_method (TreeTraversalInterface): The tree traversal interface.
+
+        Returns: None
+
+        """
+        self.root.dfs(visit_method)
+
+    def dfs_iterative(self, visit_method) -> None:
+        """
+        Depth-first search (DFS). Iterative version.
+
+        Args:
+            visit_method (TreeTraversalInterface): The tree traversal interface.
+
+        Returns: None
+
+        """
+        self.root.dfs_iterative(visit_method)
+
+    def __deepcopy__(self, memo):
+        """
+        Create a deep copy of the MultistageTree object.
+        """
+        # Create a new empty MultistageTree object
+        new_tree = MultistageTree()
+
+        # Deepcopy the root
+        new_tree.root = deepcopy(self.root, memo)
+
+        # Deepcopy the stages list
+        new_tree.stages = deepcopy(self.stages, memo)
+
+        return new_tree
+
+
+class Document:
+    """
+    Document class.
+
+    This class store the score content using an agnostic tree structure.
+
+    Attributes:
+        tree (MultistageTree): The tree structure of the document where all the nodes are stored. \
+            Each stage of the tree corresponds to a row in the Humdrum **kern file encoding.
+        measure_start_tree_stages (List[List[Node]]): The list of nodes that corresponds to the measures. \
+            Empty list by default.
+            The index of the list is starting from 1. Rows after removing empty lines and line comments
+        page_bounding_boxes (Dict[int, BoundingBoxMeasures]): The dictionary of page bounding boxes. \
+            - key: page number
+            - value: BoundingBoxMeasures object
+        header_stage (int): The index of the stage that contains the headers. None by default.
+    """
+
+    def __init__(self, tree: MultistageTree):
+        """
+        Constructor for Document class.
+
+        Args:
+            tree (MultistageTree): The tree structure of the document where all the nodes are stored.
+        """
+        self.tree = tree  # TODO: ? Should we use copy.deepcopy() here?
+        self.measure_start_tree_stages = []
+        self.page_bounding_boxes = {}
+        self.header_stage = None
+
+    FIRST_MEASURE = 1
+
+    def get_header_stage(self) -> Union[List[Node], List[List[Node]]]:
+        """
+        Get the Node list of the header stage.
+
+        Returns: (Union[List[Node], List[List[Node]]]) The Node list of the header stage.
+
+        Raises: Exception - If the document has no header stage.
+        """
+        if self.header_stage:
+            return self.tree.stages[self.header_stage]
+        else:
+            raise Exception('No header stage found')
+
+    def get_leaves(self) -> List[Node]:
+        """
+        Get the leaves of the tree.
+
+        Returns: (List[Node]) The leaves of the tree.
+        """
+        return self.tree.stages[len(self.tree.stages) - 1]
+
+    def get_spine_count(self) -> int:
+        """
+        Get the number of spines in the document.
+
+        Returns (int): The number of spines in the document.
+        """
+        return len(self.get_header_stage())  # TODO: test refactor
+
+    def get_first_measure(self) -> int:
+        """
+        Get the index of the first measure of the document.
+
+        Returns: (Int) The index of the first measure of the document.
+
+        Raises: Exception - If the document has no measures.
+
+        Examples:
+            >>> import kernpy as kp
+            >>> document, err = kp.read('score.krn')
+            >>> document.get_first_measure()
+            1
+        """
+        if len(self.measure_start_tree_stages) == 0:
+            raise Exception('No measures found')
+
+        return self.FIRST_MEASURE
+
+    def measures_count(self) -> int:
+        """
+        Get the index of the last measure of the document.
+
+        Returns: (Int) The index of the last measure of the document.
+
+        Raises: Exception - If the document has no measures.
+
+        Examples:
+            >>> document, _ = kernpy.read('score.krn')
+            >>> document.measures_count()
+            10
+            >>> for i in range(document.get_first_measure(), document.measures_count() + 1):
+            >>>   options = kernpy.ExportOptions(from_measure=i, to_measure=i+4)
+        """
+        if len(self.measure_start_tree_stages) == 0:
+            raise Exception('No measures found')
+
+        return len(self.measure_start_tree_stages)
+
+    def get_metacomments(self, KeyComment: Optional[str] = None, clear: bool = False) -> List[str]:
+        """
+        Get all metacomments in the document
+
+        Args:
+            KeyComment: Filter by a specific metacomment key: e.g. Use 'COM' to get only comments starting with\
+                '!!!COM: '. If None, all metacomments are returned.
+            clear: If True, the metacomment key is removed from the comment. E.g. '!!!COM: Coltrane' -> 'Coltrane'.\
+                If False, the metacomment key is kept. E.g. '!!!COM: Coltrane' -> '!!!COM: Coltrane'. \
+                The clear functionality is equivalent to the following code:
+                ```python
+                comment = '!!!COM: Coltrane'
+                clean_comment = comment.replace(f"!!!{KeyComment}: ", "")
+                ```
+                Other formats are not supported.
+
+        Returns: A list of metacomments.
+
+        Examples:
+            >>> document.get_metacomments()
+            ['!!!COM: Coltrane', '!!!voices: 1', '!!!OPR: Blue Train']
+            >>> document.get_metacomments(KeyComment='COM')
+            ['!!!COM: Coltrane']
+            >>> document.get_metacomments(KeyComment='COM', clear=True)
+            ['Coltrane']
+            >>> document.get_metacomments(KeyComment='non_existing_key')
+            []
+        """
+        traversal = MetacommentsTraversal()
+        self.tree.dfs_iterative(traversal)
+        result = []
+        for metacomment in traversal.metacomments:
+            if KeyComment is None or metacomment.encoding.startswith(f"!!!{KeyComment}"):
+                new_comment = metacomment.encoding
+                if clear:
+                    new_comment = metacomment.encoding.replace(f"!!!{KeyComment}: ", "")
+                result.append(new_comment)
+
+        return result
+
+    @classmethod
+    def tokens_to_encodings(cls, tokens: Sequence[AbstractToken]):
+        """
+        Get the encodings of a list of tokens.
+
+        The method is equivalent to the following code:
+            >>> tokens = kp.get_all_tokens()
+            >>> [token.encoding for token in tokens if token.encoding is not None]
+
+        Args:
+            tokens (Sequence[AbstractToken]): list - A list of tokens.
+
+        Returns: List[str] - A list of token encodings.
+
+        Examples:
+            >>> tokens = document.get_all_tokens()
+            >>> Document.tokens_to_encodings(tokens)
+            ['!!!COM: Coltrane', '!!!voices: 1', '!!!OPR: Blue Train']
+        """
+        encodings = [token.encoding for token in tokens if token.encoding is not None]
+        return encodings
+
+    def get_all_tokens(self, filter_by_categories: Optional[Sequence[TokenCategory]] = None) -> List[AbstractToken]:
+        """
+        Args:
+            filter_by_categories (Optional[Sequence[TokenCategory]]): A list of categories to filter the tokens. If None, all tokens are returned.
+
+        Returns:
+            List[AbstractToken] - A list of all tokens.
+
+        Examples:
+            >>> tokens = document.get_all_tokens()
+            >>> Document.tokens_to_encodings(tokens)
+            >>> [type(t) for t in tokens]
+            [<class 'kernpy.core.token.Token'>, <class 'kernpy.core.token.Token'>, <class 'kernpy.core.token.Token'>]
+        """
+        computed_categories = TokenCategory.valid(include=filter_by_categories)
+        traversal = TokensTraversal(False, computed_categories)
+        self.tree.dfs_iterative(traversal)
+        return traversal.tokens
+
+    def get_all_tokens_encodings(
+            self,
+            filter_by_categories: Optional[Sequence[TokenCategory]] = None
+    ) -> List[str]:
+        """
+        Args:
+            filter_by_categories (Optional[Sequence[TokenCategory]]): A list of categories to filter the tokens. If None, all tokens are returned.
+
+
+        Returns:
+            list[str] - A list of all token encodings.
+
+        Examples:
+            >>> tokens = document.get_all_tokens_encodings()
+            >>> Document.tokens_to_encodings(tokens)
+            ['!!!COM: Coltrane', '!!!voices: 1', '!!!OPR: Blue Train']
+        """
+        tokens = self.get_all_tokens(filter_by_categories)
+        return Document.tokens_to_encodings(tokens)
+
+    def get_unique_tokens(
+            self,
+            filter_by_categories: Optional[Sequence[TokenCategory]] = None
+    ) -> List[AbstractToken]:
+        """
+        Get unique tokens.
+
+        Args:
+            filter_by_categories (Optional[Sequence[TokenCategory]]): A list of categories to filter the tokens. If None, all tokens are returned.
+
+        Returns:
+            List[AbstractToken] - A list of unique tokens.
+
+        """
+        computed_categories = TokenCategory.valid(include=filter_by_categories)
+        traversal = TokensTraversal(True, computed_categories)
+        self.tree.dfs_iterative(traversal)
+        return traversal.tokens
+
+    def get_unique_token_encodings(
+            self,
+            filter_by_categories: Optional[Sequence[TokenCategory]] = None
+    ) -> List[str]:
+        """
+        Get unique token encodings.
+
+        Args:
+            filter_by_categories (Optional[Sequence[TokenCategory]]): A list of categories to filter the tokens. If None, all tokens are returned.
+
+        Returns: List[str] - A list of unique token encodings.
+
+        """
+        tokens = self.get_unique_tokens(filter_by_categories)
+        return Document.tokens_to_encodings(tokens)
+
+    def get_voices(self, clean: bool = False):
+        """
+        Get the voices of the document.
+
+        Args
+            clean (bool): Remove the first '!' from the voice name.
+
+        Returns: A list of voices.
+
+        Examples:
+            >>> document.get_voices()
+            ['!sax', '!piano', '!bass']
+            >>> document.get_voices(clean=True)
+            ['sax', 'piano', 'bass']
+            >>> document.get_voices(clean=False)
+            ['!sax', '!piano', '!bass']
+        """
+        from kernpy.core import TokenCategory
+        voices = self.get_all_tokens(filter_by_categories=[TokenCategory.INSTRUMENTS])
+
+        if clean:
+            voices = [voice[1:] for voice in voices]
+        return voices
+
+    def clone(self):
+        """
+        Create a deep copy of the Document instance.
+
+        Returns: A new instance of Document with the tree copied.
+
+        """
+        result = Document(copy(self.tree))
+        result.measure_start_tree_stages = copy(self.measure_start_tree_stages)
+        result.page_bounding_boxes = copy(self.page_bounding_boxes)
+        result.header_stage = copy(self.header_stage)
+
+        return result
+
+    def append_spines(self, spines) -> None:
+        """
+        Append the spines directly to current document tree.
+
+        Args:
+            spines(list): A list of spines to append.
+
+        Returns: None
+
+        Examples:
+            >>> import kernpy as kp
+            >>> doc, _ = kp.read('score.krn')
+            >>> spines = [
+            >>> '4e\t4f\t4g\t4a\n4b\t4c\t4d\t4e\n=\t=\t=\t=\n',
+            >>> '4c\t4d\t4e\t4f\n4g\t4a\t4b\t4c\n=\t=\t=\t=\n',
+           >>> ]
+           >>> doc.append_spines(spines)
+           None
+        """
+        raise NotImplementedError()
+        if len(spines) != self.get_spine_count():
+            raise Exception(f"Spines count mismatch: {len(spines)} != {self.get_spine_count()}")
+
+        for spine in spines:
+            return
+
+    def add(self, other: 'Document', *, check_core_spines_only: Optional[bool] = False) -> 'Document':
+        """
+        Concatenate one document to the current document: Modify the current object!
+
+        Args:
+            other: The document to concatenate.
+            check_core_spines_only: If True, only the core spines (**kern and **mens) are checked. If False, all spines are checked.
+
+        Returns ('Document'): The current document (self) with the other document concatenated.
+        """
+        if not Document.match(self, other, check_core_spines_only=check_core_spines_only):
+            raise Exception(f'Documents are not compatible for addition. '
+                            f'Headers do not match with check_core_spines_only={check_core_spines_only}. '
+                            f'self: {self.get_header_nodes()}, other: {other.get_header_nodes()}. ')
+
+        current_header_nodes = self.get_header_stage()
+        other_header_nodes = other.get_header_stage()
+
+        current_leaf_nodes = self.get_leaves()
+        flatten = lambda lst: [item for sublist in lst for item in sublist]
+        other_first_level_children = [flatten(c.children) for c in other_header_nodes]  # avoid header stage
+
+        for current_leaf, other_first_level_child in zip(current_leaf_nodes, other_first_level_children, strict=False):
+            # Ignore extra spines from other document.
+            # But if there are extra spines in the current document, it will raise an exception.
+            if current_leaf.token.encoding == TERMINATOR:
+                # remove the '*-' token from the current document
+                current_leaf_index = current_leaf.parent.children.index(current_leaf)
+                current_leaf.parent.children.pop(current_leaf_index)
+                current_leaf.parent.children.insert(current_leaf_index, other_first_level_child)
+
+            self.tree.add_node(
+                stage=len(self.tree.stages) - 1,  # TODO: check offset 0, +1, -1 ????
+                parent=current_leaf,
+                token=other_first_level_child.token,
+                last_spine_operator_node=other_first_level_child.last_spine_operator_node,
+                previous_signature_nodes=other_first_level_child.last_signature_nodes,
+                header_node=other_first_level_child.header_node
+            )
+
+        return self
+
+    def get_header_nodes(self) -> List[HeaderToken]:
+        """
+        Get the header nodes of the current document.
+
+        Returns: List[HeaderToken]: A list with the header nodes of the current document.
+        """
+        return [token for token in self.get_all_tokens(filter_by_categories=None) if isinstance(token, HeaderToken)]
+
+    def get_spine_ids(self) -> List[int]:
+        """
+                Get the indexes of the current document.
+
+                Returns List[int]: A list with the indexes of the current document.
+
+                Examples:
+                    >>> document.get_all_spine_indexes()
+                    [0, 1, 2, 3, 4]
+                """
+        header_nodes = self.get_header_nodes()
+        return [node.spine_id for node in header_nodes]
+
+    def frequencies(self, token_categories: Optional[Sequence[TokenCategory]] = None) -> Dict:
+        """
+        Frequency of tokens in the document.
+
+
+        Args:
+            token_categories (Optional[Sequence[TokenCategory]]): If None, all tokens are considered.
+        Returns (Dict):
+            A dictionary with the category and the number of occurrences of each token.
+
+        """
+        tokens = self.get_all_tokens(filter_by_categories=token_categories)
+        frequencies = {}
+        for t in tokens:
+            if t.encoding in frequencies:
+                frequencies[t.encoding]['occurrences'] += 1
+            else:
+                frequencies[t.encoding] = {
+                    'occurrences': 1,
+                    'category': t.category.name,
+                }
+
+        return frequencies
+
+    def split(self) -> List['Document']:
+        """
+        Split the current document into a list of documents, one for each **kern spine.
+        Each resulting document will contain one **kern spine along with all non-kern spines.
+
+        Returns:
+            List['Document']: A list of documents, where each document contains one **kern spine
+            and all non-kern spines from the original document.
+
+        Examples:
+            >>> document.split()
+            [<Document: score.krn>, <Document: score.krn>, <Document: score.krn>]
+        """
+        raise NotImplementedError
+        new_documents = []
+        self_document_copy = deepcopy(self)
+        kern_header_nodes = [node for node in self_document_copy.get_header_nodes() if node.encoding == '**kern']
+        other_header_nodes = [node for node in self_document_copy.get_header_nodes() if node.encoding != '**kern']
+        spine_ids = self_document_copy.get_spine_ids()
+
+        for header_node in kern_header_nodes:
+            if header_node.spine_id not in spine_ids:
+                continue
+
+            spine_ids.remove(header_node.spine_id)
+
+            new_tree = deepcopy(self.tree)
+            prev_node = new_tree.root
+            while not isinstance(prev_node, HeaderToken):
+                prev_node = prev_node.children[0]
+
+            if not prev_node or not isinstance(prev_node, HeaderToken):
+                raise Exception(f'Header node not found: {prev_node} in {header_node}')
+
+            new_children = list(filter(lambda x: x.spine_id == header_node.spine_id, prev_node.children))
+            new_tree.root = new_children
+
+            new_document = Document(new_tree)
+
+            new_documents.append(new_document)
+
+        return new_documents
+
+    @classmethod
+    def to_concat(cls, first_doc: 'Document', second_doc: 'Document', deep_copy: bool = True) -> 'Document':
+        """
+        Concatenate two documents.
+
+        Args:
+            first_doc (Document): The first document.
+            second_doc (Document: The second document.
+            deep_copy (bool): If True, the documents are deep copied. If False, the documents are shallow copied.
+
+        Returns: A new instance of Document with the documents concatenated.
+        """
+        first_doc = first_doc.clone() if deep_copy else first_doc
+        second_doc = second_doc.clone() if deep_copy else second_doc
+        first_doc.add(second_doc)
+
+        return first_doc
+
+    @classmethod
+    def match(cls, a: 'Document', b: 'Document', *, check_core_spines_only: Optional[bool] = False) -> bool:
+        """
+        Match two documents. Two documents match if they have the same spine structure.
+
+        Args:
+            a (Document): The first document.
+            b (Document): The second document.
+            check_core_spines_only (Optional[bool]): If True, only the core spines (**kern and **mens) are checked. If False, all spines are checked.
+
+        Returns: True if the documents match, False otherwise.
+
+        Examples:
+
+        """
+        if check_core_spines_only:
+            return [token.encoding for token in a.get_header_nodes() if token.encoding in CORE_HEADERS] \
+                == [token.encoding for token in b.get_header_nodes() if token.encoding in CORE_HEADERS]
+        else:
+            return [token.encoding for token in a.get_header_nodes()] \
+                == [token.encoding for token in b.get_header_nodes()]
+
+
+    def to_transposed(self, interval: str, direction: str = Direction.UP.value) -> 'Document':
+        """
+        Create a new document with the transposed notes without modifying the original document.
+
+        Args:
+            interval (str): The name of the interval to transpose. It can be 'P4', 'P5', 'M2', etc. Check the \
+             kp.AVAILABLE_INTERVALS for the available intervals.
+            direction (str): The direction to transpose. It can be 'up' or 'down'.
+
+        Returns:
+
+        """
+        if interval not in AVAILABLE_INTERVALS:
+            raise ValueError(
+                f"Interval {interval!r} is not available. "
+                f"Available intervals are: {AVAILABLE_INTERVALS}"
+            )
+
+        if direction not in (Direction.UP.value, Direction.DOWN.value):
+            raise ValueError(
+                f"Direction {direction!r} is not available. "
+                f"Available directions are: "
+                f"{Direction.UP.value!r}, {Direction.DOWN.value!r}"
+            )
+
+        new_document = self.clone()
+
+        # BFS through the tree
+        root = new_document.tree.root
+        queue = Queue()
+        queue.put(root)
+
+        while not queue.empty():
+            node = queue.get()
+
+            if isinstance(node.token, NoteRestToken):
+                orig_token = node.token
+
+                new_subtokens = []
+                transposed_pitch_encoding = None
+
+                # Transpose each pitch subtoken in the pitch–duration list
+                for subtoken in orig_token.pitch_duration_subtokens:
+                    if subtoken.category == TokenCategory.PITCH:
+                        # transpose() returns a new pitch subtoken
+                        tp = transpose(
+                            input_encoding=subtoken.encoding,
+                            interval=IntervalsByName[interval],
+                            direction=direction,
+                            input_format=NotationEncoding.HUMDRUM.value,
+                            output_format=NotationEncoding.HUMDRUM.value,
+                        )
+                        new_subtokens.append(Subtoken(tp, subtoken.category))
+                        transposed_pitch_encoding = tp
+                    else:
+                        # leave duration subtokens untouched
+                        new_subtokens.append(Subtoken(subtoken.encoding, subtoken.category))
+
+                # Replace the node’s token with a new NoteRestToken
+                node.token = NoteRestToken(
+                    encoding=transposed_pitch_encoding,
+                    pitch_duration_subtokens=new_subtokens,
+                    decoration_subtokens=orig_token.decoration_subtokens,
+                )
+
+            # enqueue children
+            for child in node.children:
+                queue.put(child)
+
+        # Return the transposed clone
+        return new_document
+
+
+    def __iter__(self):
+        """
+        Get the indexes to export all the document.
+
+        Returns: An iterator with the indexes to export the document.
+        """
+        return iter(range(self.get_first_measure(), self.measures_count() + 1))
+
+    def __next__(self):
+        """
+        Get the next index to export the document.
+
+        Returns: The next index to export the document.
+        """
+        return next(iter(range(self.get_first_measure(), self.measures_count() + 1)))
+
+
+# tree traversal utils
+class MetacommentsTraversal(TreeTraversalInterface):
+    def __init__(self):
+        self.metacomments = []
+
+    def visit(self, node):
+        if isinstance(node.token, MetacommentToken):
+            self.metacomments.append(node.token)
+
+
+class TokensTraversal(TreeTraversalInterface):
+    def __init__(
+            self,
+            non_repeated: bool,
+            filter_by_categories
+    ):
+        """
+        Create an instance of `TokensTraversal`.
+        Args:
+            non_repeated: If True, only unique tokens are returned. If False, all tokens are returned.
+            filter_by_categories: A list of categories to filter the tokens. If None, all tokens are returned.
+        """
+        self.tokens = []
+        self.seen_encodings = []
+        self.non_repeated = non_repeated
+        self.filter_by_categories = [t for t in TokenCategory] if filter_by_categories is None else filter_by_categories
+
+    def visit(self, node):
+        if (node.token
+                and (not self.non_repeated or node.token.encoding not in self.seen_encodings)
+                and (self.filter_by_categories is None or node.token.category in self.filter_by_categories)
+        ):
+            self.tokens.append(node.token)
+            if self.non_repeated:
+                self.seen_encodings.append(node.token.encoding)
+
+
+class TraversalFactory:
+    class Categories(Enum):
+        METACOMMENTS = "metacomments"
+        TOKENS = "tokens"
+
+    @classmethod
+    def create(
+            cls,
+            traversal_type: str,
+            non_repeated: bool,
+            filter_by_categories: Optional[Sequence[TokenCategory]]
+    ) -> TreeTraversalInterface:
+        """
+        Create an instance of `TreeTraversalInterface` based on the `traversal_type`.
+        Args:
+            non_repeated:
+            filter_by_categories:
+            traversal_type: The type of traversal to use. Possible values are:
+                - "metacomments"
+                - "tokens"
+
+        Returns: An instance of `TreeTraversalInterface`.
+        """
+        if traversal_type == cls.Categories.METACOMMENTS.value:
+            return MetacommentsTraversal()
+        elif traversal_type == cls.Categories.TOKENS.value:
+            return TokensTraversal(non_repeated, filter_by_categories)
+
+        raise ValueError(f"Unknown traversal type: {traversal_type}")
+