"""
# Display Settings.

Display options are used by
`tf.advanced.display.plain`
and
`tf.advanced.display.pretty`
and other display functions.

This class manages

*   the provisioning of options with defaults,
*   the overriding options by passing options as arguments to display functions
*   the retrieval of option values by the rest of the application.

!!! note "distinction between interface options and display options"
    *   all interface options are also display options and can be passed as arguments
        to display functions
    *   interface options have a checkbox in the TF browser

Parameters
----------
baseTypes: string | iterable, optional None
    **interface option**
    Node types at the bottom of pretty displays.
    The default is app dependent, usually the slot type of the corpus.

colorMap: dict, optional None
    Which nodes of a tuple (or list of tuples) will be highlighted.
    If `colorMap` is `None` or missing, all nodes will be highlighted with
    the default highlight color, which is yellow.

    But you can assign different colours to the members of the tuple:
    `colorMap` must be a dictionary that maps the positions in a tuple
    to a color.

    *   If a position is not mapped, it will not be highlighted.
    *   If it is mapped to the empty string, it gets the default highlight color.
    *   Otherwise, it should be mapped to a string that is a valid
        [CSS color](https://developer.mozilla.org/en-US/docs/Web/CSS/color_value).

    !!! hint "color names"
        The link above points to a series of handy color names and their previews.

    !!! note "`highlights` takes precedence over `colorMap`"
        If both `highlights` and `colorMap` are given, `colorMap` is ignored.

        If you need to micro-manage, `highlights` is your thing.
        Whenever possible, use `colorMap`.

condensed: boolean, optional False
    **interface option**
    indicates one of two modes of displaying the result list:

    *   `True`: instead of showing all results one by one,
        we show container nodes with all results in it highlighted.
        That way, we blur the distinction between the individual results,
        but it is easier to oversee where the results are.
        This is how SHEBANQ displays its query results.
        **See also the parameter `condenseType`**.
    *   `False`: make a separate display for each result tuple.
        This gives the best account of the exact result set.

    !!! caution "mixing up highlights"
        Condensing may mix-up the highlight colouring.
        If a node occurs in two results, at different positions
        in the tuple, the `colorMap` wants to assign it two colours!
        Yet one color will be chosen, and it is unpredictable which one.

condenseType: string, optional None
    **interface option**
    The type of container to be used for condensing results.
    The default is app dependent, usually `verse` or `tablet`.

edgeFeatures: string | iterable, optional None
    **interface option**
    Edge features that will be shown on the interface, in pretty displays.
    The selection of edge features to show will be in effect
    if `forceEdges` is `True`.
    Otherwise, the display of edge features depends on whether it is in the
    `extraFeatures` or `tupleFeatures` (which will be set after a query).

    The default is app dependent, it consist of all the edge features in the
    dataset, except `oslots`.

edgeHighlights: dict
    Custom highlighting of edges.

    The keys are the names of edge features. For each such key, the value is
    a set or mapping of edges defined by that edge feature that should be highlighted.
    Only edges that are involved in the display will be highlighted.

    If the value is a set, the specified edges will be highlighted
    with a default color (blue).

    If it is a dictionary, it should map edges to colours.
    Any color that is a valid
    [CSS color](https://developer.mozilla.org/en-US/docs/Web/CSS/color_value)
    qualifies.

    If you map an edge to the empty string, it will get the default highlight color.

    An edge is a pair of nodes, represented as a tuple of two node numbers.

    You can also map the following pseudo edges to colours:

    *   `(fromNode, None)`: all outgoing edges of `fromNode`;
    *   `(None, toNode)`: all incoming edges of `toNode`.

    There are no edge highlights in plain display.

end: integer, optional None
    `end` is the end point in the iterable of results.
    If `None`, displaying will stop after the end of the iterable.

extraFeatures: string | iterable, optional ()
    A string or iterable of (node- / edge-) feature names.
    These features will be loaded automatically.
    In pretty displays these features will show up as `feature=value`,
    provided the value is not `None`, or something like None.

    !!! hint "Automatic loading"
        These features will load automatically, no explicit loading is
        necessary.

    !!! hint "values from other nodes"
        Suppose you want to display a value from a related node, e.g. a `gloss`
        that is available on `lex` nodes but not on `word` nodes, and you
        want to show it on the word nodes.
        Then you may specify `lex:gloss`, meaning that TF will
        look up a `lex` node from the current node (by means of `L.u(w, otype='lex')`,
        and if it finds one, it will read the `gloss` feature from it.

fmt: string, optional None
    **interface option**
    `fmt` is the text format that will be used for the representation.
    E.g. `text-orig-full`.

    !!! hint "Text formats"
        Use `T.formats` to inspect what text formats are available in your corpus.

full: boolean, optional False
    For pretty displays: indicates that the whole object should be
    displayed, even if it is big.

    !!! hint "Big objects"
        Big objects are objects of a type that is bigger than the default condense type.

hiddenTypes: string | iterable, optional None
    **interface option**
    Node types that will not be shown in displays.
    All node types can be hidden, except the slot type and the section types.
    Structure types can be hidden.

    !!! hint "Meaning"
        Nodes of hidden types will not be skipped, but they do not add visible
        structure to the display. The material under those nodes will still
        be displayed. For example, if the corpus has verses divided into half verses,
        and you are not interested in the half verse division, you can make
        half verses hidden. The content of the half verses is still shown,
        but the half verse division is gone.

    The default is app dependent, usually the empty set.

showEdges: boolean | optional True
    **interface option**
    If `True`, edges are shown in pretty displays, provided they are also selected
    in `edgeFeatures`.
    If `False` all edges are hidden.

hideTypes: boolean | optional True
    **interface option**
    If `True`, hidden types are in fact hidden, otherwise the hiding of types
    has no effect.

highlights: dict | set, optional {}
    When nodes such as verses and sentences and lines and cases are displayed
    by `plain()` or `pretty()`,
    their contents is also displayed. You can selectively highlight
    those parts.

    `highlights={}` is a set or mapping of nodes that should be highlighted.
    Only nodes that are involved in the display will be highlighted.

    If `highlights` is a set, its nodes will be highlighted
    with a default color (yellow).

    If it is a dictionary, it should map nodes to colours.
    Any color that is a valid
    [CSS color](https://developer.mozilla.org/en-US/docs/Web/CSS/color_value)
    qualifies.

    If you map a node to the empty string, it will get the default highlight color.

    Highlights in plain display will be done, also for nodes deeply buried in the top node.
    Slots are highlighted by colouring the background,
    all other nodes by coloured frames around their material.

    !!! note "one big highlights dictionary"
        It is OK to first compose a big highlights dictionary
        for many tuples of nodes,
        and then run `prettyTuple()` for many different tuples
        with the same `highlights`.
        It does not harm performance if `highlights` maps
        lots of nodes outside the tuple as well.

lineNumbers: boolean, optional False
    **interface option**
    indicates whether line numbers should be displayed.

    !!! note "source data"
        Line numbers are with respect to the source data file that is contains the
        origin material of the node in question, if a data source provides
        a feature that contains line numbers.

    !!! caution "configuration"
        Whether a corpus has line numbers, and in which feature they are stored
        for which node types is configured in a corpus dependent app.

        If the corpus has no line numbers, the default is `None`.

multiFeatures: boolean, optional False
    **interface option**
    indicates whether pretty displays should show all possible node features
    provided they have a non-empty value. The feature `otype` is excluded from this.

noneValues: set, optional None
    A set of values for which no display should be generated.
    The default set is `None` and the strings `NA`, `none`, `unknown`.

    !!! hint "None is useful"
        Keep `None` in the set. If not, all custom features will be displayed
        for all kinds of nodes. So you will see clause types on words,
        and part of speech on clause atoms, all with value `None`.

    !!! hint "Suppress common values"
        You can use `noneValues` also to suppress the normal values of a feature,
        in order to attract attention to the more special values, e.g.

            noneValues={None, 'NA', 'unknown', 'm', 'sg', 'p3'}

    !!! caution "None values affect all features"
        Beware of putting to much in `noneValues`.
        The contents of `noneValues` affect the display of
        all features, not only the custom features.

plainGaps: boolean, optional True
    **interface option**
    indicates whether gaps types should be displayed in plain displays.
    In pretty displays gaps are marked by dotted left-right borders of the nodes
    around the gaps. In plain displays such borders are generally disruptive,
    but it is possible to show them.

prettyTypes: boolean, optional False
    **interface option**
    indicates whether node types should always be displayed in pretty displays.
    The node type of slot nodes is never displayed.

queryFeatures: boolean, optional True
    **interface option**
    indicates whether pretty displays should show the features
    mentioned in the last query and their values.

showGraphics: boolean, optional True
    **interface option**
    indicates whether plain and pretty displays should include associated
    graphic elements.

    !!! caution "configuration"
        Whether a corpus has graphics for some node types and how to get them is
        configured in a corpus dependent app.

        If the corpus has no graphics, the default is `None`.

showMath: boolean, optional False
    **interface option**
    indicates whether plain and pretty displays should interpret the `$`
    to mark mathematical formulas in TeX notation.

    !!! caution "real dollars"
        If you have a corpus where the character `$` is used in an ordinary way,
        this option should be set to `False`.

skipCols: set, optional set()
    indicates columns to skip in `show()`, `table()`;
    no effect on `prettyTuple()` and `plainTuple()`.
    Also no effect if `condensed` is True.

    The columns are not really removed, but the cells in the columns
    become empty.

    So it will not disturb the highlighting of the tuples involved,
    and it does not disturb the working of `colorMap`.

    The value may be a space-separated string of numbers, or an iterable of integers.
    Columns start at 1.

standardFeatures: boolean, optional True
    **interface option**
    indicates whether pretty displays should show standard features and their values.
    If a feature is both a standard feature and it is mentioned in the last query,
    it will still be shown if `standardFeatures` is off and `queryFeatures` is on.

start: integer, optional None
    `start` is the starting point for displaying the iterable of results.
    (1 is the first one).
    If `None`, displaying starts at the first element of the iterable.

suppress: set, optional set()
    a set of names of features that should NOT be displayed.
    By default, quite a number of features is displayed for a node.
    If you find they clutter the display, you can turn them off
    selectively.

tupleFeatures: iterable of 2-tuples, optional ()
    A bit like `extraFeatures` above, but more intricate.

    In the first place, after running a query by means of `A.search()`,
    an automatic `displaySetup(tupleFeatures=...)` will be performed with
    the features mentioned in the query for node atoms.

    This will influence the effect of subsequent display and export operations.

    When you set it yourself, it should be a tuple of pairs

        (i, features)

    which means that to member `i` of a result tuple we assign extra `features`.

    `features` may be given as an iterable or a space separated string of feature names.

withPassage: boolean or set, optional True
    indicates whether a passage label should be put next to a displayed node
    or tuple of nodes.
    When passed with `table()`, or `plainTuple()`,
    the value may also be a set of integers, indicating the columns whose
    nodes will be linked with a web link
    (the first column is column 1).

withLabels: boolean, optional True
    **interface option**
    indicates whether pretty displays should show labels.

withNodes: boolean, optional False
    **interface option**
    indicates whether node numbers should be displayed.

    !!! hint "zooming in"
        If you are in a Jupyter notebook, you can inspect in a powerful way by
        setting `withNodes=True`. Then every part of a pretty display shows
        its node number, and you can use the following APIs
        to look up all information
        about each node that the corpus has to offer:

        *   `F`: `tf.core.nodefeature.NodeFeature`
        *   `E`: `tf.core.edgefeature.EdgeFeature`
        *   `L`: `tf.core.locality.Locality`
        *   `T`: `tf.core.text.Text`

withTypes: boolean, optional False
    **interface option**
    indicates whether node types should be displayed.
    The node type of slot nodes is never displayed.
"""

from ..core.helpers import setFromValue
from .helpers import parseFeatures, SEQ_TYPES1, SEQ_TYPES2


INTERFACE_OPTIONS = (
    (
        "condensed",
        False,
        "cond",
        "condense results",
        "Group query results into containers of the selected type.",
        True,
    ),
    (
        "hideTypes",
        True,
        "hidet",
        "hide types",
        "Do not show the outer structure of nodes "
        "of the selected types."
        "The contents of those nodes are still shown.",
        True,
    ),
    (
        "forceEdges",
        False,
        "forcee",
        "show / hide the selected edge features",
        "Show selected edge features in pretty displays, hide the unselected ones. "
        "If this is off, only the edge features that occur in the query are shown.",
        True,
    ),
    (
        "withLabels",
        True,
        "withl",
        "show labels",
        "Show additional labels in nodes.",
        False,
    ),
    (
        "withNodes",
        False,
        "withn",
        "show nodes",
        "Show the node number for every node in the results."
        " The node number is your access to all information about that node."
        " If you click on it, it will be copied to the <i>node pad</i>.",
        False,
    ),
    (
        "withTypes",
        False,
        "witht",
        "show types",
        "Show the node type for every node in the results.",
        False,
    ),
    (
        "prettyTypes",
        True,
        "withtp",
        "always show types when expanded",
        "Show the node type for every node in the expanded view, "
        "even if <b>show types</b> is off.",
        False,
    ),
    (
        "plainGaps",
        True,
        "plaing",
        "show gaps",
        "In plain displays, show the gaps in nodes by means of dotted lines.",
        False,
    ),
    (
        "standardFeatures",
        False,
        "shows",
        "show standard features",
        "Show the standard feature values for every node in the results.",
        False,
    ),
    (
        "queryFeatures",
        True,
        "showf",
        "show query features",
        "Show the features mentioned in the last query for every node in the results.",
        False,
    ),
    (
        "multiFeatures",
        False,
        "showm",
        "show all possible node features (except otype)",
        "Show all possible node features with non-empty values, except otype.",
        False,
    ),
    (
        "lineNumbers",
        False,
        "linen",
        "source lines",
        "Show source line numbers with the nodes."
        " Only if the TF data has a feature for line numbers.",
        False,
    ),
    (
        "showGraphics",
        True,
        "graphics",
        "graphic elements",
        "Show graphical companion elements with the nodes."
        " Only if the data set implements the logic for it.",
        False,
    ),
    (
        "showMath",
        False,
        "math",
        "mathematical formulas",
        "Interpret material within pairs of `$` as formulas in TeX notation.",
        False,
    ),
)
"""Options that can be set on the interface of the TF browser.

Every option is a tuple with members

*   *option*: the name by which you can control this option in API functions;
*   *default*: the value that is used if this option is nowhere explicitly given;
*   *acro*: acronym for this option, used in the HTML as value for an id attribute;
*   *desc*: short description;
*   *long*: long description;
*   *move*: whether to move this option into a separate box in the TF browser.
"""


DISPLAY_OPTIONS = dict(
    baseTypes=None,
    colorMap=None,
    condensed=False,
    condenseType=None,
    edgeFeatures=None,
    edgeHighlights=None,
    end=None,
    extraFeatures=((), {}),
    full=False,
    fmt=None,
    hideTypes=True,
    hiddenTypes=None,
    highlights={},
    noneValues={None},
    forceEdges=False,
    skipCols=set(),
    start=None,
    suppress=set(),
    tupleFeatures=(),
    withPassage=True,
)

DISPLAY_OPTIONS.update({o[0]: o[1] for o in INTERFACE_OPTIONS})


class OptionsCurrent:
    def __init__(self, options):
        self.allKeys = set(options)
        for (k, v) in options.items():
            setattr(self, k, v)

    def get(self, k, v=None):
        return getattr(self, k, v)

    def set(self, k, v):
        self.allKeys.add(k)
        setattr(self, k, v)


class Options:
    def __init__(self, app):
        self.app = app

        aContext = app.context
        interfaceDefaults = aContext.interfaceDefaults

        self.defaults = {}
        defaults = self.defaults

        for (k, v) in DISPLAY_OPTIONS.items():
            value = (
                interfaceDefaults[k] if k in interfaceDefaults else aContext.get(k, v)
            )
            defaults[k] = value

        self.reset()

    def reset(self, *options):
        app = self.app
        error = app.error
        # info = app.info
        defaults = self.defaults

        if options:
            current = self.current
            for option in options:
                if option not in defaults:
                    # info(defaults)
                    error(f'WARNING: unknown display option "{option}" will be ignored')
                    continue
                current[option] = defaults[option]
        else:
            self.current = {k: v for (k, v) in defaults.items()}

    def setup(self, *options, **overrides):
        app = self.app
        info = app.info
        current = self.current

        for option in options:
            info(f"{option} = {current.get(option, None)}", tm=False)

        for (option, value) in overrides.items():
            normValue = self.normalize(option, value)
            if not normValue:
                continue
            current[option] = normValue[1]

    def normalize(self, option, value):
        app = self.app
        api = app.api
        aContext = app.context
        allowedValues = aContext.allowedValues
        error = app.error
        defaults = self.defaults

        if option not in defaults:
            error(f'WARNING: unknown display option "{option}" will be ignored')
            return None

        if option == "extraFeatures":
            (bare, indirect) = parseFeatures(value)
            api.ensureLoaded(bare)
            value = (bare, indirect)
        elif option == "tupleFeatures":
            api.ensureLoaded(value)
            if type(value) is str:
                value = value.split() if value else []
        elif option in {"suppress"}:
            if type(value) is str:
                value = set(value.split()) if value else set()
        elif option in {"skipCols"}:
            if not value:
                value = set()
            elif type(value) is str:
                value = {int(v) for v in value.split()} if value else set()
            elif type(value) not in {set, frozenset}:
                value = set(value)
        elif option in {"withPassage"}:
            if not value:
                value = False
            elif type(value) is str:
                value = {int(v) for v in value.split()} if value else set()
            elif type(value) in {list, tuple, dict}:
                value = set(value)
            elif type(value) is not set:
                value = True
        elif option == "highlights":
            if value is not None and type(value) is not dict:
                value = {m: "" for m in value}
        elif option == "edgeHighlights":
            if value is None:
                value = {}
            for (edge, highlights) in value.items():
                if highlights is not None and type(highlights) is not dict:
                    value[edge] = {m: "" for m in highlights}
        elif option in {"baseTypes", "hiddenTypes", "edgeFeatures"}:
            legalValues = set(allowedValues[option])
            values = setFromValue(value)
            value = {tp for tp in values if tp in legalValues}
        return (True, value)

    def check(self, msg, options):
        app = self.app
        api = app.api
        Fotype = api.F.otype
        aContext = app.context
        allowedValues = aContext.allowedValues
        error = app.error
        current = self.current

        good = True
        for (option, value) in options.items():
            if option not in current:
                error(f'ERROR in {msg}(): unknown display option "{option}={value}"')
                good = False
            if option in {"baseTypes", "condenseType", "hiddenTypes", "edgeFeatures"}:
                legalValues = set(allowedValues[option])
                if value is not None:
                    if option in {"baseTypes", "hiddenTypes", "edgeFeatures"}:
                        testVal = setFromValue(value)
                        isLegal = all(v in legalValues for v in testVal)
                    else:
                        isLegal = value in legalValues
                    if not isLegal:
                        thing = (
                            "edge feature" if option == "edgeFeatures" else "node type"
                        )
                        error(
                            f'ERROR in {msg}(): illegal {thing} in "{option}={value}"'
                        )
                        legalRep = ", ".join(sorted(legalValues))
                        error(f"Legal values are: {legalRep}")
                        good = False
            elif option == "extraFeatures":
                if value is not None:
                    if (
                        type(value) in SEQ_TYPES1
                        and len(value) == 2
                        and type(value[0]) in SEQ_TYPES2
                        and type(value[1]) is dict
                    ):
                        indirect = value[1]
                        legalValues = set(Fotype.all)
                        isLegal = all(v in legalValues for v in indirect)
                        if not isLegal:
                            error(
                                f"ERROR in {msg}(): illegal node type in"
                                f' "{option}={value}"'
                            )
                            good = False
        return good

    def distill(self, options):
        defaults = self.defaults
        current = self.current

        normOptions = {}

        for option in defaults:
            value = options.get(option, current.get(option, defaults[option]))
            normValue = self.normalize(option, value)
            if normValue:
                normOptions[option] = normValue[1]

        return OptionsCurrent(normOptions)

    def consume(self, options, *remove):
        return {o: options[o] for o in options if o not in remove}