コード例 #1
0
ファイル: translate.py プロジェクト: BatataWada1/NOP-commerce
    def translate_formula(self, dest=None, row=None, col=None):
        """
        Convert the formula into A1 notation, or as row and column coordinates

        The formula is converted into A1 assuming it is assigned to the cell
        whose address is `dest` (no worksheet name).

        """
        if not any([dest, row, col]):
            raise TypeError("You must provide coordinates for the target")
        tokens = self.get_tokens()
        if not tokens:
            return ""
        elif tokens[0].type == Token.LITERAL:
            return tokens[0].value
        out = ['=']
        # per the spec:
        # A compliant producer or consumer considers a defined name in the
        # range A1-XFD1048576 to be an error. All other names outside this
        # range can be defined as names and overrides a cell reference if an
        # ambiguity exists. (I.18.2.5)
        if dest:
            row, col = coordinate_to_tuple(dest)
        row_delta = row - self.row
        col_delta = col - self.col
        for token in tokens:
            if (token.type == Token.OPERAND and token.subtype == Token.RANGE):
                out.append(
                    self.translate_range(token.value, row_delta, col_delta))
            else:
                out.append(token.value)
        return "".join(out)
コード例 #2
0
    def freeze_panes(self, topLeftCell=None):
        if isinstance(topLeftCell, Cell):
            topLeftCell = topLeftCell.coordinate
        if topLeftCell == 'A1':
            topLeftCell = None

        if not topLeftCell:
            self.sheet_view.pane = None
            return

        row, column = coordinate_to_tuple(topLeftCell)

        view = self.sheet_view
        view.pane = Pane(topLeftCell=topLeftCell,
                         activePane="topRight",
                         state="frozen")
        view.selection[0].pane = "topRight"

        if column > 1:
            view.pane.xSplit = column - 1
        if row > 1:
            view.pane.ySplit = row - 1
            view.pane.activePane = 'bottomLeft'
            view.selection[0].pane = "bottomLeft"
            if column > 1:
                view.selection[0].pane = "bottomRight"
                view.pane.activePane = 'bottomRight'

        if row > 1 and column > 1:
            sel = list(view.selection)
            sel.insert(0,
                       Selection(pane="topRight", activeCell=None, sqref=None))
            sel.insert(
                1, Selection(pane="bottomLeft", activeCell=None, sqref=None))
            view.selection = sel
コード例 #3
0
def collapse_cell_addresses(cells, input_ranges=()):
    """ Collapse a collection of cell co-ordinates down into an optimal
        range or collection of ranges.

        E.g. Cells A1, A2, A3, B1, B2 and B3 should have the data-validation
        object applied, attempt to collapse down to a single range, A1:B3.

        Currently only collapsing contiguous vertical ranges (i.e. above
        example results in A1:A3 B1:B3).
    """

    ranges = list(input_ranges)

    # convert cell into row, col tuple
    raw_coords = (coordinate_to_tuple(cell) for cell in cells)

    # group by column in order
    grouped_coords = defaultdict(list)
    for row, col in sorted(raw_coords, key=itemgetter(1)):
        grouped_coords[col].append(row)

    # create range string from first and last row in column
    for col, cells in grouped_coords.items():
        col = get_column_letter(col)
        fmt = "{0}{1}:{2}{3}"
        if len(cells) == 1:
            fmt = "{0}{1}"
        r = fmt.format(col, min(cells), col, max(cells))
        ranges.append(r)

    return " ".join(ranges)
コード例 #4
0
ファイル: read_only.py プロジェクト: BatataWada1/NOP-commerce
    def _get_row(self, element, min_col=1, max_col=None, row_counter=None):
        """Return cells from a particular row"""
        col_counter = min_col
        data_only = getattr(self.parent, 'data_only', False)

        for cell in safe_iterator(element, CELL_TAG):
            coordinate = cell.get('r')
            if coordinate:
                row, column = coordinate_to_tuple(coordinate)
            else:
                row, column = row_counter, col_counter

            if max_col is not None and column > max_col:
                break

            if min_col <= column:
                if col_counter < column:
                    for col_counter in range(max(col_counter, min_col), column):
                        # pad row with missing cells
                        yield EMPTY_CELL

                data_type = cell.get('t', 'n')
                style_id = int(cell.get('s', 0))
                value = None
                cached_value = None
                formula = cell.findtext(FORMULA_TAG)
                if formula is not None and not data_only:
                    data_type = 'f'
                    value = "=%s" % formula
                    cached_value = cell.findtext(VALUE_TAG) or None
                elif data_type == 'inlineStr':
                    child = cell.find(INLINE_TAG)
                    if child is not None:
                        richtext = Text.from_tree(child)
                        value = richtext.content

                else:
                    value = cell.findtext(VALUE_TAG) or None

                yield ReadOnlyCell(self, row, column,
                                   value, data_type, style_id, cached_value)
            col_counter = column + 1

        if max_col is not None:
            for _ in range(max(min_col, col_counter), max_col+1):
                yield EMPTY_CELL
コード例 #5
0
def _check_anchor(obj):
    """
    Check whether an object has an existing Anchor object
    If not create a OneCellAnchor using the provided coordinate
    """
    anchor = obj.anchor
    if not isinstance(anchor, _AnchorBase):
        row, col = coordinate_to_tuple(anchor)
        anchor = OneCellAnchor()
        anchor._from.row = row - 1
        anchor._from.col = col - 1
        if isinstance(obj, ChartBase):
            anchor.ext.width = cm_to_EMU(obj.width)
            anchor.ext.height = cm_to_EMU(obj.height)
        elif isinstance(obj, Image):
            anchor.ext.width = pixels_to_EMU(obj.width)
            anchor.ext.height = pixels_to_EMU(obj.height)
    return anchor
コード例 #6
0
    def parse_cell(self, element):
        value = element.find(self.VALUE_TAG)
        cached_value = None
        if value is not None:
            value = value.text
        formula = element.find(self.FORMULA_TAG)
        data_type = element.get('t', 'n')
        coordinate = element.get('r')
        self._col_count += 1
        style_id = element.get('s')

        # assign formula to cell value unless only the data is desired
        if formula is not None and not self.data_only:
            data_type = 'f'
            cached_value = value

            if formula.text:
                value = "=" + formula.text
            else:
                value = "="

            formula_type = formula.get('t')
            if formula_type:
                if formula_type != "shared":
                    self.ws.formula_attributes[coordinate] = dict(formula.attrib)
                else:
                    si = formula.get('si')  # Shared group index for shared formulas

                    # The spec (18.3.1.40) defines shared formulae in
                    # terms of the following:
                    #
                    # `master`: "The first formula in a group of shared
                    #            formulas"
                    # `ref`: "Range of cells which the formula applies
                    #        to." It's a required attribute on the master
                    #        cell, forbidden otherwise.
                    # `shared cell`: "A cell is shared only when si is
                    #                 used and t is `shared`."
                    #
                    # Whether to use the cell's given formula or the
                    # master's depends on whether the cell is shared,
                    # whether it's in the ref, and whether it defines its
                    # own formula, as follows:
                    #
                    #  Shared?   Has formula? | In ref    Not in ref
                    # ========= ==============|======== ===============
                    #   Yes          Yes      | master   impl. defined
                    #    No          Yes      |  own         own
                    #   Yes           No      | master   impl. defined
                    #    No           No      |  ??          N/A
                    #
                    # The ?? is because the spec is silent on this issue,
                    # though my inference is that the cell does not
                    # receive a formula at all.
                    #
                    # For this implementation, we are using the master
                    # formula in the two "impl. defined" cases and no
                    # formula in the "??" case. This choice of
                    # implementation allows us to disregard the `ref`
                    # parameter altogether, and does not require
                    # computing expressions like `C5 in A1:D6`.
                    # Presumably, Excel does not generate spreadsheets
                    # with such contradictions.
                    if si in self.shared_formula_masters:
                        trans = self.shared_formula_masters[si]
                        value = trans.translate_formula(coordinate)
                    else:
                        self.shared_formula_masters[si] = Translator(value, coordinate)


        style_array = None
        if style_id is not None:
            style_id = int(style_id)
            style_array = self.styles[style_id]

        if coordinate:
            row, column = coordinate_to_tuple(coordinate)
        else:
            row, column = self._row_count, self._col_count

        cell = Cell(self.ws, row=row, col_idx=column, style_array=style_array)
        self.ws._cells[(row, column)] = cell

        if value is not None:
            if data_type == 'n':
                value = _cast_number(value)
                if is_date_format(cell.number_format):
                    data_type = 'd'
                    value = from_excel(value, self.epoch)

            elif data_type == 'b':
                value = bool(int(value))
            elif data_type == 's':
                value = self.shared_strings[int(value)]
            elif data_type == 'str':
                data_type = 's'
            elif data_type == 'd':
                value = from_ISO8601(value)

        else:
            if data_type == 'inlineStr':
                child = element.find(self.INLINE_STRING)
                if child is not None:
                    data_type = 's'
                    richtext = Text.from_tree(child)
                    value = richtext.content

        if self.guess_types or value is None:
            cell.value = value
        else:
            cell._value = value
            cell.data_type = data_type
        cell.cached_value = cached_value
コード例 #7
0
ファイル: translate.py プロジェクト: BatataWada1/NOP-commerce
 def __init__(self, formula, origin):
     # Excel errors out when a workbook has formulae in R1C1 notation,
     # regardless of the calcPr:refMode setting, so I'm assuming the
     # formulae stored in the workbook must be in A1 notation.
     self.row, self.col = coordinate_to_tuple(origin)
     self.tokenizer = Tokenizer(formula)
コード例 #8
0
 def __delitem__(self, key):
     row, column = coordinate_to_tuple(key)
     if (row, column) in self._cells:
         del self._cells[(row, column)]