tableparser.py

Requirement =====================================================================================

# Author: David Goodger
# Contact: [e-mail]
# Revision: $Revision: 4152 $
# Date: $Date: 2005-12-08 00:46:30 +0100 (Thu, 08 Dec 2005) $
# Copyright: This module has been placed in the public domain.

"""
This module defines table parser classes,which parse plaintext-graphic tables
and produce a well-formed data structure suitable for building a CALS table.

:Classes:
    - `GridTableParser`: Parse fully-formed tables represented with a grid.
    - `SimpleTableParser`: Parse simple tables, delimited by top & bottom
      borders.

:Exception class: `TableMarkupError`

:Function:
    `update_dict_of_lists()`: Merge two dictionaries containing list values.
"""

__docformat__ = 'reStructuredText'


import re
import sys
from docutils import DataError


class TableMarkupError(DataError): pass


class TableParser:

    """
    Abstract superclass for the common parts of the syntax-specific parsers.
    """

    head_body_separator_pat = None
    """Matches the row separator between head rows and body rows."""

    double_width_pad_char = '\x00'
    """Padding character for East Asian double-width text."""

    def parse(self, block):
        """
        Analyze the text `block` and return a table data structure.

        Given a plaintext-graphic table in `block` (list of lines of text; no
        whitespace padding), parse the table, construct and return the data
        necessary to construct a CALS table or equivalent.

        Raise `TableMarkupError` if there is any problem with the markup.
        """
        self.setup(block)
        self.find_head_body_sep()
        self.parse_table()
        structure = self.structure_from_cells()
        return structure

    def find_head_body_sep(self):
        """Look for a head/body row separator line; store the line index."""
        for i in range(len(self.block)):
            line = self.block[i]
            if self.head_body_separator_pat.match(line):
                if self.head_body_sep:
                    raise TableMarkupError(
                        'Multiple head/body row separators in table (at line '
                        'offset %s and %s); only one allowed.'
                        % (self.head_body_sep, i))
                else:
                    self.head_body_sep = i
                    self.block[i] = line.replace('=', '-')
        if self.head_body_sep == 0 or self.head_body_sep == (len(self.block)
                                                             - 1):
            raise TableMarkupError('The head/body row separator may not be '
                                   'the first or last line of the table.')


class GridTableParser(TableParser):

    """
    Parse a grid table using `parse()`.

    Here's an example of a grid table::

        +------------------------+------------+----------+----------+
        | Header row, column 1   | Header 2   | Header 3 | Header 4 |
        +========================+============+==========+==========+
        | body row 1, column 1   | column 2   | column 3 | column 4 |
        +------------------------+------------+----------+----------+
        | body row 2             | Cells may span columns.          |
        +------------------------+------------+---------------------+
        | body row 3             | Cells may  | - Table cells       |
        +------------------------+ span rows. | - contain           |
        | body row 4             |            | - body elements.    |
        +------------------------+------------+---------------------+

    Intersections use '+', row separators use '-' (except for one optional
    head/body row separator, which uses '='), and column separators use '|'.

    Passing the above table to the `parse()` method will result in the
    following data structure::

        ([24, 12, 10, 10],
         [[(0, 0, 1, ['Header row, column 1']),
           (0, 0, 1, ['Header 2']),
           (0, 0, 1, ['Header 3']),
           (0, 0, 1, ['Header 4'])]],
         [[(0, 0, 3, ['body row 1, column 1']),
           (0, 0, 3, ['column 2']),
           (0, 0, 3, ['column 3']),
           (0, 0, 3, ['column 4'])],
          [(0, 0, 5, ['body row 2']),
           (0, 2, 5, ['Cells may span columns.']),
           None,
           None],
          [(0, 0, 7, ['body row 3']),
           (1, 0, 7, ['Cells may', 'span rows.', '']),
           (1, 1, 7, ['- Table cells', '- contain', '- body elements.']),
           None],
          [(0, 0, 9, ['body row 4']), None, None, None]])

    The first item is a list containing column widths (colspecs). The second
    item is a list of head rows, and the third is a list of body rows. Each
    row contains a list of cells. Each cell is either None (for a cell unused
    because of another cell's span), or a tuple. A cell tuple contains four
    items: the number of extra rows used by the cell in a vertical span
    (morerows); the number of extra columns used by the cell in a horizontal
    span (morecols); the line offset of the first line of the cell contents;
    and the cell contents, a list of lines of text.
    """

    head_body_separator_pat = re.compile(r'\+=[=+]+=\+ *$')

    def setup(self, block):
        self.block = block[:]           # make a copy; it may be modified
        self.block.disconnect()         # don't propagate changes to parent
        self.bottom = len(block) - 1
        self.right = len(block[0]) - 1
        self.head_body_sep = None
        self.done = [-1] * len(block[0])
        self.cells = []
        self.rowseps = {0: [0]}
        self.colseps = {0: [0]}

    def parse_table(self):
        """
        Start with a queue of upper-left corners, containing the upper-left
        corner of the table itself. Trace out one rectangular cell, remember
        it, and add its upper-right and lower-left corners to the queue of
        potential upper-left corners of further cells. Process the queue in
        top-to-bottom order, keeping track of how much of each text column has
        been seen.

        We'll end up knowing all the row and column boundaries, cell positions
        and their dimensions.
        """
        corners = [(0, 0)]
        while corners:
            top, left = corners.pop(0)
            if top == self.bottom or left == self.right \
                  or top <= self.done[left]:
                continue
            result = self.scan_cell(top, left)
            if not result:
                continue
            bottom, right, rowseps, colseps = result
            update_dict_of_lists(self.rowseps, rowseps)
            update_dict_of_lists(self.colseps, colseps)
            self.mark_done(top, left, bottom, right)
            cellblock = self.block.get_2D_block(top + 1, left + 1,
                                                bottom, right)
            cellblock.disconnect()      # lines in cell can't sync with parent
            cellblock.replace(self.double_width_pad_char, '')
            self.cells.append((top, left, bottom, right, cellblock))
            corners.extend([(top, right), (bottom, left)])
            corners.sort()
        if not self.check_parse_complete():
            raise TableMarkupError('Malformed table; parse incomplete.')

    def mark_done(self, top, left, bottom, right):
        """For keeping track of how much of each text column has been seen."""
        before = top - 1
        after = bottom - 1
        for col in range(left, right):
            assert self.done[col] == before
            self.done[col] = after

    def check_parse_complete(self):
        """Each text column should have been completely seen."""
        last = self.bottom - 1
        for col in range(self.right):
            if self.done[col] != last:
                return None
        return 1

    def scan_cell(self, top, left):
        """Starting at the top-left corner, start tracing out a cell."""
        assert self.block[top][left] == '+'
        result = self.scan_right(top, left)
        return result

    def scan_right(self, top, left):
        """
        Look for the top-right corner of the cell, and make note of all column
        boundaries ('+').
        """
        colseps = {}
        line = self.block[top]
        for i in range(left + 1, self.right + 1):
            if line[i] == '+':
                colseps[i] = [top]
                result = self.scan_down(top, left, i)
                if result:
                    bottom, rowseps, newcolseps = result
                    update_dict_of_lists(colseps, newcolseps)
                    return bottom, i, rowseps, colseps
            elif line[i] != '-':
                return None
        return None

    def scan_down(self, top, left, right):
        """
        Look for the bottom-right corner of the cell, making note of all row
        boundaries.
        """
        rowseps = {}
        for i in range(top + 1, self.bottom + 1):
            if self.block[i][right] == '+':
                rowseps[i] = [right]
                result = self.scan_left(top, left, i, right)
                if result:
                    newrowseps, colseps = result
                    update_dict_of_lists(rowseps, newrowseps)
                    return i, rowseps, colseps
            elif self.block[i][right] != '|':
                return None
        return None

    def scan_left(self, top, left, bottom, right):
        """
        Noting column boundaries, look for the bottom-left corner of the cell.
        It must line up with the starting point.
        """
        colseps = {}
        line = self.block[bottom]
        for i in range(right - 1, left, -1):
            if line[i] == '+':
                colseps[i] = [bottom]
            elif line[i] != '-':
                return None
        if line[left] != '+':
            return None
        result = self.scan_up(top, left, bottom, right)
        if result is not None:
            rowseps = result
            return rowseps, colseps
        return None

    def scan_up(self, top, left, bottom, right):
        """
        Noting row boundaries, see if we can return to the starting point.
        """