MoinLight

Annotated moinformat/parsers/table.py

42:083967e32406
40:bc22df23b35e
38:427c66773470
2017-12-12 Paul Boddie file changeset files shortlog Added a parser manifest and exposed manifests via subpackage roots.
paul@38 1 
#!/usr/bin/env python
paul@38 2 
 
 paul@38      3 
"""
 
 paul@38      4 
Moin wiki table parser.
 
 paul@38      5 
 
 paul@38      6 
Copyright (C) 2017 Paul Boddie <paul@boddie.org.uk>
 
 paul@38      7 
 
 paul@38      8 
This program is free software; you can redistribute it and/or modify it under
 
 paul@38      9 
the terms of the GNU General Public License as published by the Free Software
 
 paul@38     10 
Foundation; either version 3 of the License, or (at your option) any later
 
 paul@38     11 
version.
 
 paul@38     12 
 
 paul@38     13 
This program is distributed in the hope that it will be useful, but WITHOUT
 
 paul@38     14 
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
 
 paul@38     15 
FOR A PARTICULAR PURPOSE.  See the GNU General Public License for more
 
 paul@38     16 
details.
 
 paul@38     17 
 
 paul@38     18 
You should have received a copy of the GNU General Public License along with
 
 paul@38     19 
this program.  If not, see <http://www.gnu.org/licenses/>.
 
 paul@38     20 
"""
 
 paul@38     21 
 
 paul@38     22 
from moinformat.parsing import get_patterns
 
 paul@38     23 
from moinformat.tree import Table, TableAttrs, TableCell, TableRow, Text
 
 paul@38     24 
from moinformat import Parser
 
 paul@38     25 
 
 paul@38     26 
 
 paul@38     27 
 
 paul@38     28 
# Parser functionality.
 
 paul@38     29 
 
 paul@38     30 
class TableParser(Parser):
 
 paul@38     31 
 
 paul@38     32 
    "A parser for improved table syntax."
 
 paul@38     33 
 
 paul@38     34 
    # Principal parser methods.
 
 paul@38     35 
 
 paul@38     36 
    def parse_region_content(self, items, region):
 
 paul@38     37 
 
 paul@38     38 
        "Parse the data provided by 'items' to populate the given 'region'."
 
 paul@38     39 
 
 paul@38     40 
        self.set_region(items, region)
 
 paul@38     41 
        self.parse_table_region()
 
 paul@38     42 
 
 paul@38     43 
    def parse_table_region(self):
 
 paul@38     44 
 
 paul@38     45 
        # Start to populate table rows.
 
 paul@38     46 
 
 paul@38     47 
        cell = TableCell([])
 
 paul@38     48 
        row = TableRow([cell])
 
 paul@38     49 
        table = Table([row])
 
 paul@38     50 
        self.region.append(table)
 
 paul@38     51 
 
 paul@38     52 
        while True:
 
 paul@38     53 
            self.parse_region_details(cell, self.table_region_pattern_names)
 
 paul@38     54 
 
 paul@38     55 
            # Detect the end of the table.
 
 paul@38     56 
 
 paul@38     57 
            if self.read_matching() == "regionend":
 
 paul@38     58 
                break
 
 paul@38     59 
 
 paul@38     60 
            if self.read_matching() == "columnsep":
 
 paul@38     61 
                cell = TableCell([])
 
 paul@38     62 
                row.append(cell)
 
 paul@38     63 
 
 paul@38     64 
            elif self.read_matching() == "rowsep":
 
 paul@38     65 
                row = TableRow([])
 
 paul@38     66 
                table.append(row)
 
 paul@38     67 
                cell = TableCell([])
 
 paul@38     68 
                row.append(cell)
 
 paul@38     69 
 
 paul@38     70 
    # Parser handler methods.
 
 paul@38     71 
 
 paul@38     72 
    def parse_continuation(self, cell):
 
 paul@38     73 
        pass
 
 paul@38     74 
 
 paul@38     75 
    def parse_table_end(self, cell):
 
 paul@38     76 
 
 paul@38     77 
        "Handle the end of a region within 'cell'."
 
 paul@38     78 
 
 paul@38     79 
        feature = self.read_match()
 
 paul@38     80 
        if self.region.have_end(feature):
 
 paul@38     81 
            raise StopIteration
 
 paul@38     82 
        else:
 
 paul@38     83 
            cell.append_inline(Text(feature))
 
 paul@38     84 
 
 paul@38     85 
    # Regular expressions.
 
 paul@38     86 
 
 paul@38     87 
    syntax = {}
 
 paul@38     88 
    syntax.update(Parser.syntax)
 
 paul@38     89 
    syntax.update({
 
 paul@38     90 
        # At start of line:
 
 paul@38     91 
        "rowsep"        : r"^==(?!.*==\s*?$)(?=\N*?)",  # == not-heading ws-excl-nl
 
 paul@38     92 
        "continuation"  : r"^(\N*)\.\.(?!\.)(?=\N)",    # .. ws-excl-nl or .. not-dot
 
 paul@38     93 
 
 paul@38     94 
        # Within text:
 
 paul@38     95 
        "columnsep"     : r"\|\|(?!\|)(?=\N)",          # || ws-excl-nl or || not-pipe
 
 paul@38     96 
        })
 
 paul@38     97 
 
 paul@38     98 
    patterns = get_patterns(syntax)
 
 paul@38     99 
 
 paul@38    100 
 
 paul@38    101 
 
 paul@38    102 
    # Pattern details.
 
 paul@38    103 
 
 paul@38    104 
    table_region_pattern_names = Parser.region_pattern_names + [
 
 paul@38    105 
        "columnsep", "continuation", "regionend", "rowsep",
 
 paul@38    106 
        ]
 
 paul@38    107 
 
 paul@38    108 
 
 paul@38    109 
 
 paul@38    110 
    # Pattern handlers.
 
 paul@38    111 
 
 paul@38    112 
    handlers = {}
 
 paul@38    113 
    handlers.update(Parser.handlers)
 
 paul@38    114 
    handlers.update({
 
 paul@38    115 
        "columnsep" : Parser.end_region,
 
 paul@38    116 
        "continuation" : parse_continuation,
 
 paul@38    117 
        "rowsep" : Parser.end_region,
 
 paul@38    118 
        "regionend" : parse_table_end,
 
 paul@38    119 
        })
 
 paul@38    120 
 
 paul@40    121 
parser = TableParser
 
 paul@40    122 
 
 paul@38    123 
# vim: tabstop=4 expandtab shiftwidth=4