Class UnicodeData

Tools/unicode/makeunicodedata.py:1018–1247

# load a unicode-data file from disk

class UnicodeData:
    # table: List[Optional[UcdRecord]]  # index is codepoint; None means unassigned

    def __init__(self, version, ideograph_check=True):
        self.changed = []
        table = [None] * 0x110000
        for s in UcdFile(UNICODE_DATA, version):
            char = int(s[0], 16)
            table[char] = from_row(s)

        self.derived_name_ranges = []
        self.derived_name_prefixes = {
            prefix: i
            for i, (_, prefix) in enumerate(derived_name_range_names)
        }

        # expand first-last ranges
        field = None
        for i in range(0, 0x110000):
            # The file UnicodeData.txt has its own distinct way of
            # expressing ranges.  See:
            #   https://www.unicode.org/reports/tr44/#Code_Point_Ranges
            s = table[i]
            if s:
                if s.name[-6:] == "First>":
                    s.name = ""
                    field = dataclasses.astuple(s)[:15]
                elif s.name[-5:] == "Last>":
                    for j, (rangename, _) in enumerate(derived_name_range_names):
                        if s.name.startswith("<" + rangename):
                            self.derived_name_ranges.append(
                                (field[0], s.codepoint, j))
                            break
                    s.name = ""
                    field = None
                else:
                    codepoint = s.codepoint
                    if s.name.endswith(codepoint):
                        prefix = s.name[:-len(codepoint)]
                        j = self.derived_name_prefixes.get(prefix)
                        if j is None:
                            j = len(self.derived_name_prefixes)
                            self.derived_name_prefixes[prefix] = j
                        if (self.derived_name_ranges
                                and self.derived_name_ranges[-1][2] == j
                                and int(self.derived_name_ranges[-1][1], 16) == i - 1):
                            self.derived_name_ranges[-1][1] = codepoint
                        else:
                            self.derived_name_ranges.append(
                                [codepoint, codepoint, j])
                        s.name = ""
            elif field:
                table[i] = from_row(('%04X' % i,) + field[1:])

        # public attributes
        self.filename = UNICODE_DATA % ''
        self.table = table
        self.chars = list(range(0x110000)) # unicode 3.2