def read_from_ccj(self,
f,
title,
author,
puzzle_number,
copyright_message,
verbose=False):
d = f.read()
# i is the index into the file for the rest of this script:
i = 2
# I think these must be the list of buttons on the left:
while byte_at(d, i) != 0:
s, i = read_string(d, i)
if verbose:
print("got button string:", s)
# Then the congratulations message, I think:
i += 1
s, i = read_string(d, i)
if verbose:
print("got congratulations message:", s)
# Skip another byte; 0x02 in the Independent it seems, but 0x00 in the
# Herald puzzle I tried.
i += 1
# I think we get the grid dimensions in the next two:
self.width = byte_at(d, i)
i += 1
self.height = byte_at(d, i)
i += 1
self.grid = Grid(self.width, self.height)
# Now skip over everything until we think we see the grid, since I've
# no idea what it's meant to mean:
while byte_at(d, i) != 0x3f and byte_at(d, i) != 0x23:
i += 1
for y in range(0, self.height):
for x in range(0, self.width):
# Lights seem to be indicated by: '?' (or 'M' very occasionally)
if byte_at(d, i) in (0x3f, 0x4d):
self.grid.cells[y][x] = Cell(y, x)
# Blocked-out squares seem to be always '#'
elif byte_at(d, i) == 0x23:
pass
else:
message = "Unknown value {0} at {1}"
raise Exception(message.format(str(byte_at(d, i)),
coord_str(x, y)))
i += 1
if verbose:
print("grid is:\n" + self.grid.to_grid_string(True))
# Next there's a grid structure the purpose of which I don't
# understand:
grid_unknown_purpose = Grid(self.width, self.height)
for y in range(0, self.height):
for x in range(0, self.width):
grid_unknown_purpose.cells[y][x] = Cell(y, x)
if byte_at(d, i) == 0:
grid_unknown_purpose.cells[y][x].set_letter(' ')
elif byte_at(d, i) < 10:
letter = str(byte_at(d, i))
grid_unknown_purpose.cells[y][x].set_letter(letter)
else:
truncated = str(byte_at(d, i) % 10)
if verbose:
message = "Warning, truncating {0} to {1} at {2}"
print(message.format(byte_at(d, i),
truncated,
coord_str(x, y)))
grid_unknown_purpose.cells[y][x].set_letter(truncated)
i += 1
# Seem to need to skip over an extra byte (0x01) here before the
# answers. Maybe it indicates whether there are answers next or not:
if byte_at(d, i) != 1:
raise Exception("So far we expect a 0x01 before the answers...")
i += 1
if verbose:
print("grid_unknown_purpose is:\n" +
grid_unknown_purpose.to_grid_string(False))
# Now there's the grid with the answers:
for y in range(0, self.height):
for x in range(0, self.width):
if self.grid.cells[y][x]:
self.grid.cells[y][x].set_letter(chr(byte_at(d, i)))
i += 1
if verbose:
print("grid with answers is:\n" + self.grid.to_grid_string(False))
skipped_blocks_of_four = 0
while skippable_block_of_four(d, i):
i += 4
skipped_blocks_of_four += 1
if skipped_blocks_of_four > 0:
if verbose:
print("Skipped over",
str(skipped_blocks_of_four),
"ignorable blocks")
# I expect the next one to be 0x02:
if byte_at(d, i) != 0x02:
message = "Expect the first of the block of 16 always to be 0x02, "
message += "in fact was: {0}"
raise Exception(message.format(byte_at(d, i)))
# Always just 16?
i += 16
self.across_clues, i = parse_list_of_clues(d, i)
if verbose:
print("Now do down clues:")
self.down_clues, i = parse_list_of_clues(d, i)
m = re.search(r'^(.*)-([0-9]+)', self.across_clues.label)
if m:
self.setter = m.group(1)
self.puzzle_number = m.group(2)
if (not self.setter) and author:
self.setter = author
if (not self.puzzle_number) and puzzle_number:
self.puzzle_number = puzzle_number
self.title = "Crossword"
if title:
self.title = title
if self.setter and self.puzzle_number:
self.title += " " + self.puzzle_number + " / " + self.setter
elif self.setter:
self.title += " / " + self.setter
elif self.puzzle_number:
self.title += " " + self.puzzle_number
if date_string:
self.title += " (" + date_string + ")"
self.author = "Unknown Setter"
if author:
self.author = author
self.copyright_message = "© Unknown"
if copyright_message:
self.copyright_message = copyright_message
class ParsedCCJ:
def __init__(self):
self.width = None
self.height = None
self.across_clues = None
self.down_clues = None
self.grid = None
self.title = None
self.author = None
self.copyright_message = None
self.setter = None
self.puzzle_number = None
def read_from_ccj(self,
f,
title,
author,
puzzle_number,
copyright_message,
verbose=False):
d = f.read()
# i is the index into the file for the rest of this script:
i = 2
# I think these must be the list of buttons on the left:
while byte_at(d, i) != 0:
s, i = read_string(d, i)
if verbose:
print("got button string:", s)
# Then the congratulations message, I think:
i += 1
s, i = read_string(d, i)
if verbose:
print("got congratulations message:", s)
# Skip another byte; 0x02 in the Independent it seems, but 0x00 in the
# Herald puzzle I tried.
i += 1
# I think we get the grid dimensions in the next two:
self.width = byte_at(d, i)
i += 1
self.height = byte_at(d, i)
i += 1
self.grid = Grid(self.width, self.height)
# Now skip over everything until we think we see the grid, since I've
# no idea what it's meant to mean:
while byte_at(d, i) != 0x3f and byte_at(d, i) != 0x23:
i += 1
for y in range(0, self.height):
for x in range(0, self.width):
# Lights seem to be indicated by: '?' (or 'M' very occasionally)
if byte_at(d, i) in (0x3f, 0x4d):
self.grid.cells[y][x] = Cell(y, x)
# Blocked-out squares seem to be always '#'
elif byte_at(d, i) == 0x23:
pass
else:
message = "Unknown value {0} at {1}"
raise Exception(message.format(str(byte_at(d, i)),
coord_str(x, y)))
i += 1
if verbose:
print("grid is:\n" + self.grid.to_grid_string(True))
# Next there's a grid structure the purpose of which I don't
# understand:
grid_unknown_purpose = Grid(self.width, self.height)
for y in range(0, self.height):
for x in range(0, self.width):
grid_unknown_purpose.cells[y][x] = Cell(y, x)
if byte_at(d, i) == 0:
grid_unknown_purpose.cells[y][x].set_letter(' ')
elif byte_at(d, i) < 10:
letter = str(byte_at(d, i))
grid_unknown_purpose.cells[y][x].set_letter(letter)
else:
truncated = str(byte_at(d, i) % 10)
if verbose:
message = "Warning, truncating {0} to {1} at {2}"
print(message.format(byte_at(d, i),
truncated,
coord_str(x, y)))
grid_unknown_purpose.cells[y][x].set_letter(truncated)
i += 1
# Seem to need to skip over an extra byte (0x01) here before the
# answers. Maybe it indicates whether there are answers next or not:
if byte_at(d, i) != 1:
raise Exception("So far we expect a 0x01 before the answers...")
i += 1
if verbose:
print("grid_unknown_purpose is:\n" +
grid_unknown_purpose.to_grid_string(False))
# Now there's the grid with the answers:
for y in range(0, self.height):
for x in range(0, self.width):
if self.grid.cells[y][x]:
self.grid.cells[y][x].set_letter(chr(byte_at(d, i)))
i += 1
if verbose:
print("grid with answers is:\n" + self.grid.to_grid_string(False))
skipped_blocks_of_four = 0
while skippable_block_of_four(d, i):
i += 4
skipped_blocks_of_four += 1
if skipped_blocks_of_four > 0:
if verbose:
print("Skipped over",
str(skipped_blocks_of_four),
"ignorable blocks")
# I expect the next one to be 0x02:
if byte_at(d, i) != 0x02:
message = "Expect the first of the block of 16 always to be 0x02, "
message += "in fact was: {0}"
raise Exception(message.format(byte_at(d, i)))
# Always just 16?
i += 16
self.across_clues, i = parse_list_of_clues(d, i)
if verbose:
print("Now do down clues:")
self.down_clues, i = parse_list_of_clues(d, i)
m = re.search(r'^(.*)-([0-9]+)', self.across_clues.label)
if m:
self.setter = m.group(1)
self.puzzle_number = m.group(2)
if (not self.setter) and author:
self.setter = author
if (not self.puzzle_number) and puzzle_number:
self.puzzle_number = puzzle_number
self.title = "Crossword"
if title:
self.title = title
if self.setter and self.puzzle_number:
self.title += " " + self.puzzle_number + " / " + self.setter
elif self.setter:
self.title += " / " + self.setter
elif self.puzzle_number:
self.title += " " + self.puzzle_number
if date_string:
self.title += " (" + date_string + ")"
self.author = "Unknown Setter"
if author:
self.author = author
self.copyright_message = "© Unknown"
if copyright_message:
self.copyright_message = copyright_message
def write_to_puz_file(self, output_filename, verbose=False):
"""Write the crossword in AcrossLite .puz format to output_filename
Note that the version for the file format that this outputs
doesn't include checksums, so a strict loader will reject such
a file - it's fine in xword, though."""
# In the AcrossLite .PUZ format we need to make sure that there's one
# "clue" for each clue number, even if it's just "See 6" for clues
# whose answers are split over different clue numbers in the grid.
# So, go through the clue dictionaries and make sure that there is
# something for every clue. (So we don't miss the "See 6" type of
# clue.)
# We take a deep copy of the across clues and down clues first
# so that we don't add unnecessary fake clues to the
# attributes of this instance.
clue_groups = {
True: copy.deepcopy(self.across_clues),
False: copy.deepcopy(self.down_clues)}
for group_across in (True, False):
clue_dictionary = clue_groups[group_across].clue_dictionary
for clue in clue_dictionary.values():
first_clue_entry = str(clue.all_clue_numbers[0][0])
for entry_n, entry_across in clue.all_clue_numbers:
clue_string = "See " + first_clue_entry
if entry_across != group_across:
clue_string += entry_across and " across" or " down"
expected_dictionary = {
True: clue_groups[True].clue_dictionary,
False: clue_groups[False].clue_dictionary
}[entry_across]
if entry_n not in expected_dictionary.keys():
fake_clue = ParsedClue()
fake_clue.across = entry_across
fake_clue.text_including_enumeration = clue_string
fake_clue.set_number(str(entry_n))
expected_dictionary[entry_n] = fake_clue
if verbose:
print("**** Added missing clue with index ", str(entry_n),
fake_clue.tidied_text_including_enumeration())
# Now the file can be written out:
with io.FileIO(output_filename, 'wb') as f:
f.write(bytearray(0x2C))
dimensions_etc = bytearray(2)
dimensions_etc[0] = self.width
dimensions_etc[1] = self.height
f.write(dimensions_etc)
f.write(struct.pack("<h",
clue_groups[True].real_number_of_clues() +
clue_groups[False].real_number_of_clues()))
f.write(bytearray(4))
solutions = bytearray(self.width*self.height)
empty_grid = bytearray(self.width*self.height)
i = 0
for y in range(0, self.height):
for x in range(0, self.width):
c = self.grid.cells[y][x]
if c:
solutions[i] = ord(c.letter)
empty_grid[i] = ord('-')
else:
solutions[i] = ord('.')
empty_grid[i] = ord('.')
i += 1
f.write(solutions)
f.write(empty_grid)
nul = bytearray(1)
f.write(self.title.encode('UTF-8'))
f.write(nul)
f.write(self.author.encode('UTF-8'))
f.write(nul)
f.write(self.copyright_message.encode('UTF-8'))
f.write(nul)
all_clues = clue_groups[True].ordered_list_of_clues()
all_clues += clue_groups[False].ordered_list_of_clues()
all_clues.sort(key=keyfunc_clues)
for c in all_clues:
number_string_tidied = re.sub(r'/', ',', c.number_string)
number_string_tidied = number_string_tidied.lower()
clue_text = c.tidied_text_including_enumeration()
# We have to stick the number string at the beginning
# otherwise it won't be clear when the answers to clues cover
# several entries in the grid.
f.write(("[" + number_string_tidied + "] ").encode('UTF-8'))
# Encode the clue text as UTF-8, because it's not defined what
# the character set should be anywhere that I've seen. (xword
# currently assumes ISO-8859-1, but that doesn't strike me as
# a good enough reason in itself, since it's easily patched.)
f.write(clue_text.encode('UTF-8'))
f.write(nul)
f.write(nul)