def annotate(data, header):
cache = _optcache()
if cache is None:
return None
for field in header.structs[0].fields:
if field.label == 10003:
rowtype = field.type.elem_type
Structrows = gff4._structtype(rowtype.fourcc, rowtype.fields, rowtype.size)
field10003 = gff4.Field(field.label, gff4._listtype(Structrows, False), False, field.offset)
elif field.label == 10002:
field10002 = gff4.Field(field.label, ListStructcolm, False, field.offset)
Structgtop = gff4._structtype(header.structs[0].fourcc, (field10002, field10003), header.structs[0].size)
header = header._replace(structs=(Structgtop, Structcolm, Structrows))
gtop = Structgtop()
colms = gtop[10002]
existing = []
for colm in data[10002]:
colm = Structcolm(colm)
if not colm[10000]:
colname = lookupcolname(colm[10001], cache)
if colname is not None:
colm[10000] = colname
else:
existing.append(colm[10000])
colms.append(colm)
gtop[10003].extend(Structrows(row) for row in data[10003])
addnames(existing)
return gtop, header
def csv2gda(source, dest=None, platform='PC '):
f = source
close = False
if isinstance(source, basestring):
f = open(source, 'rb')
close = True
csvr = csv.reader(f)
names = csvr.next()
colcount = len(names)
types = csvr.next()
if len(types) != colcount:
raise RuntimeError, 'inconsistent number of columns'
if names[0].lower() != 'id':
raise RuntimeError, 'first column must be ID'
if types[0] != 'int':
raise RuntimeError, 'ID column must be int'
goodnames = [s.decode('utf_8') for s in names if not CRC_NAME.match(s)]
addnames(goodnames)
column_list = ListStructcolm()
row_fields = []
field_offset = 0
field_label = G2DA_COLUMN_1_LABEL
for name, typename in zip(names, types):
if typename == 'comment':
continue
colm = Structcolm()
if CRC_NAME.match(name):
colm[10001] = int(name[1:], 16)
else:
colm[10001] = checksum(name)
coltypeid = coltypes.index(typename)
colm[10999] = coltypeid
column_list.append(colm)
field = gff4.Field(field_label, gfftypes[coltypeid], False,
field_offset)
field_label += 1
field_offset += 4
row_fields.append(field)
Structrows = gff4._structtype('rows', row_fields, field_offset)
ListStructrows = gff4._listtype(Structrows, False)
G2DA_ROW_LIST = gff4.Field(10003, ListStructrows, False, 4)
Structgtop = gff4._structtype('gtop', (G2DA_COLUMN_LIST, G2DA_ROW_LIST), 8)
header = gff4.Header('V4.0', platform, 'G2DA', 'V0.2', 0, 0, 0,
(Structgtop, Structcolm, Structrows))
rows_list = ListStructrows()
for cells in csvr:
if len(cells) != colcount:
raise RuntimeError, 'inconsistent number of columns'
field_label = G2DA_COLUMN_1_LABEL
row = Structrows()
for cell, typename in zip(cells, types):
if typename == 'comment':
continue
if cell != '****':
row[field_label] = cell
field_label += 1
rows_list.append(row)
gtop = Structgtop()
gtop[10002] = column_list
gtop[10003] = rows_list
if close:
f.close()
if dest is None:
return gff4.GFF4Data(header, gtop)
elif isinstance(dest, basestring):
with open(dest, 'wb') as f:
gff4.write_gff4(f, gtop, header, 4)
else:
gff4.write_gff4(dest, gtop, header, 4)
csvw.writerow([coltypes[col[10999]] for col in gda[10002]])
for row in gda[10003]:
csvw.writerow([colvalue(v) for k, v in sorted(row.iteritems())])
finally:
if close:
f.close()
#G2DA_COLUMN_NAME = gff4.Field(10000, gff4.ECString, False, 0)
#G2DA_COLUMN_HASH = gff4.Field(10001, gff4.UINT32, False, 4)
#G2DA_COLUMN_TYPE = gff4.Field(10999, gff4.UINT8, False, 8)
#Structcolm = gff4._structtype('colm', (G2DA_COLUMN_NAME, G2DA_COLUMN_HASH, G2DA_COLUMN_TYPE), 12)
G2DA_COLUMN_HASH = gff4.Field(10001, gff4.UINT32, False, 0)
G2DA_COLUMN_TYPE = gff4.Field(10999, gff4.UINT8, False, 4)
Structcolm = gff4._structtype('colm', (G2DA_COLUMN_HASH, G2DA_COLUMN_TYPE), 8)
ListStructcolm = gff4._listtype(Structcolm, False)
G2DA_COLUMN_LIST = gff4.Field(10002, ListStructcolm, False, 0)
G2DA_COLUMN_1_LABEL = 10005
def csv2gda(source, dest=None, platform='PC '):
f = source
close = False
if isinstance(source, basestring):
f = open(source, 'rb')
close = True
csvr = csv.reader(f)
names = csvr.next()
from array import array
import gff4
field19002 = gff4.Field(19002, gff4.UINT32, False, 0)
field19003 = gff4.Field(19003, gff4.ECString, False, 4)
StructSTRN = gff4._structtype('STRN', [field19002, field19003], 8)
field19001 = gff4.Field(19001, gff4._listtype(StructSTRN, False), False, 0)
StructTLK_20 = gff4._structtype('TLK ', [field19001], 4)
tlkheader = gff4.Header(version='V4.0', platform='PC ', file_type='TLK ', file_version='V0.2', string_count=0, string_offset=96, data_offset=96, structs=(StructTLK_20, StructSTRN))
def htlk2tlk(htlk):
tlk = StructTLK_20()
table = htlk[19007]
data = htlk[19008]
for hstr in htlk[19006]:
id = hstr[19004]
start = hstr[19005]
tlk[19001].append(StructSTRN({19002: id, 19003: _decompress(start, table, data)}))
return tlk
def getstring(htlk, id):
for hstr in htlk[19006]:
if id == hstr[19004]:
return _decompress(hstr[19005], htlk=htlk)
raise KeyError
def _decompress(start, table=None, data=None, htlk=None):
if table is None:
table = htlk[19007]
if data is None:
data = htlk[19008]
cache.executemany("INSERT OR IGNORE INTO names (crc, string) VALUES (?, ?);", (_tuple(s) for s in ss))
cache.commit()
def dumpnames(f):
cache = _optcache()
count = 0
if cache is not None:
for row in cache.execute("SELECT string FROM names ORDER BY string ASC;"):
print >>f, row[0]
count += 1
return count
field10000 = gff4.Field(10000, gff4.ECString, False, 0)
field10001 = gff4.Field(10001, gff4.UINT32, False, 4)
field10999 = gff4.Field(10999, gff4.UINT8, False, 8)
Structcolm = gff4._structtype('colm', (field10000, field10001, field10999), 12)
ListStructcolm = gff4._listtype(Structcolm, False)
def annotate(data, header):
cache = _optcache()
if cache is None:
return None
for field in header.structs[0].fields:
if field.label == 10003:
rowtype = field.type.elem_type
Structrows = gff4._structtype(rowtype.fourcc, rowtype.fields, rowtype.size)
field10003 = gff4.Field(field.label, gff4._listtype(Structrows, False), False, field.offset)
elif field.label == 10002:
field10002 = gff4.Field(field.label, ListStructcolm, False, field.offset)
Structgtop = gff4._structtype(header.structs[0].fourcc, (field10002, field10003), header.structs[0].size)
header = header._replace(structs=(Structgtop, Structcolm, Structrows))