def read_labelsst_record(off, size):
r = integer(wbk[off : off+2])
c = integer(wbk[off+2 : off+4])
j = integer(wbk[off+4 : off+6])
fs=xf[j]
i = integer(wbk[off+6 : off+10])
return r, c, sst[i], 's', fs
def read_xf_record(off, size):
ans=''
i=integer(wbk[off : off+2])
(b, it, u, h) = fonts[i]
ans+=str(b)
ans+=str(it)
ans+=str(u)
i=integer(wbk[off+6 : off+7])
n = i & 0x07
if n==2:
ans+='10' # center
elif n==3:
ans+='11' # right
else:
ans+='01' #left
n = (i & 0x70)/16
if n==1:
ans+='10' #center
elif n==2:
ans+='11' # bottom
else:
ans+='01' # top
n = (i & 0x08)/8
ans+=str(n) # 1 for word wrap
return ans, h
def read_label_record(off, size):
r = integer(wbk[off : off+2])
c = integer(wbk[off+2 : off+4])
i = integer(wbk[off+4 : off+6])
fs=xf[i]
s, bytes = read_string(off+6, 2)
return r, c, s, 's', fs
def read_index_record(off, size):
rf = integer(wbk[off+4:off+8])
rl = integer(wbk[off+8:off+12])
l=[]
for i in range(0, (size - 16) / 4):
l.append(integer(wbk[off+16+4*i:off+16+4*(i+1)]))
return rf, rl, l
def skip_attr(s, pos): # tAttr token: this token has many different tokens
t = integer(s[pos +
1]) # tAttrChoose token ,# IF control, choose, skip etc
if t == 0x04:
nc = integer(s[pos + 2:pos + 4])
return nc * 2 + 6, t
return 4, t
def read_label_record(off, size):
r = integer(wbk[off : off + 2])
c = integer(wbk[off + 2 : off + 4])
i = integer(wbk[off + 4 : off + 6])
fs = xf[i]
s, bytes = read_string(off + 6, 2)
return r, c, s, "s", fs
def read_float_record(off, size):
r = integer(wbk[off : off + 2])
c = integer(wbk[off + 2 : off + 4])
i = integer(wbk[off + 4 : off + 6])
fs = xf[i]
fv = floating(hexa(wbk[off + 6 : off + 14]))
return r, c, fv, "f", fs
def read_index_record(off, size):
rf = integer(wbk[off + 4 : off + 8])
rl = integer(wbk[off + 8 : off + 12])
l = []
for i in range(0, (size - 16) / 4):
l.append(integer(wbk[off + 16 + 4 * i : off + 16 + 4 * (i + 1)]))
return rf, rl, l
def read_font_record(off, size):
h = integer(wbk[off : off + 2]) / 20
i = integer(wbk[off + 2 : off + 4])
bold = (i & 0x0001) == 1
italic = (i & 0x0002) == 2
underlined = (i & 0x0004) == 4
return int(bold), int(italic), int(underlined), h
def read_xf_record(off, size):
ans = ""
i = integer(wbk[off : off + 2])
(b, it, u, h) = fonts[i]
ans += str(b)
ans += str(it)
ans += str(u)
i = integer(wbk[off + 6 : off + 7])
n = i & 0x07
if n == 2:
ans += "10" # center
elif n == 3:
ans += "11" # right
else:
ans += "01" # left
n = (i & 0x70) / 16
if n == 1:
ans += "10" # center
elif n == 2:
ans += "11" # bottom
else:
ans += "01" # top
n = (i & 0x08) / 8
ans += str(n) # 1 for word wrap
return ans, h
def read_font_record(off, size):
h = integer(wbk[off : off+2])/20
i=integer(wbk[off+2 : off+4])
bold = ((i & 0x0001)==1)
italic = ((i & 0x0002)==2)
underlined = ((i & 0x0004)==4)
return int(bold), int(italic), int(underlined), h
def read_labelsst_record(off, size):
r = integer(wbk[off : off + 2])
c = integer(wbk[off + 2 : off + 4])
j = integer(wbk[off + 4 : off + 6])
fs = xf[j]
i = integer(wbk[off + 6 : off + 10])
return r, c, sst[i], "s", fs
def read_float_record(off, size):
r = integer(wbk[off : off+2])
c = integer(wbk[off+2 : off+4])
i = integer(wbk[off+4 : off+6])
fs=xf[i]
fv = floating(hexa(wbk[off+6:off+14]))
return r, c, fv, 'f', fs
def read_formula_record(off, size):
r = integer(wbk[off : off + 2])
c = integer(wbk[off + 2 : off + 4])
i = integer(wbk[off + 4 : off + 6])
fs = xf[i]
val, typ = calc_formula_result(off + 6, 8)
formula = read_formula(wbk[off + 20 : off + size])
return r, c, val, fs, typ, formula
def read_cellrange_address(w):
r1 = str(1 + integer(w[:2]))
r2 = str(1 + integer(w[2:4]))
i = (integer(w[4:6])) & 0x00FF
c1 = get_label(i)
i = (integer(w[6:8])) & 0x00FF
c2 = get_label(i)
return c1 + r1 + ':' + c2 + r2
def read_cellrange_address(w):
r1= str(1+integer(w[:2]))
r2= str(1+integer(w[2:4]))
i= (integer(w[4:6])) & 0x00FF
c1= get_label(i)
i= (integer(w[6:8])) & 0x00FF
c2= get_label(i)
return c1+r1+ ':' +c2+r2
def read_formula_record(off, size):
r = integer(wbk[off : off+2])
c = integer(wbk[off+2 : off+4])
i = integer(wbk[off+4 : off+6])
fs=xf[i]
val, typ = calc_formula_result(off+6, 8)
formula = read_formula(wbk[off+20 : off + size])
return r, c, val, fs, typ, formula
def read_additional(s, pos, tid):
ans=''
np=pos
if tid == 0x20 or tid == 0x40 or tid == 0x60:
ans+='{'
nc= integer(s[pos])+1
nr= integer(s[pos+1:pos+3])+1
np+=3
for i in range(0, nr):
for j in range(0, nc):
if integer(s[np])==0x00: # empty
ans += ','
np += 9
elif integer(s[np])==0x01: # float
ans += str(floating(hexa(s[np+1:np+9])))+','
np+=9
elif integer(s[np])==0x02: # string
st,bytes = read_string(s, np+1, 2)
ans+= '"' + st + '"' + ','
np+=1+bytes
elif integer(s[np])==0x04: # boolean
if integer(s[np+1])==1:
ans+='TRUE,'
else:
ans+='FALSE,'
np+=9
elif integer(s[np])==0x10: # error value
ans+=err_value[integer(s[np+1])]+','
np+=9
else:
pass
ans=ans[:len(ans)-1]+';'
ans=ans[:len(ans)-1]+'}'
return ans, np-pos
return '', 0
def read_additional(s, pos, tid):
ans = ''
np = pos
if tid == 0x20 or tid == 0x40 or tid == 0x60:
ans += '{'
nc = integer(s[pos]) + 1
nr = integer(s[pos + 1:pos + 3]) + 1
np += 3
for i in range(0, nr):
for j in range(0, nc):
if integer(s[np]) == 0x00: # empty
ans += ','
np += 9
elif integer(s[np]) == 0x01: # float
ans += str(floating(hexa(s[np + 1:np + 9]))) + ','
np += 9
elif integer(s[np]) == 0x02: # string
st, bytes = read_string(s, np + 1, 2)
ans += '"' + st + '"' + ','
np += 1 + bytes
elif integer(s[np]) == 0x04: # boolean
if integer(s[np + 1]) == 1:
ans += 'TRUE,'
else:
ans += 'FALSE,'
np += 9
elif integer(s[np]) == 0x10: # error value
ans += err_value[integer(s[np + 1])] + ','
np += 9
else:
pass
ans = ans[:len(ans) - 1] + ';'
ans = ans[:len(ans) - 1] + '}'
return ans, np - pos
return '', 0
def calc_formula_result(off, size):
i = integer(wbk[off : off+1])
if i == 0x01:
return bool(integer(wbk[off+2 : off+3])), 'b'
if i == 0x02:
ec = integer(wbk[off+2 : off+3])
return err_value[ec], 'e'
if i == 0x03:
return '', 's'
else:
return floating( integer(wbk[off : off+8]) ), 'f'
def calc_formula_result(off, size):
i = integer(wbk[off : off + 1])
if i == 0x01:
return bool(integer(wbk[off + 2 : off + 3])), "b"
if i == 0x02:
ec = integer(wbk[off + 2 : off + 3])
return err_value[ec], "e"
if i == 0x03:
return "", "s"
else:
return floating(integer(wbk[off : off + 8])), "f"
def read_rk_record(off, size):
r = integer(wbk[off : off+2])
c = integer(wbk[off+2 : off+4])
i = integer(wbk[off+4 : off+6])
fs=xf[i]
rk_val = hexa(wbk[off+6 : off+10])
val, typ = read_rk_value(rk_val)
if int(val) == val:
val = int(val)
typ = 'i'
return r, c, str(val), typ, fs
def read_boolerr_record(off, size):
r = integer(wbk[off : off+2])
c = integer(wbk[off+2 : off+4])
i = integer(wbk[off+4 : off+6])
fs=xf[i]
typ = integer(wbk[off+7 : off+8])
val = integer(wbk[off+6 : off+7])
if typ == 0: # 0 for boolean
return r, c, bool(val), 'b', fs
else:
return r, c, err_value[val], 'e', fs
def read_rk_record(off, size):
r = integer(wbk[off : off + 2])
c = integer(wbk[off + 2 : off + 4])
i = integer(wbk[off + 4 : off + 6])
fs = xf[i]
rk_val = hexa(wbk[off + 6 : off + 10])
val, typ = read_rk_value(rk_val)
if int(val) == val:
val = int(val)
typ = "i"
return r, c, str(val), typ, fs
def read_boolerr_record(off, size):
r = integer(wbk[off : off + 2])
c = integer(wbk[off + 2 : off + 4])
i = integer(wbk[off + 4 : off + 6])
fs = xf[i]
typ = integer(wbk[off + 7 : off + 8])
val = integer(wbk[off + 6 : off + 7])
if typ == 0: # 0 for boolean
return r, c, bool(val), "b", fs
else:
return r, c, err_value[val], "e", fs
def read_mulrk_record(off, size):
r = integer(wbk[off : off+2])
fc = integer(wbk[off+2 : off+4])
nc = (size - 6) / 6
l = []
for i in range(0, nc):
j = integer(wbk[off+4+6*i : off+4+6*i+2])
fs=xf[j]
rk_val = integer(wbk[off + 4+6*i+2 : off+ 4+6*i+2+4])
val, typ = read_rk_value(rk_val)
if int(val) == val:
val = int(val)
typ = 'i'
c = fc + i
l.append( (r, c, val, typ, fs) )
return l
def next_record_info():
global ptr
identity = hexa(wbk[ptr : ptr+2])
off = ptr + 4
size = integer(wbk[ptr+2 : ptr+4])
ptr = ptr +4 +size
return (identity, off, size)
def read_mulrk_record(off, size):
r = integer(wbk[off : off + 2])
fc = integer(wbk[off + 2 : off + 4])
nc = (size - 6) / 6
l = []
for i in range(0, nc):
j = integer(wbk[off + 4 + 6 * i : off + 4 + 6 * i + 2])
fs = xf[j]
rk_val = integer(wbk[off + 4 + 6 * i + 2 : off + 4 + 6 * i + 2 + 4])
val, typ = read_rk_value(rk_val)
if int(val) == val:
val = int(val)
typ = "i"
c = fc + i
l.append((r, c, val, typ, fs))
return l
def next_record_info():
global ptr
identity = hexa(wbk[ptr : ptr + 2])
off = ptr + 4
size = integer(wbk[ptr + 2 : ptr + 4])
ptr = ptr + 4 + size
return (identity, off, size)
def read_sst_record(off, size):
nm = integer(wbk[off + 4 : off + 8])
off += 8
bytes = 0
l = []
s = ""
for i in range(0, nm):
(s, bytes) = read_string(off, 2)
off += bytes
l.append(s)
return l
def read_sst_record(off, size):
nm = integer(wbk[off+4:off+8])
off += 8
bytes = 0
l=[]
s=''
for i in range(0, nm):
(s, bytes) = read_string(off, 2)
off += bytes
l.append(s)
return l
def read_string(pos, len_size):
save=pos
ln=integer(wbk[pos:pos+len_size])
pos+=len_size
flag=integer(wbk[pos:pos+1])
pos+=1
rt = 0
sz = 0
if flag&8==8:
rt = integer(wbk[pos : pos+2])
pos+=2
if flag&4==4:
sz = integer(wbk[pos : pos+4])
pos+=4
if flag&1==1:
s = wbk[pos:pos+2*ln:2]
pos = pos+2*ln
else:
s = wbk[pos:pos+ln]
pos = pos + ln
pos += 4*rt
pos += sz
return s, pos-save
def read_string(pos, len_size):
save = pos
ln = integer(wbk[pos : pos + len_size])
pos += len_size
flag = integer(wbk[pos : pos + 1])
pos += 1
rt = 0
sz = 0
if flag & 8 == 8:
rt = integer(wbk[pos : pos + 2])
pos += 2
if flag & 4 == 4:
sz = integer(wbk[pos : pos + 4])
pos += 4
if flag & 1 == 1:
s = wbk[pos : pos + 2 * ln : 2]
pos = pos + 2 * ln
else:
s = wbk[pos : pos + ln]
pos = pos + ln
pos += 4 * rt
pos += sz
return s, pos - save
def read_standardwidth_record(off, size):
return integer(wbk[off : off + 2]) / 16 + 1
def read_colinfo_record(off, size):
cf = integer(wbk[off : off + 2])
cl = integer(wbk[off + 2 : off + 4])
w = integer(wbk[off + 4 : off + 6]) / 256 + 1
return cf, cl, w
def read_defcol_record(off, size):
return integer(wbk[off : off + 2])
def read_formula(s):
rpn = [] # rpn array
add_list = []
size = integer(s[:2])
pos = 2
while pos != 2 + size:
tid = integer(s[pos])
if tid >= 0x00 and tid <= 0x02: return 'UNKNOWN'
elif tid >= 0x03 and tid <= 0x16:
rpn += [token_list[tid]]
pos += 1
elif tid == 0x17: # string
st, bytes = read_string(s, pos + 1, 1)
pos = pos + 1 + bytes
rpn += ['"' + st + '"']
elif tid == 0x1C: # error value
er = err_value[integer(s[pos + 1])]
rpn += [er]
pos += 2
elif tid == 0x1D: # bool
nt = integer(s[pos + 1])
if nt == 0: rpn += ['FALSE']
else: rpn += ['TRUE']
pos += 2
elif tid == 0x1E: # integer
rpn += [str(integer(s[pos + 1:pos + 3]))]
pos += 3
elif tid == 0x1F: # float
fv = floating(hexa(s[pos + 1:pos + 9]))
pos += 9
rpn += [str(fv)]
elif tid == 0x21 or tid == 0x41 or tid == 0x61: # const argc function name
(fn, argc) = fn_list[integer(s[pos + 1:pos + 3])]
rpn += [(fn, argc)]
pos += 3
elif tid == 0x22 or tid == 0x42 or tid == 0x62: # variable argc function name
try:
fn, argc = fn_list[integer(s[pos + 2:pos + 4])]
except:
return 'UNKNOWN'
argc = integer(s[pos + 1])
rpn += [(fn, argc)]
pos += 4
elif tid == 0x24 or tid == 0x44 or tid == 0x64: # tRef
rpn += [read_cell_address(s[pos + 1:pos + 5])]
pos += 5
elif tid == 0x25 or tid == 0x45 or tid == 0x65: # tArea
rpn += [read_cellrange_address(s[pos + 1:pos + 9])]
pos += 9
elif tid == 0x20 or tid == 0x40 or tid == 0x60: # tArray
add_list += [(len(rpn), tid)]
pos += 8
elif tid == 0x26 or tid == 0x46 or tid == 0x66:
pos += 7
elif tid == 0x19:
bytes, t = skip_attr(s, pos)
pos += bytes
if t == 0x10:
rpn += [('SUM', 1)
] # tAttrSum replaces the tFuncVar token here
else:
return 'UNKNOWN' # has to take care of some other tokens too
for i in range(0, len(add_list)): # additional data
index, tid = add_list[i]
dt, bytes = read_additional(s, pos, tid)
pos += bytes
rpn.insert(index, dt)
l = len(rpn)
stk = []
for i in range(0, l): # converting RPN to formula using stack
if rpn[i] in token_list[:18]:
t2 = stk.pop()
t1 = stk.pop()
stk += [t1 + rpn[i] + t2]
elif rpn[i] == 'u+' or rpn[i] == 'u-':
t1 = stk.pop()
ch = rpn[i]
stk += [ch[1] + t1]
elif rpn[i] == '%':
t1 = stk.pop()
stk += [t1 + '%']
elif rpn[i] == '()':
t1 = stk.pop()
stk += ['(' + t1 + ')']
elif type(rpn[i]) == tuple:
fname, argc = rpn[i]
fnstr = ')'
for j in range(0, argc):
fnstr = ',' + stk.pop() + fnstr
if argc != 0:
fnstr = fnstr[1:]
fnstr = fname + '(' + fnstr
stk += [fnstr]
else:
stk += [rpn[i]]
return stk[0]
def read_cell_address(w):
r = str(1 + integer(w[:2]))
i = (integer(w[2:])) & 0x00FF
c = get_label(i)
return c + r
def read_blank_record(off, size):
r = integer(wbk[off : off + 2])
c = integer(wbk[off + 2 : off + 4])
return r, c, "", "s"
def skip_attr(s, pos): # tAttr token: this token has many different tokens
t = integer(s[pos+1]) # tAttrChoose token ,# IF control, choose, skip etc
if t == 0x04:
nc=integer(s[pos+2:pos+4])
return nc*2 + 6 ,t
return 4, t
def read_colinfo_record(off, size):
cf = integer(wbk[off : off+2])
cl = integer(wbk[off+2 : off+4])
w = integer(wbk[off+4 : off+6])/256 + 1
return cf, cl, w
def read_standardwidth_record(off, size):
return integer(wbk[off:off+2])/16 + 1
def bin2data(s):
global wbk, ptr, fonts, xf
wbk = return_wbk(s)
ptr=0
if wbk == 'file not in correct format':
return wbk
# reading the workbook global substream
(f, off, size) = next_record_info()# BOF record
#####################################################
while f != 0x0031: # reaching to FONT Records
(f, off, size) = next_record_info()
fonts=[]
while f == 0x0031: # reading font records
tpl=read_font_record(off, size)
fonts.append(tpl)
if len(fonts)==4: # since font with index 4 has been ommited
fonts.append((0, 0, 0, 0)) # by excel :)
(f, off, size) = next_record_info()
while f != 0x00E0: # reaching to XF record
(f, off, size) = next_record_info()
xf=[]
while f == 0x00E0: # reading XF records
fs = read_xf_record(off, size) # fn is formatting no.
xf.append(fs)
(f, off, size) = next_record_info()
#####################################################
while f != 0x0085: #reaching to SHEET Records
(f, off, size) = next_record_info()
sheets = [] # contains the info of all the worksheets
while f == 0x0085: #reading the SHEET records
record = wbk[off : off+size]
pos = integer(record[:4])
sheet_type = integer(record[5:6]);
(name, bytes) = read_string(off + 6, 1)
if sheet_type==0: # sheettype 0 for workssheet
sheets.append((pos, sheet_type, name))
(f, off, size) = next_record_info()
#building the SST shared string table
while f != 0x00FC: #reaching to SST record
(f, off, size) = next_record_info()
global sst
sst = [] # list of used strings
sst = read_sst_record(off, size)
workbook_data=[]
workbook_col_widths=[]
for sheet in sheets:
ptr = sheet[0] # offset of 1st worksheet
sheet_name = sheet[2]
next_record_info() # skipping the BOF record
(f, off, size) = next_record_info()
if f == 0x020B: # INDEX record
(rf, rl, dbcell_posns) = read_index_record(off, size)
###################################################
colinfo=[]
def_width = 0
standard_width = 0
while f != 0x0200: #reaching to DIMENSION record
if f==0x0055: #reading DEFCOLWIDTH
def_width = read_defcol_record(off, size)
if f==0x007D: #reading COLINFO record
cnf = read_colinfo_record(off, size)
colinfo.append( cnf )
(f, off, size) = next_record_info()
#####################################################
rf,rl,cf,cl = read_dimension_record(off, size)
if (rl-rf)%32==0: row_block_count = (rl-rf)/32
else: row_block_count = int((rl-rf)/32)+1
if row_block_count == 0: # sheet is empty
continue
#reading the data now
sheet_data = [[()]*cl]*rl # empty database
for i in range(0, row_block_count):
(f, off, size) = next_record_info()
while f == 0x0208:
(f, off, size) = next_record_info() # skipping the ROW record
while f != 0x00D7: # f==0x00D7 for DBCELL record
if f == 0x0205:
x, y, val, typ, fs = read_boolerr_record(off, size)
elif f == 0x0204:
x, y, val, typ, fs = read_label_record(off, size)
elif f == 0x00FD:
x, y, val, typ, fs = read_labelsst_record(off, size)
elif f == 0x0203:
x, y, val, typ, fs = read_float_record(off, size)
elif f == 0x027E:
x, y, val, typ, fs = read_rk_record(off, size)
elif f == 0x0006: # formula record
x, y, val, fs, typ, formula= read_formula_record(off, size)
buf = list(sheet_data[x])
buf[y] = (val, 'for', fs, typ, formula)
sheet_data[x] = buf
(f, off, size) = next_record_info()
continue
elif f == 0x00BD:
mul_cell_data = read_mulrk_record(off, size) # list of tuples
for tple in mul_cell_data:
x, y, val, typ, fs = tple
buf = list(sheet_data[x])
buf[y] = (val, typ, fs)
sheet_data[x] = buf
(f, off, size) = next_record_info()
continue
else:
(f, off, size) = next_record_info()
continue
buf = list(sheet_data[x])
buf[y] = (val, typ, fs)
sheet_data[x] = buf
(f, off, size) = next_record_info()
workbook_data.append(sheet_data)
#########################################################
while f != 0x000A: # 000A for EOF record
if f==0x0099: # standardwidth record
standard_width = read_standardwidth_record(off, size)
(f, off, size) = next_record_info()
col_widths = [0]*cl
for tpl in colinfo:
for j in range(tpl[0], tpl[1]+1):
col_widths[j] = tpl[2]
if standard_width!=0:
default = standard_width
else:
default = def_width
for j in range(0, cl):
if col_widths[j] == 0:
col_widths[j]=default
workbook_col_widths.append(col_widths)
###########################################################
return workbook_data_to_scalc_string(workbook_data, workbook_col_widths,0)
def read_row_record(off, size):
r = integer(wbk[off : off + 2])
c1 = integer(wbk[off + 2 : off + 4])
c2 = integer(wbk[off + 4 : off + 6])
return r, c1, c2
def read_cell_address(w):
r=str(1+integer(w[:2]))
i=(integer(w[2:]))&0x00FF
c =get_label(i)
return c + r
def read_dimension_record(off, size):
rf = integer(wbk[off:off+4])
rl = integer(wbk[off+4:off+8])
cf = integer(wbk[off+8:off+10])
cl = integer(wbk[off+10:off+12])
return rf,rl,cf,cl
def read_row_record(off, size):
r = integer(wbk[off:off+2])
c1 = integer(wbk[off+2:off+4])
c2 = integer(wbk[off+4:off+6])
return r, c1, c2
def bin2data(s):
global wbk, ptr, fonts, xf
wbk = return_wbk(s)
ptr = 0
if wbk == "file not in correct format":
return wbk
# reading the workbook global substream
(f, off, size) = next_record_info() # BOF record
#####################################################
while f != 0x0031: # reaching to FONT Records
(f, off, size) = next_record_info()
fonts = []
while f == 0x0031: # reading font records
tpl = read_font_record(off, size)
fonts.append(tpl)
if len(fonts) == 4: # since font with index 4 has been ommited
fonts.append((0, 0, 0, 0)) # by excel :)
(f, off, size) = next_record_info()
while f != 0x00E0: # reaching to XF record
(f, off, size) = next_record_info()
xf = []
while f == 0x00E0: # reading XF records
fs = read_xf_record(off, size) # fn is formatting no.
xf.append(fs)
(f, off, size) = next_record_info()
#####################################################
while f != 0x0085: # reaching to SHEET Records
(f, off, size) = next_record_info()
sheets = [] # contains the info of all the worksheets
while f == 0x0085: # reading the SHEET records
record = wbk[off : off + size]
pos = integer(record[:4])
sheet_type = integer(record[5:6])
(name, bytes) = read_string(off + 6, 1)
if sheet_type == 0: # sheettype 0 for workssheet
sheets.append((pos, sheet_type, name))
(f, off, size) = next_record_info()
# building the SST shared string table
while f != 0x00FC: # reaching to SST record
(f, off, size) = next_record_info()
global sst
sst = [] # list of used strings
sst = read_sst_record(off, size)
workbook_data = []
workbook_col_widths = []
for sheet in sheets:
ptr = sheet[0] # offset of 1st worksheet
sheet_name = sheet[2]
next_record_info() # skipping the BOF record
(f, off, size) = next_record_info()
if f == 0x020B: # INDEX record
(rf, rl, dbcell_posns) = read_index_record(off, size)
###################################################
colinfo = []
def_width = 0
standard_width = 0
while f != 0x0200: # reaching to DIMENSION record
if f == 0x0055: # reading DEFCOLWIDTH
def_width = read_defcol_record(off, size)
if f == 0x007D: # reading COLINFO record
cnf = read_colinfo_record(off, size)
colinfo.append(cnf)
(f, off, size) = next_record_info()
#####################################################
rf, rl, cf, cl = read_dimension_record(off, size)
if (rl - rf) % 32 == 0:
row_block_count = (rl - rf) / 32
else:
row_block_count = int((rl - rf) / 32) + 1
if row_block_count == 0: # sheet is empty
continue
# reading the data now
sheet_data = [[()] * cl] * rl # empty database
for i in range(0, row_block_count):
(f, off, size) = next_record_info()
while f == 0x0208:
(f, off, size) = next_record_info() # skipping the ROW record
while f != 0x00D7: # f==0x00D7 for DBCELL record
if f == 0x0205:
x, y, val, typ, fs = read_boolerr_record(off, size)
elif f == 0x0204:
x, y, val, typ, fs = read_label_record(off, size)
elif f == 0x00FD:
x, y, val, typ, fs = read_labelsst_record(off, size)
elif f == 0x0203:
x, y, val, typ, fs = read_float_record(off, size)
elif f == 0x027E:
x, y, val, typ, fs = read_rk_record(off, size)
elif f == 0x0006: # formula record
x, y, val, fs, typ, formula = read_formula_record(off, size)
buf = list(sheet_data[x])
buf[y] = (val, "for", fs, typ, formula)
sheet_data[x] = buf
(f, off, size) = next_record_info()
continue
elif f == 0x00BD:
mul_cell_data = read_mulrk_record(off, size) # list of tuples
for tple in mul_cell_data:
x, y, val, typ, fs = tple
buf = list(sheet_data[x])
buf[y] = (val, typ, fs)
sheet_data[x] = buf
(f, off, size) = next_record_info()
continue
else:
(f, off, size) = next_record_info()
continue
buf = list(sheet_data[x])
buf[y] = (val, typ, fs)
sheet_data[x] = buf
(f, off, size) = next_record_info()
workbook_data.append(sheet_data)
#########################################################
while f != 0x000A: # 000A for EOF record
if f == 0x0099: # standardwidth record
standard_width = read_standardwidth_record(off, size)
(f, off, size) = next_record_info()
col_widths = [0] * cl
for tpl in colinfo:
for j in range(tpl[0], tpl[1] + 1):
col_widths[j] = tpl[2]
if standard_width != 0:
default = standard_width
else:
default = def_width
for j in range(0, cl):
if col_widths[j] == 0:
col_widths[j] = default
workbook_col_widths.append(col_widths)
###########################################################
return workbook_data_to_scalc_string(workbook_data, workbook_col_widths, 0)
def read_defcol_record(off, size):
return integer(wbk[off:off+2])
def read_blank_record(off, size):
r = integer(wbk[off : off+2])
c = integer(wbk[off+2 : off+4])
return r, c, '', 's'
def read_dimension_record(off, size):
rf = integer(wbk[off : off + 4])
rl = integer(wbk[off + 4 : off + 8])
cf = integer(wbk[off + 8 : off + 10])
cl = integer(wbk[off + 10 : off + 12])
return rf, rl, cf, cl
def read_dbcell_record(off, size):
l=[]
for i in range(0, (size-4)/2):
l.append(integer(wbk[off+4+2*i : off+4+2*i+2]))
return l
def read_dbcell_record(off, size):
l = []
for i in range(0, (size - 4) / 2):
l.append(integer(wbk[off + 4 + 2 * i : off + 4 + 2 * i + 2]))
return l
def read_formula(s):
rpn=[] # rpn array
add_list=[]
size=integer(s[:2])
pos = 2
while pos != 2+size:
tid = integer(s[pos])
if tid>=0x00 and tid<=0x02: return 'UNKNOWN'
elif tid>=0x03 and tid<=0x16:
rpn+=[token_list[tid]]
pos+=1
elif tid==0x17: # string
st, bytes = read_string(s, pos+1, 1)
pos=pos+1+bytes
rpn+=['"'+st+'"']
elif tid==0x1C: # error value
er=err_value[integer(s[pos+1])]
rpn+=[er]
pos+=2
elif tid==0x1D: # bool
nt=integer(s[pos+1])
if nt==0: rpn+=['FALSE']
else: rpn+=['TRUE']
pos+=2
elif tid==0x1E: # integer
rpn+=[str(integer(s[pos+1:pos+3]))]
pos+=3
elif tid==0x1F: # float
fv = floating(hexa(s[pos+1:pos+9]))
pos+=9
rpn+=[str(fv)]
elif tid==0x21 or tid==0x41 or tid==0x61: # const argc function name
( fn, argc )= fn_list[integer(s[pos+1:pos+3])]
rpn+=[(fn, argc)]
pos+=3
elif tid==0x22 or tid==0x42 or tid==0x62: # variable argc function name
try:
fn, argc = fn_list[integer(s[pos+2:pos+4])]
except:
return 'UNKNOWN'
argc=integer(s[pos+1])
rpn+=[(fn, argc)]
pos+=4
elif tid == 0x24 or tid==0x44 or tid==0x64: # tRef
rpn+=[read_cell_address(s[pos+1:pos+5])]
pos+=5
elif tid == 0x25 or tid==0x45 or tid==0x65: # tArea
rpn+=[read_cellrange_address(s[pos+1:pos+9])]
pos+=9
elif tid == 0x20 or tid == 0x40 or tid == 0x60: # tArray
add_list+=[ (len(rpn), tid) ]
pos+=8
elif tid == 0x26 or tid == 0x46 or tid == 0x66:
pos+=7
elif tid == 0x19:
bytes, t = skip_attr(s, pos)
pos += bytes
if t==0x10:
rpn+=[('SUM', 1)] # tAttrSum replaces the tFuncVar token here
else:
return 'UNKNOWN' # has to take care of some other tokens too
for i in range(0, len(add_list)): # additional data
index, tid = add_list[i]
dt, bytes = read_additional(s, pos, tid)
pos+=bytes
rpn.insert(index, dt)
l=len(rpn)
stk=[]
for i in range(0,l): # converting RPN to formula using stack
if rpn[i] in token_list[:18]:
t2=stk.pop()
t1=stk.pop()
stk+=[t1+rpn[i]+t2]
elif rpn[i]=='u+' or rpn[i]=='u-':
t1=stk.pop()
ch=rpn[i]
stk+=[ch[1]+t1]
elif rpn[i]=='%':
t1=stk.pop()
stk+=[t1+'%']
elif rpn[i]=='()':
t1=stk.pop()
stk+=['('+t1+')']
elif type(rpn[i])==tuple:
fname, argc = rpn[i]
fnstr=')'
for j in range(0, argc):
fnstr=',' + stk.pop() + fnstr
if argc != 0:
fnstr=fnstr[1:]
fnstr= fname+'('+fnstr
stk+=[fnstr]
else:
stk+=[rpn[i]]
return stk[0]
