def set_formula(self, addr, formula):
if addr in self.cellmap:
cell = self.cellmap[addr]
else:
raise Exception('Cell %s not in cellmap' % addr)
seeds = [cell]
if cell.is_range:
for index, c in enumerate(cell.range.cells): # for each cell of the range, translate the formula
if index == 0:
c.formula = formula
translator = Translator(unicode('=' + formula), c.address().split('!')[1]) # the Translator needs a reference without sheet
else:
translated = translator.translate_formula(c.address().split('!')[1]) # the Translator needs a reference without sheet
c.formula = translated[1:] # to get rid of the '='
seeds.append(c)
else:
cell.formula = formula
cellmap, G = graph_from_seeds(seeds, self)
self.cellmap = cellmap
self.G = G
should_eval = self.cellmap[addr].should_eval
self.cellmap[addr].should_eval = 'always'
self.evaluate(addr)
self.cellmap[addr].should_eval = should_eval
print("Graph construction updated, %s nodes, %s edges, %s cellmap entries" % (len(G.nodes()),len(G.edges()),len(cellmap)))
def set_formula(self, addr, formula):
if addr in self.cellmap:
cell = self.cellmap[addr]
else:
raise Exception('Cell %s not in cellmap' % addr)
seeds = [cell]
if cell.is_range:
for index, c in enumerate(cell.range.cells): # for each cell of the range, translate the formula
if index == 0:
c.formula = formula
translator = Translator(unicode('=' + formula), c.address().split('!')[1]) # the Translator needs a reference without sheet
else:
translated = translator.translate_formula(c.address().split('!')[1]) # the Translator needs a reference without sheet
c.formula = translated[1:] # to get rid of the '='
seeds.append(c)
else:
cell.formula = formula
cellmap, G = graph_from_seeds(seeds, self)
self.cellmap = cellmap
self.G = G
should_eval = self.cellmap[addr].should_eval
self.cellmap[addr].should_eval = 'always'
self.evaluate(addr)
self.cellmap[addr].should_eval = should_eval
print "Graph construction updated, %s nodes, %s edges, %s cellmap entries" % (len(G.nodes()),len(G.edges()),len(cellmap))
def read_cells(archive, ignore_sheets=[], ignore_hidden=False):
global debug
print('___### Reading Cells from XLSX ###___')
cells = {}
functions = set()
cts = dict(read_content_types(archive))
strings_path = cts.get(
SHARED_STRINGS
) # source: https://bitbucket.org/openpyxl/openpyxl/src/93604327bce7aac5e8270674579af76d390e09c0/openpyxl/reader/excel.py?at=default&fileviewer=file-view-default
if strings_path is not None:
if strings_path.startswith("/"):
strings_path = strings_path[1:]
shared_strings = read_string_table(archive.read(strings_path))
else:
shared_strings = []
for sheet in detect_worksheets(archive):
sheet_name = sheet['title']
function_map = {}
if sheet_name in ignore_sheets: continue
root = fromstring(
archive.read(sheet['path'])
) # it is necessary to use cElementTree from xml module, otherwise root.findall doesn't work as it should
hidden_cols = False
nb_hidden = 0
if ignore_hidden:
hidden_col_min = None
hidden_col_max = None
for col in root.findall('.//{%s}cols/*' % SHEET_MAIN_NS):
if 'hidden' in col.attrib and col.attrib['hidden'] == '1':
hidden_cols = True
hidden_col_min = int(col.attrib['min'])
hidden_col_max = int(col.attrib['max'])
for c in root.findall('.//{%s}c/*/..' % SHEET_MAIN_NS):
cell_data_type = c.get('t',
'n') # if no type assigned, assign 'number'
cell_address = c.attrib['r']
skip = False
if hidden_cols:
found = re.search(CELL_REF_RE, cell_address)
col = col2num(found.group(1))
if col >= hidden_col_min and col <= hidden_col_max:
nb_hidden += 1
skip = True
if not skip:
cell = {
'a': '%s!%s' % (sheet_name, cell_address),
'f': None,
'v': None
}
if debug: print('Cell', cell['a'])
for child in c:
child_data_type = child.get(
't', 'n') # if no type assigned, assign 'number'
if child.tag == '{%s}f' % SHEET_MAIN_NS:
if 'ref' in child.attrib: # the first cell of a shared formula has a 'ref' attribute
if debug:
print(
'*** Found definition of shared formula ***',
child.text, child.attrib['ref'])
if "si" in child.attrib:
function_map[child.attrib['si']] = (
child.attrib['ref'],
Translator(str('=' + child.text),
cell_address)
) # translator of openpyxl needs a unicode argument that starts with '='
# else:
# print "Encountered cell with ref but not si: ", sheet_name, child.attrib['ref']
if child_data_type == 'shared':
if debug:
print(
'*** Found child %s of shared formula %s ***'
% (cell_address, child.attrib['si']))
ref = function_map[child.attrib['si']][0]
formula = function_map[child.attrib['si']][1]
translated = formula.translate_formula(
cell_address)
cell['f'] = translated[
1:] # we need to get rid of the '='
else:
cell['f'] = child.text
elif child.tag == '{%s}v' % SHEET_MAIN_NS:
if cell_data_type == 's' or cell_data_type == 'str': # value is a string
try: # if it fails, it means that cell content is a string calculated from a formula
cell['v'] = shared_strings[int(child.text)]
except:
cell['v'] = child.text
elif cell_data_type == 'b':
cell['v'] = bool(int(child.text))
elif cell_data_type == 'n':
cell['v'] = _cast_number(child.text)
elif child.text is None:
continue
if cell['f'] is not None:
pattern = re.compile(r"([A-Z][A-Z0-9]*)\(")
found = re.findall(pattern, cell['f'])
map(lambda x: functions.add(x), found)
if cell['f'] is not None or cell['v'] is not None:
should_eval = 'always' if cell[
'f'] is not None and 'OFFSET' in cell['f'] else 'normal'
# cleaned_formula = cell['f']
cleaned_formula = cell['f'].replace(
", ", ",") if cell['f'] is not None else None
if "!" in cell_address:
cells[cell_address] = Cell(cell_address,
sheet_name,
value=cell['v'],
formula=cleaned_formula,
should_eval=should_eval)
else:
cells[sheet_name + "!" + cell_address] = Cell(
cell_address,
sheet_name,
value=cell['v'],
formula=cleaned_formula,
should_eval=should_eval)
if nb_hidden > 0:
print('Ignored %i hidden cells in sheet %s' %
(nb_hidden, sheet_name))
print('Nb of different functions %i' % len(functions))
print(functions)
for f in functions:
if f not in existing:
print('== Missing function: %s' % f)
return cells