def parse(self, filename):
try:
workbook = openpyxl.reader.excel.load_workbook(filename)
except:
print "Loading %s failed" % filename
sys.exit()
for sheetname in workbook.get_sheet_names():
if sheetname == "Legend-Key":
continue
obsclass = self.observationclasses.setdefault(sheetname, SDTMObservationClass(sheetname))
sheet = workbook.get_sheet_by_name(sheetname)
for row in sheet.rows:
# header row
if si(row[0]).startswith("SDTM v3.1"):
for (col_idx, col) in enumerate(row):
if col_idx >= 4:
obsclass.add_domain(col_idx, si(col))
elif si(row[1]) == "":
continue
else:
for (col_idx, col) in enumerate([si(x) for x in row]):
if obsclass.domains.get(col_idx):
if col.lower() != "not used":
# add the column to the dataset
obsclass.domains.get(col_idx).add_variable(row[0],
row[1],
row[2],
row[3],
col)
print "Loaded %s" % sheetname
for obclass in self.observationclasses.itervalues():
for domain in obclass.domains.itervalues():
print "%s: %s (%s variables)" % (obclass.name, domain.name, len(domain.variables))
def load_files(self, path):
# the SHARE content exists in a number of sheets
if not os.path.exists(path):
print "No such path %s" % path
sys.exit()
for document in glob.glob(os.path.join(path, "*.xlsx")):
if not re.match("^[^~](.*) Template.xlsx$", os.path.basename(document)):
continue
print "Checking %s" % document
try:
workbook = openpyxl.reader.excel.load_workbook(os.path.join(path, document))
except Exception, exc:
print "Loading %s failed: %s" % (document, exc)
sys.exit()
for sheetname in workbook.get_sheet_names():
if not re.match("^Generic ([A-Z\-\s]+) Template$", sheetname.strip()):
continue
sheet = workbook.get_sheet_by_name(sheetname)
domain_or_class = None
# set this to something large
start_offset = 99
for (row_idx, row) in enumerate(sheet.rows):
if si(row[0]) == "Domain":
for col in [si(x) for x in row[1:]]:
if col != "":
if "GENERIC" in col.upper():
domain_or_class = SDTMObservationClass(col)
else:
domain_or_class = SDTMDomain(col)
elif si(row[0]) == "Variable Name":
start_offset = row_idx
elif (row_idx > start_offset):
if si(row[0]) != "":
domain_or_class.add_row(dict(zip(COLUMNS, [si(x) for x in row])))
else:
self.domains[domain_or_class.name] = domain_or_class