def test_detect_dialect_using_json(self):
temp = tempfile.NamedTemporaryFile(delete=False)
filename = '{}.{}'.format(temp.name, self.file_extension)
encoding = 'utf-8'
self.files_to_delete.append(filename)
# Using JSON will force the sniffer to do not include ':', '}' in the
# possible delimiters
table = rows.Table(fields=OrderedDict([
('jsoncolumn1', rows.fields.JSONField),
('jsoncolumn2', rows.fields.JSONField),
]))
table.append({
'jsoncolumn1': '{"a": 42}',
'jsoncolumn2': '{"b": 43}',
})
table.append({
'jsoncolumn1': '{"c": 44}',
'jsoncolumn2': '{"d": 45}',
})
rows.export_to_csv(table, filename, encoding=encoding)
table = rows.import_from_csv(filename, encoding=encoding)
self.assertEqual(table.field_names, ['jsoncolumn1', 'jsoncolumn2'])
self.assertDictEqual(table[0].jsoncolumn1, {'a': 42})
self.assertDictEqual(table[0].jsoncolumn2, {'b': 43})
self.assertDictEqual(table[1].jsoncolumn1, {'c': 44})
self.assertDictEqual(table[1].jsoncolumn2, {'d': 45})
def parse_state_file(self, response):
state = response.meta["state"]
self.errors = []
try:
self.parse_boletim(state, response.body)
except Exception as exp:
self.errors.append(
("boletim", state, f"{exp.__class__.__name__}: {exp}"))
try:
self.parse_caso(state, response.body)
except Exception as exp:
self.errors.append(
("caso", state, f"{exp.__class__.__name__}: {exp}"))
if self.errors:
error_counter = Counter(error[0] for error in self.errors)
error_counter_str = ", ".join(
f"{error_type}: {count}"
for error_type, count in error_counter.items())
self.logger.error(
f"{len(self.errors)} errors found when parsing {state} ({error_counter_str})"
)
error_header = ("sheet", "state", "message")
errors = rows.import_from_dicts(
[dict(zip(error_header, row)) for row in self.errors])
rows.export_to_csv(errors, f"errors-{state}.csv")
exit(255)
def test_quotes(self):
temp = tempfile.NamedTemporaryFile(delete=False)
filename = '{}.{}'.format(temp.name, self.file_extension)
self.files_to_delete.append(filename)
table = rows.Table(fields=OrderedDict([
('field_1', rows.fields.TextField),
('field_2', rows.fields.TextField),
('field_3', rows.fields.TextField),
('field_4', rows.fields.TextField), ]))
table.append({
'field_1': '"quotes"',
'field_2': 'test "quotes"',
'field_3': '"quotes" test',
'field_4': 'test "quotes" test',
})
# we need this line row since `"quotes"` on `field_1` could be
# `JSONField` or `TextField`
table.append({
'field_1': 'noquotes',
'field_2': 'test "quotes"',
'field_3': '"quotes" test',
'field_4': 'test "quotes" test',
})
rows.export_to_csv(table, filename)
table2 = rows.import_from_csv(filename)
self.assert_table_equal(table, table2)
def main():
parser = argparse.ArgumentParser()
parser.add_argument("state", choices=spiders.keys())
parser.add_argument("--year", type=int)
parser.add_argument("--action", type=int)
parser.add_argument("--quiet", action="store_true")
parser.add_argument("--headless", action="store_true")
args = parser.parse_args()
if args.year is None and args.action is None:
actions = get_actions_for_state(args.state)
elif args.action is None:
actions = [
action
for action in get_actions_for_state(args.state)
if action.year == args.year
]
else:
actions = [
Action(year=args.year, code=args.action, state=args.state, name="Unknown")
]
spider = spiders[args.state](headless=args.headless)
for action in actions:
if not args.quiet:
print(
f"Downloading budget execution for {action.state} ({action.code} @ {action.year})"
)
table = spider.execute(action.year, action.code)
output_filename = f"{action.state}-{action.year}-{action.code}.csv"
rows.export_to_csv(table, output_filename)
if not args.quiet:
print(f" done (saved to {output_filename})")
spider.close()
def test_detect_dialect_using_json(self):
temp = tempfile.NamedTemporaryFile(delete=False)
filename = '{}.{}'.format(temp.name, self.file_extension)
encoding = 'utf-8'
self.files_to_delete.append(filename)
# Using JSON will force the sniffer to do not include ':', '}' in the
# possible delimiters
table = rows.Table(fields=OrderedDict([
('jsoncolumn1', rows.fields.JSONField),
('jsoncolumn2', rows.fields.JSONField),
]))
table.append({
'jsoncolumn1': '{"a": 42}',
'jsoncolumn2': '{"b": 43}',
})
table.append({
'jsoncolumn1': '{"c": 44}',
'jsoncolumn2': '{"d": 45}',
})
rows.export_to_csv(table, filename, encoding=encoding)
table = rows.import_from_csv(filename, encoding=encoding)
self.assertEqual(table.field_names, ['jsoncolumn1', 'jsoncolumn2'])
self.assertDictEqual(table[0].jsoncolumn1, {'a': 42})
self.assertDictEqual(table[0].jsoncolumn2, {'b': 43})
self.assertDictEqual(table[1].jsoncolumn1, {'c': 44})
self.assertDictEqual(table[1].jsoncolumn2, {'d': 45})
def parse_state_file(self, response):
state = response.meta["state"]
self.errors = []
try:
self.parse_boletim(state, response.body)
except Exception as exp:
self.errors.append(
("boletim", state, f"{exp.__class__.__name__}: {exp}"))
try:
self.parse_caso(state, response.body)
except Exception as exp:
self.errors.append(
("caso", state, f"{exp.__class__.__name__}: {exp}"))
if self.errors:
error_counter = Counter(error[0] for error in self.errors)
error_counter_str = ", ".join(
f"{error_type}: {count}"
for error_type, count in error_counter.items())
self.logger.error(
f"{len(self.errors)} errors found when parsing {state} ({error_counter_str})"
)
error_header = ("sheet", "state", "message")
errors = rows.import_from_dicts(
[dict(zip(error_header, row)) for row in self.errors])
filename = ERROR_PATH / f"errors-{state}.csv"
if not filename.parent.exists():
filename.parent.mkdir(parents=True)
rows.export_to_csv(errors, filename)
# Force crawler to stop
os.kill(os.getpid(), SIGINT)
os.kill(os.getpid(), SIGINT)
raise CloseSpider(f"Error found on {state} (see {filename}).")
def test_quotes(self):
temp = tempfile.NamedTemporaryFile(delete=False)
filename = '{}.{}'.format(temp.name, self.file_extension)
self.files_to_delete.append(filename)
table = rows.Table(fields=OrderedDict([
('field_1', rows.fields.TextField),
('field_2', rows.fields.TextField),
('field_3', rows.fields.TextField),
('field_4', rows.fields.TextField),
]))
table.append({
'field_1': '"quotes"',
'field_2': 'test "quotes"',
'field_3': '"quotes" test',
'field_4': 'test "quotes" test',
})
# we need this line row since `"quotes"` on `field_1` could be
# `JSONField` or `TextField`
table.append({
'field_1': 'noquotes',
'field_2': 'test "quotes"',
'field_3': '"quotes" test',
'field_4': 'test "quotes" test',
})
rows.export_to_csv(table, filename)
table2 = rows.import_from_csv(filename)
self.assert_table_equal(table, table2)
def create_download_script(filename=Path('data/download.sh'),
output_path=Path('data/output'),
download_path=Path('data/download')):
if not filename.parent.exists():
filename.parent.mkdir()
if not output_path.exists():
output_path.mkdir()
links = discover_links()
today = datetime.datetime.now()
date = f'{today.year}-{today.month:02d}-{today.day:02d}'
rows.export_to_csv(links, output_path / f'links_{date}.csv')
with open(filename, mode='w', encoding='utf8') as fobj:
fobj.write('#!/bin/sh\n')
fobj.write(
f'# Arquivo gerado em {today.year}-{today.month}-{today.day}\n')
fobj.write(
'# Visite o site da Receita Federal para verificar se existem atualizações.\n\n'
)
fobj.write('mkdir -p {}\n\n'.format(str(download_path)))
for row in links:
path = download_path / (row.uf + '.txt')
fobj.write(f'wget -O "{path}" "{row.url}"\n')
meta = os.stat(filename)
os.chmod(filename, meta.st_mode | stat.S_IEXEC)
def download_and_save(names, filename):
result = []
for name in names:
print(name)
result.append(classify_by_sex(name))
table = rows.import_from_dicts(result)
rows.export_to_csv(table, filename)
def parse_state_file(self, response):
meta = response.meta
state = meta["state"]
caso_filename = meta["caso_filename"]
if response.status >= 400:
self.errors[state].append(
("connection", state, f"HTTP status code: {response.status}"))
else:
response_data = json.load(io.BytesIO(response.body))
try:
self.parse_boletim(state, response_data["reports"])
except Exception as exp:
self.errors[state].append(
("boletim", state, f"{exp.__class__.__name__}: {exp}"))
try:
self.parse_caso(state, caso_filename, response_data["cases"])
except Exception as exp:
self.errors[state].append(
("caso", state, f"{exp.__class__.__name__}: {exp}"))
if self.errors[state]:
error_counter = Counter(error[0] for error in self.errors[state])
error_counter_str = ", ".join(
f"{error_type}: {count}"
for error_type, count in error_counter.items())
self.logger.error(
f"{len(self.errors[state])} errors found when parsing {state} ({error_counter_str})"
)
error_header = ("sheet", "state", "message")
errors = rows.import_from_dicts(
[dict(zip(error_header, row)) for row in self.errors[state]])
filename = ERROR_PATH / f"errors-{state}.csv"
if not filename.parent.exists():
filename.parent.mkdir(parents=True)
rows.export_to_csv(errors, filename)
def main():
parser = argparse.ArgumentParser()
parser.add_argument('html_entrada')
parser.add_argument('csv_saida')
args = parser.parse_args()
table = sum_iof_into_entries(html_to_table(args.html_entrada))
rows.export_to_csv(table, args.csv_saida)
def test_export_to_csv_filename(self):
# TODO: may test file contents
temp = tempfile.NamedTemporaryFile(delete=False)
self.files_to_delete.append(temp.name)
rows.export_to_csv(utils.table, temp.name)
table = rows.import_from_csv(temp.name)
self.assert_table_equal(table, utils.table)
def main():
parser = argparse.ArgumentParser()
parser.add_argument('html_entrada')
parser.add_argument('csv_saida')
args = parser.parse_args()
table = sum_iof_into_entries(html_to_table(args.html_entrada))
rows.export_to_csv(table, args.csv_saida)
def create_final_headers(header_type, order_columns, final_filename):
final_headers = {}
filenames = sorted(
[
(REGEXP_HEADER_YEAR.findall(filename)[0], filename)
for filename in glob(str(settings.HEADERS_PATH / f"{header_type}-*.csv"))
if REGEXP_HEADER_YEAR.findall(filename)
]
)
# TODO: check if schema is according to final header. if there are diffs,
# warn user.
for index, (header_year, filename) in enumerate(filenames):
header = read_header(filename)
for row in header:
if not row.nome_final:
continue
if row.nome_final not in final_headers:
row_data = row._asdict()
if index > 0:
row_data["introduced_on"] = header_year
row_data["original_names"] = [(header_year, row_data.pop("nome_tse"))]
final_headers[row.nome_final] = row_data
else:
original_name = (header_year, row.nome_tse)
original_names = final_headers[row.nome_final]["original_names"]
should_add = True
for original in original_names:
if original[1] == original_name[1]:
should_add = False
break
if should_add:
original_names.append(original_name)
table = rows.Table(
fields=OrderedDict(
[
("nome_final", rows.fields.TextField),
("descricao", rows.fields.TextField),
]
)
)
header_list = sorted(
final_headers.values(), key=lambda row: order_columns(row["nome_final"])
)
for row in header_list:
row_data = {"descricao": row["descricao"] or "", "nome_final": row["nome_final"]}
introduced_on = row.get("introduced_on", None)
original_names = ", ".join(
f"{item[1]} ({item[0]})" for item in row.get("original_names")
)
row_data["descricao"] += f". Aparece no TSE como: {original_names}"
if introduced_on:
row_data["descricao"] += f". Coluna adicionada em {introduced_on}"
if row_data["descricao"][-1] != ".":
row_data["descricao"] += "."
table.append(row_data)
rows.export_to_csv(table, final_filename)
def test_export_callback(self):
table = rows.import_from_dicts([{
'id': number
} for number in range(10)])
myfunc = mock.Mock()
rows.export_to_csv(table, callback=myfunc, batch_size=3)
self.assertEqual(myfunc.call_count, 4)
self.assertEqual([x[0][0] for x in myfunc.call_args_list],
[3, 6, 9, 10])
def write_csv(self):
filename = OUTPUT_PATH / Path("caso-sc.csv")
# Adicionando a 'row' do estado de SC
data = {}
data["date"] = date.today()
data["state"] = "SC"
data["city"] = ""
data["place_type"] = "state"
data["notified"] = sum([self.data_cities[i]["notified"] for i in range(len(self.data_cities))])
data["confirmed"] = sum([self.data_cities[i]["confirmed"] for i in range(len(self.data_cities))])
data["discarded"] = sum([self.data_cities[i]["discarded"] for i in range(len(self.data_cities))])
data["suspect"] = sum([self.data_cities[i]["suspect"] for i in range(len(self.data_cities))])
data["deaths"] = sum([self.data_cities[i]["deaths"] for i in range(len(self.data_cities))])
data["city_ibge_code"] = ""
data["estimated_population_2019"] = sum([self.data_cities[i]["estimated_population_2019"] for i in range(len(self.data_cities))])
data["confirmed_per_100k_inhabitants"] = sum([self.data_cities[i]["confirmed_per_100k_inhabitants"] for i in range(len(self.data_cities))])
data["death_rate"] = sum([self.data_cities[i]["death_rate"] for i in range(len(self.data_cities))]) / len(self.data_cities)
data["notes"] = ""
data["source_url"] = ""
self.data_cities.append(data)
# Adicionando as cidades que nao foram parseadas ainda
for each_city in self.cities_sc_ibge:
if each_city.municipio not in self.cidades_url.keys():
data = {}
data["date"] = ""
data["state"] = "SC"
data["city"] = each_city.municipio
data["place_type"] = "city"
data["notified"] = ""
data["confirmed"] = ""
data["discarded"] = ""
data["suspect"] = ""
data["deaths"] = ""
data["notes"] = ""
data["city_ibge_code"] = ""
data["estimated_population_2019"] = ""
data["confirmed_per_100k_inhabitants"] = ""
data["death_rate"] = ""
data["source_url"] = ""
self.data_cities.append(data)
# keyorder = ['date', 'state', 'city', 'place_type', 'notified', 'confirmed', 'discarded', 'suspect', 'deaths', 'notes', 'city_ibge_code', 'estimated_population_2019', 'confirmed_per_100k_inhabitants', 'death_rate', 'source_url']
#
# data_cities_ordered = []
# for d in self.data_cities:
# data_cities_ordered.append(sorted(d.items(), key=lambda i:keyorder.index(i[0])))
rows_data = rows.import_from_dicts(self.data_cities)
rows_data.order_by("city")
rows.export_to_csv(rows_data, filename)
def test_import_field_limit(self):
temp = tempfile.NamedTemporaryFile(delete=False)
filename = "{}.{}".format(temp.name, self.file_extension)
self.files_to_delete.append(filename)
table = rows.import_from_dicts([{"f1": "a" * 132000}])
rows.export_to_csv(table, filename)
# The following line must not raise the exception:
# `_csv.Error: field larger than field limit (131072)`
new = rows.import_from_csv(filename)
def test_issue_168(self):
temp = tempfile.NamedTemporaryFile(delete=False)
filename = "{}.{}".format(temp.name, self.file_extension)
self.files_to_delete.append(filename)
table = rows.Table(fields=OrderedDict([("jsoncolumn", rows.fields.JSONField)]))
table.append({"jsoncolumn": '{"python": 42}'})
rows.export_to_csv(table, filename)
table2 = rows.import_from_csv(filename)
self.assert_table_equal(table, table2)
def export_prob_predictions_to_csv(self, filename, ids, predictions):
new_rows = []
# classifier = {
# 'Return_to_owner': 0,
# 'Euthanasia': 1,
# 'Adoption': 2,
# 'Transfer': 3,
# 'Died': 4
# }
#print(predictions)
#count = [0, 0, 0, 0]
#m = []
for i, prediction in enumerate(predictions):
#ID Adoption Died Euthanasia Return_to_owner Transfer
# print(type(prediction))
# print(prediction == '3')
# print(int(prediction == '3'))
new_row = OrderedDict({})
# print(prediction[0])
# print(type(prediction[0]))
# print numpy.argmax(prediction)
# m.append(numpy.argmax(prediction))
# if numpy.argmax(prediction) == 0:
# count[0]+=1
# if numpy.argmax(prediction) == 1:
# count[1]+=1
# if numpy.argmax(prediction) == 2:
# count[2]+=1
# if numpy.argmax(prediction) == 3:
# count[3]+=1
new_row['ID'] = ids[i]
new_row['Adoption'] = prediction[2]
new_row['Died'] = prediction[4]
new_row['Euthanasia'] = prediction[1]
new_row['Return_to_owner'] = prediction[0]
new_row['Transfer'] = prediction[3]
new_rows.append(new_row)
#print(count)
#print(set(m))
new_rows.sort(key=lambda e: e['ID'])
#print(new_rows)
new_fields = [(key, rows.fields.UnicodeField) for key in new_rows[0].keys()]
table_to = rows.Table(fields=OrderedDict(new_fields))
for row in new_rows:
table_to.append(row)
rows.export_to_csv(table_to, filename)
def test_export_to_csv_uses_serialize(self, mocked_serialize):
temp = tempfile.NamedTemporaryFile(delete=False)
self.files_to_delete.append(temp.name)
kwargs = {"test": 123, "parameter": 3.14}
mocked_serialize.return_value = iter([utils.table.fields.keys()])
rows.export_to_csv(utils.table, temp.name, encoding="utf-8", **kwargs)
self.assertTrue(mocked_serialize.called)
self.assertEqual(mocked_serialize.call_count, 1)
call = mocked_serialize.call_args
self.assertEqual(call[0], (utils.table,))
self.assertEqual(call[1], kwargs)
def test_export_to_csv_filename(self):
# TODO: may test file contents
temp = tempfile.NamedTemporaryFile(delete=False)
self.files_to_delete.append(temp.name)
rows.export_to_csv(utils.table, temp.name)
table = rows.import_from_csv(temp.name)
self.assert_table_equal(table, utils.table)
temp.file.seek(0)
result = temp.file.read()
export_in_memory = rows.export_to_csv(utils.table, None)
self.assertEqual(result, export_in_memory)
def test_export_to_csv_filename(self):
# TODO: may test file contents
temp = tempfile.NamedTemporaryFile(delete=False)
self.files_to_delete.append(temp.name)
rows.export_to_csv(utils.table, temp.name)
table = rows.import_from_csv(temp.name)
self.assert_table_equal(table, utils.table)
temp.file.seek(0)
result = temp.file.read()
export_in_memory = rows.export_to_csv(utils.table, None)
self.assertEqual(result, export_in_memory)
def test_export_to_csv_uses_serialize(self, mocked_serialize):
temp = tempfile.NamedTemporaryFile(delete=False)
self.files_to_delete.append(temp.name)
kwargs = {'test': 123, 'parameter': 3.14, }
mocked_serialize.return_value = iter([utils.table.fields.keys()])
rows.export_to_csv(utils.table, temp.name, encoding='utf-8', **kwargs)
self.assertTrue(mocked_serialize.called)
self.assertEqual(mocked_serialize.call_count, 1)
call = mocked_serialize.call_args
self.assertEqual(call[0], (utils.table, ))
self.assertEqual(call[1], kwargs)
def main():
now = datetime.datetime.now()
today = datetime.date(now.year, now.month, now.day)
download_path = pathlib.Path('download')
output_path = pathlib.Path('output')
if not download_path.exists():
download_path.mkdir()
if not output_path.exists():
output_path.mkdir()
# Get spreadsheet links
links = get_links(date=today)
rows.export_to_csv(links, output_path / f'links-{today}.csv')
# Download all the links
result = []
for link in links:
print(link.name)
filename = download_path / urlparse(link.url).path.split('/')[-1]
# Download file
print(f' Downloading ({link.url})...', end='', flush=True)
if filename.exists():
print(f' already downloaded.')
else:
try:
download(link.url, filename)
except RuntimeError as exception:
print(f' {exception.args[0]}')
continue
else:
print(' done.')
# Extract data
print(f' Extracting ({filename})...', end='', flush=True)
try:
data = extract(filename, link)
except Exception as exp:
import traceback
print(f' ERROR! {traceback.format_exc().splitlines()[-1]}')
else:
print(f' done (rows extracted: {len(data)}).')
result.extend(data)
# Extract everything to a new CSV
output = output_path / f'salarios-magistrados-{today}.csv'
print(f'Extracting result to {output}...')
export_csv(result, output)
def pdf_to_csv(input_filename, output_filename):
total_pages = rows.plugins.pdf.number_of_pages(input_filename)
pdf = rows.plugins.pdf.PyMuPDFBackend(input_filename)
result = []
for page_number in range(1, total_pages + 1):
page = list(next(pdf.objects(page_numbers=(page_number, ))))
data = list(rows.plugins.utils.ipartition(page, 4))
header = [obj.text for obj in data[0]]
for row in data[1:]:
row = dict(zip(header, [obj.text for obj in row]))
row["codigo_ibge"] = row.pop("IBGE")
row["perfil"] = row.pop("Perfil Município")
result.append(row)
table = rows.import_from_dicts(result)
rows.export_to_csv(table, output_filename)
def test_import_from_xpath_filename(self):
table = rows.import_from_xpath(self.filename,
encoding=self.encoding,
**self.kwargs)
expected_meta = {'imported_from': 'xpath', 'filename': self.filename,}
self.assertEqual(table.meta, expected_meta)
temp = tempfile.NamedTemporaryFile(delete=False)
self.files_to_delete.append(temp.name)
fobj = temp.file
rows.export_to_csv(table, fobj)
fobj.seek(0)
table = rows.import_from_csv(fobj)
self.assert_table_equal(table, self.expected_table)
def __call__(self):
view = getMultiAdapter((self.context, self.request), name='view')
table = view.table()
filename = "%s.csv" % view.filename_prefix()
data = rows.export_to_csv(table)
self.request.response.setHeader('Content-Type', '"%s"' % EXTENSIONS_TYPES.get('csv'))
self.request.response.setHeader('Content-Disposition', 'attachment; filename="%s"' % filename)
return data
def convert(state, input_filename, output_filename):
table = rows.import_from_csv(
input_filename,
force_types={
"confirmed": rows.fields.IntegerField,
"deaths": rows.fields.IntegerField,
},
)
state_cities = ["TOTAL NO ESTADO", "Importados/Indefinidos"] + sorted(
row.municipio for row in cities if row.uf == state
)
confirmed, deaths, dates = {}, {}, []
for row in table:
row_confirmed = row.confirmed or 0
row_date = row.date
row_deaths = row.deaths or 0
row_name = row.city if row.place_type == "city" else "TOTAL NO ESTADO"
if row_name not in state_cities:
print(f"ERRO: município {repr(row_name)} não encontrado.")
continue
if row_confirmed == 0 and row_deaths == 0:
# No data for this city in this day
continue
if row_date not in confirmed:
confirmed[row_date] = {}
if row_date not in deaths:
deaths[row_date] = {}
if row_name in confirmed[row_date] or row_name in deaths[row_date]:
print(f"ERRO: conflito em {repr(row_name)} para {row_date}.")
continue
confirmed[row_date][row_name] = row_confirmed
deaths[row_date][row_name] = row_deaths
result = []
dates = sorted(confirmed.keys(), reverse=True)
for city in state_cities:
row = {"municipio": city}
for date in dates:
date_str = f"{date.day:02d}_{date.month:02d}"
row[f"confirmados_{date_str}"] = confirmed[date].get(city, None)
row[f"mortes_{date_str}"] = deaths[date].get(city, None)
result.append(row)
rows.export_to_csv(rows.import_from_dicts(result), output_filename)
def main():
now = datetime.datetime.now()
today = datetime.date(now.year, now.month, now.day)
download_path = pathlib.Path('download')
output_path = pathlib.Path('output')
if not download_path.exists():
download_path.mkdir()
if not output_path.exists():
output_path.mkdir()
# Get spreadsheet links
links = get_links(date=today)
rows.export_to_csv(links, output_path / f'links-{today}.csv')
# Download all the links
filenames = []
for link in links:
save_path = download_path / urlparse(link.url).path.split('/')[-1]
filenames.append(save_path)
if not save_path.exists():
print(f'Downloading {link.url}...', end='', flush=True)
download(link.url, save_path)
print(' done.')
else:
print(f'Skipping {save_path.name}...')
# Extract data from all the spreadsheets
result = []
for filename in filenames:
print(f'Extracting {filename.name}...', end='', flush=True)
try:
data = extract(filename)
except Exception as exp:
import traceback
print(f' ERROR! {traceback.format_exc().splitlines()[-1]}')
else:
print(' done.')
result.extend(data)
# Extract everything to a new CSV
output = output_path / f'salarios-magistrados-{today}.csv'
print(f'Extracting result to {output}...')
export_csv(result, output)
def test_import_from_xpath_fobj(self):
# TODO: may test with codecs.open passing an encoding
with open(self.filename, mode='rb') as fobj:
table = rows.import_from_xpath(fobj,
encoding=self.encoding,
**self.kwargs)
expected_meta = {'imported_from': 'xpath',
'filename': self.filename,
'encoding': self.encoding, }
self.assertEqual(table.meta, expected_meta)
temp = tempfile.NamedTemporaryFile(delete=False)
self.files_to_delete.append(temp.name)
fobj = temp.file
rows.export_to_csv(table, fobj)
fobj.seek(0)
table = rows.import_from_csv(fobj)
self.assert_table_equal(table, self.expected_table)
def download_game_data_for_country(path, year, country_code):
'Download country athlete data for a specific year if not downloaded yet'
filename = path.joinpath(_make_filename(year, country_code))
if filename.exists():
print(' (already downloaded, skipping)')
return
url = URL_DATA.format(year=year, country_code=country_code)
response = requests.get(url)
if '404' in response.url: # country didn't played this year
print(" (didn't play this year, skipping)")
return
html = response.content
table = rows.import_from_html(BytesIO(html),
encoding='utf-8',
fields=FIELDS)
rows.export_to_csv(table, str(filename.absolute()), encoding='utf-8')
print(' ok')
def export_exact_predictions_to_csv(self, filename, ids, predictions):
new_rows = []
# classifier = {
# 'Return_to_owner': 0,
# 'Euthanasia': 1,
# 'Adoption': 2,
# 'Transfer': 3,
# 'Died': 4
# }
for i, prediction in enumerate(predictions):
#ID Adoption Died Euthanasia Return_to_owner Transfer
# print(type(prediction))
# print(prediction == '3')
# print(int(prediction == '3'))
new_row = OrderedDict({})
new_row['ID'] = ids[i]
new_row['Adoption'] = int(prediction == '2')
new_row['Died'] = int(prediction == '4')
new_row['Euthanasia'] = int(prediction == '1')
new_row['Return_to_owner'] = int(prediction == '0')
new_row['Transfer'] = int(prediction == '3')
new_rows.append(new_row)
new_rows.sort(key=lambda e: e['ID'])
#print(new_rows)
new_fields = [(key, rows.fields.UnicodeField) for key in new_rows[0].keys()]
table_to = rows.Table(fields=OrderedDict(new_fields))
for row in new_rows:
table_to.append(row)
rows.export_to_csv(table_to, filename)
def parse_licitacao(year, city_code):
# TODO: adicionar codibge, nome municipio e ano nos 3 arquivos
filename = DOWNLOAD_PATH / f"{year}_{city_code}_Licitacao.zip"
result1 = parse(filename, "licitacao")
rows.export_to_csv(result1, OUTPUT_PATH / f"licitacao-{city_code}-{year}.csv")
result2 = parse(filename, "licitacao_participante")
rows.export_to_csv(
result2, OUTPUT_PATH / f"licitacao-participante-{city_code}-{year}.csv"
)
result3 = parse(filename, "licitacao_vencedor")
rows.export_to_csv(
result3, OUTPUT_PATH / f"licitacao-vencedor-{city_code}-{year}.csv"
)
if row.animaltype == 'Dog':
new_row.update(get_dog_age_columns(row))
# Demora muito
#new_row.update(get_dog_breed_columns(row))
#new_row.update(get_dog_color_columns(row))
new_row['outcome'] = get_animal_outcome(row)
new_dog_rows.append(new_row)
new_fields = [(key, rows.fields.UnicodeField) for key in new_cat_rows[0].keys()]
table_to = rows.Table(fields=OrderedDict(new_fields))
for row in new_cat_rows:
table_to.append(row)
rows.export_to_csv(table_to, "clean_data3_no_color_no_breed_cat.csv")
new_fields = [(key, rows.fields.UnicodeField) for key in new_dog_rows[0].keys()]
table_to = rows.Table(fields=OrderedDict(new_fields))
for row in new_dog_rows:
table_to.append(row)
rows.export_to_csv(table_to, "clean_data3_no_color_no_breed_dog.csv")
##################
# Limpando dados de teste
##################
table_from = rows.import_from_csv("../test.csv")
import requests
import rows
url = "http://balneabilidade.inema.ba.gov.br/index.php/relatoriodebalneabilidade/geraBoletim?idcampanha=42041"
print("*** Downloading PDF...")
response = requests.get(url)
# The line below will automatically identify the table in all PDF pages - it
# works for this file but not for all cases. You can be more specific defining
# the page numbers, a start/end string (like the header/footer strings) and
# also change the table identification algorithm. Check `backend`, `algorithm`,
# `starts_after`, `ends_before` and `page_numbers` parameters.
# For this simple case you could also install rows' CLI (`pip install
# rows[cli]`) and run: `rows print <url>`
table = rows.import_from_pdf(io.BytesIO(response.content))
rows.export_to_csv(table, "beach-data.csv")
print("*** Table exported to beach-data.csv")
print("*** Extracted table:")
print(rows.export_to_txt(table))
# You could also iterate over the object, like:
# for row in table: print(row)
print("\n\n*** Extracted text:")
text_pages = rows.plugins.pdf.pdf_to_text(io.BytesIO(response.content))
print("\n\n".join(text_pages))
# Get data from Portuguese Wikipedia
city_list_url = 'https://pt.wikipedia.org/wiki/Lista_de_munic%C3%ADpios_do_Brasil'
response = requests.get(city_list_url)
html = response.content
# Extract desired data using XPath
cities = rows.import_from_xpath(
BytesIO(html),
rows_xpath='//table/tr/td/ul/li',
fields_xpath=OrderedDict([('name', './/text()'),
('link', './/a/@href')]))
regexp_city_state = re.compile(r'(.*) \(([A-Z]{2})\)')
def transform(row, table):
'Transform row "link" into full URL and add "state" based on "name"'
data = row._asdict()
data['link'] = urlparse.urljoin('https://pt.wikipedia.org', data['link'])
data['name'], data['state'] = regexp_city_state.findall(data['name'])[0]
return data
new_fields = OrderedDict()
new_fields['name'] = cities.fields['name']
new_fields['state'] = rows.fields.TextField # new field
new_fields['link'] = cities.fields['link']
cities = rows.transform(new_fields, transform, cities)
rows.export_to_csv(cities, 'brazilian-cities.csv')
table_2 = rows.Table(fields=new_fields)
for row in table_1:
if row.sexuponoutcome and row.sexuponoutcome != u'Unknown':
castration, sex = row.sexuponoutcome.split()
else:
castration, sex = u'Unknown', u'Unknown'
week_day = calendar.day_name[row.datetime.weekday()]
us_holidays = holidays.UnitedStates()
holiday = row.datetime in us_holidays
age_in_days = get_animal_age(row.ageuponoutcome)
if row.animaltype == "Cat":
age_group = get_cat_age_group(age_in_days)
else:
age_group = get_dog_age_group(age_in_days)
table_2.append({'animalid': row.animalid,
'name': row.name,
'datetime': week_day,
'holiday': holiday,
'outcometype': row.outcometype,
'outcomesubtype': row.outcomesubtype,
'animaltype': row.animaltype,
'sex': sex,
'castration': castration,
'agegroup': age_group,
'breed': row.breed,
'color': row.color})
rows.export_to_csv(table_2, "clean_data.csv")
# coding: utf-8
from __future__ import unicode_literals
import os
from collections import OrderedDict
import rows
# taken from:
# http://www.supercom.gob.ec/es/informate-y-participa/directorio-de-medios/21-radiodifusoras
filename = os.path.join(
os.path.dirname(__file__), "../../tests/data/ecuador-medios-radiodifusoras.html"
)
rows_xpath = '//*[@class="entry-container"]/*[@class="row-fluid"]/*[@class="span6"]'
fields_xpath = OrderedDict(
[
("url", ".//h2/a/@href"),
("name", ".//h2/a/text()"),
("address", './/div[@class="spField field_direccion"]/text()'),
("phone", './/div[@class="spField field_telefono"]/text()'),
("website", './/div[@class="spField field_sitio_web"]/text()'),
("email", './/div[@class="spField field_email"]/text()'),
]
)
table = rows.import_from_xpath(filename, rows_xpath, fields_xpath)
rows.export_to_csv(table, "ecuador-radiodifusoras.csv")
# coding: utf-8
from __future__ import unicode_literals
import os
from collections import OrderedDict
import rows
# taken from:
# http://www.supercom.gob.ec/es/informate-y-participa/directorio-de-medios/21-radiodifusoras
filename = os.path.join(os.path.dirname(__file__),
'../../tests/data/ecuador-medios-radiodifusoras.html')
rows_xpath = '//*[@class="entry-container"]/*[@class="row-fluid"]/*[@class="span6"]'
fields_xpath = OrderedDict([
('url', './/h2/a/@href'),
('name', './/h2/a/text()'),
('address', './/div[@class="spField field_direccion"]/text()'),
('phone', './/div[@class="spField field_telefono"]/text()'),
('website', './/div[@class="spField field_sitio_web"]/text()'),
('email', './/div[@class="spField field_email"]/text()'), ])
table = rows.import_from_xpath(filename, rows_xpath, fields_xpath)
rows.export_to_csv(table, 'ecuador-radiodifusoras.csv')
drow = row.__dict__
for key, value in drow.items():
if key not in statistics:
statistics[key] = {}
if value not in statistics[key]:
statistics[key][value] = 0
statistics[key][value] += 1
string = rows.fields.UnicodeField
columns = {}
columns['value'] = string
for key in statistics.keys():
for value in statistics[key].keys():
columns[key + '_' + value] = string
table_2d_analize = rows.Table(fields=columns)
drows = map(lambda r: r.__dict__, table)
for key in statistics.keys():
for value in statistics[key].keys():
data = {}
data['value'] = key + '_' + value
for key2 in statistics.keys():
for value2 in statistics[key2].keys():
data[key2 + '_' + value2] = len(filter(lambda d: d[key] == value and d[key2] == value2, drows))
table_2d_analize.append(data)
rows.export_to_csv(table_2d_analize, '2d_column_analize.csv')
def test_export_to_csv_accepts_dialect(self):
result_1 = rows.export_to_csv(utils.table, dialect=csv.excel_tab)
result_2 = rows.export_to_csv(utils.table, dialect=csv.excel)
self.assertEqual(result_1.replace(b'\t', b','), result_2)
if row.animaltype == 'Dog':
new_row.update(get_dog_age_columns(row))
# Demora muito
new_row.update(get_dog_breed_columns(row))
new_row.update(get_dog_color_columns(row))
new_row['outcome'] = get_animal_outcome(row)
new_dog_rows.append(new_row)
new_fields = [(key, rows.fields.UnicodeField) for key in new_cat_rows[0].keys()]
table_to = rows.Table(fields=OrderedDict(new_fields))
for row in new_cat_rows:
table_to.append(row)
rows.export_to_csv(table_to, "clean_data3_cat.csv")
new_fields = [(key, rows.fields.UnicodeField) for key in new_dog_rows[0].keys()]
table_to = rows.Table(fields=OrderedDict(new_fields))
for row in new_dog_rows:
table_to.append(row)
rows.export_to_csv(table_to, "clean_data3_dog.csv")
##################
# Limpando dados de teste
##################
table_from = rows.import_from_csv("../test.csv")
# Get data from Portuguese Wikipedia
city_list_url = "https://pt.wikipedia.org/wiki/Lista_de_munic%C3%ADpios_do_Brasil"
response = requests.get(city_list_url)
html = response.content
# Extract desired data using XPath
cities = rows.import_from_xpath(
BytesIO(html),
rows_xpath="//table/tr/td/ul/li",
fields_xpath=OrderedDict([("name", ".//text()"), ("link", ".//a/@href")]),
)
regexp_city_state = re.compile(r"(.*) \(([A-Z]{2})\)")
def transform(row, table):
'Transform row "link" into full URL and add "state" based on "name"'
data = row._asdict()
data["link"] = urljoin("https://pt.wikipedia.org", data["link"])
data["name"], data["state"] = regexp_city_state.findall(data["name"])[0]
return data
new_fields = OrderedDict()
new_fields["name"] = cities.fields["name"]
new_fields["state"] = rows.fields.TextField # new field
new_fields["link"] = cities.fields["link"]
cities = rows.transform(new_fields, transform, cities)
rows.export_to_csv(cities, "brazilian-cities.csv")
def csv_claims_by_state():
counter = count_claims_by_state()
result = import_from_dicts(counter)
result.order_by('label')
return rows.export_to_csv(result)
def test_export_callback(self):
table = rows.import_from_dicts([{"id": number} for number in range(10)])
myfunc = mock.Mock()
rows.export_to_csv(table, callback=myfunc, batch_size=3)
self.assertEqual(myfunc.call_count, 4)
self.assertEqual([x[0][0] for x in myfunc.call_args_list], [3, 6, 9, 10])
if value not in statistics[key]:
statistics[key][value] = 0
statistics[key][value] += 1
string = rows.fields.UnicodeField
table_output = rows.Table(fields=OrderedDict({'column': string, 'value': string, 'amount': string, 'percent': string}))
for key in statistics.keys():
for value in statistics[key].keys():
table_output.append({
'column': key,
'value': value,
'amount': statistics[key][value],
'percent': "{0:.2f}".format(statistics[key][value] / quantidade_de_exemplos * 100)
})
rows.export_to_csv(table_output, '1d_column_analize.csv')
# columns = {}
# columns['value'] = string
# for key in statistics.keys():
# for value in statistics[key].keys():
# columns[key + '_' + value] = string
# table_2d_analize = rows.Table(fields=columns)
# drows = map(lambda r: r.__dict__, table)
# for key in statistics.keys():
# for value in statistics[key].keys():
# data = {}
# data['value'] = key + '_' + value
# for key2 in statistics.keys():
# for value2 in statistics[key2].keys():
import rows
extract_links = rows.plugins.html.extract_links
extract_text = rows.plugins.html.extract_text
# Get the HTML
url = "http://wnpp.debian.net/"
response = requests.get(url)
html = response.content
# Import data, preserving cell's HTML
packages = rows.import_from_html(BytesIO(html), index=10, preserve_html=True)
def transform(row, table):
'Extract links from "project" field and remove HTML from all'
data = row._asdict()
data["links"] = " ".join(extract_links(row.project))
for key, value in data.items():
if isinstance(value, six.text_type):
data[key] = extract_text(value)
return data
new_fields = packages.fields.copy()
new_fields["links"] = rows.fields.TextField
packages = rows.transform(new_fields, transform, packages)
rows.export_to_csv(packages, "debian-wnpp.csv")
import requests
import rows
import six
extract_links = rows.plugins.html.extract_links
extract_text = rows.plugins.html.extract_text
# Get the HTML
url = 'http://wnpp.debian.net/'
response = requests.get(url)
html = response.content
# Import data, preserving cell's HTML
packages = rows.import_from_html(BytesIO(html), index=10, preserve_html=True)
def transform(row, table):
'Extract links from "project" field and remove HTML from all'
data = row._asdict()
data['links'] = ' '.join(extract_links(row.project))
for key, value in data.items():
if isinstance(value, six.text_type):
data[key] = extract_text(value)
return data
new_fields = packages.fields.copy()
new_fields['links'] = rows.fields.TextField
packages = rows.transform(new_fields, transform, packages)
rows.export_to_csv(packages, 'debian-wnpp.csv')
def test_export_to_csv_accepts_dialect(self):
result_1 = rows.export_to_csv(utils.table, dialect=csv.excel_tab)
result_2 = rows.export_to_csv(utils.table, dialect=csv.excel)
self.assertEqual(result_1.replace(b'\t', b','), result_2)
def csv_claims_by_tag():
counter = count_claims_by_tag()
result = import_from_dicts(counter)
result.order_by('count')
return rows.export_to_csv(result)
('animaltype', rows.fields.UnicodeField),
('sex', rows.fields.UnicodeField),
('castration', rows.fields.UnicodeField),
#('ageuponoutcome', rows.fields.UnicodeField),
#('breed', rows.fields.UnicodeField),
#('color', rows.fields.UnicodeField)
])
table_3 = rows.Table(fields=table3_fields)
for row in table_2:
if len(row.name):
has_name = 'Yes'
else:
has_name = 'No'
if (row.datetime == "Sunday" or row.datetime == "Saturday" or row.holiday == 'True'):
free_day = True
else:
free_day = False
table_3.append({
'has_name': has_name,
'free_day': free_day,
'outcometype': row.outcometype,
'animaltype': row.animaltype,
'sex': row.sex,
'castration': row.castration,
})
rows.export_to_csv(table_3, "clean_data2.csv")
