def accept(self,
key: str = None,
timeout: int = 3600,
**opt) -> Union[tpsup.nettools.encryptedsocket, None]:
verbose = opt.get('verbose', 0)
if verbose:
tplog(
f"waiting for new client connection. time out after {timeout} idle seconds"
)
selector = _ServerSelector()
selector.register(self.socket, selectors.EVENT_READ)
poll_interval = 1
waited_so_far = 0
while waited_so_far < timeout:
if verbose > 2:
print("looping")
ready = selector.select(poll_interval)
if ready:
(clientsocket, address) = self.socket.accept()
if verbose:
tplog(
f"accepted client socket {clientsocket}, address={address}"
)
return tpsup.nettools.encryptedsocket(
established_socket=clientsocket, key=key)
else:
waited_so_far += poll_interval
return None
def timeout_child(conn: multiprocessing.connection.Connection,
func: types.FunctionType, *args, **kwargs):
"""
wrapper function for timeout_func
:param conn:
:param func:
:param args:
:param kwargs:
:return:
"""
verbose = kwargs.get('verbose', 0)
if verbose:
tplog(f"func={func}")
if not func:
# https://stackoverflow.com/questions/43369648/cant-get-attribute-function-inner-on-module-mp-main-from-e-python
# when the multiprocessing library copies your main module, it won't run it as the __main__ script and
# therefore anything defined inside the if __name__ == '__main__' is not defined in the child process
# namespace. Hence, the AttributeError
message = f"both func={func} is not initialized. note: func cannot be defined in __main__"
tb = pformat(
traceback.format_stack()) # outside exception use format_stack()
conn.send(RuntimeError(f"{message}\n{tb}"))
else:
result = None
try:
result = func(*args, **kwargs)
conn.send(result)
except Exception as e:
tb = pformat(
traceback.format_exc()) # within exception use format_exc()
conn.send(
RuntimeError(
f"child process exception, pid={os.getpid()}\n{tb}"))
conn.close()
def __init__(self,
key: str,
established_socket: socket.socket = None,
host_port: str = None,
maxtry: int = 5,
try_interval: int = 3,
**opt):
if established_socket:
self.socket = established_socket
elif host_port:
host, port = host_port.split(':')
if not port:
raise RuntimeError(
f"bad format at host_port='{host_port}'; expected host:port"
)
sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
if sock:
for i in range(0, maxtry):
try:
sock.connect((host, int(port)))
# tplog(f"connected to {sock}")
# __init__ connected to <socket.socket fd=3, family=AddressFamily.AF_INET,
# type=SocketKind.SOCK_STREAM, proto=0, laddr=('127.0.0.1', 36690), raddr=('127.0.0.1', 29999)>
laddr = sock.getsockname()
raddr = sock.getpeername()
tplog(
f"connected: local {laddr[0]}:{laddr[1]}, remote {raddr[0]}:{raddr[1]}"
)
tplog(
f"connected: local {laddr[0]}:{laddr[1]}, remote {raddr[0]}:{raddr[1]}"
)
self.socket = sock
break
except Exception as e:
tplog(
f'{i + 1} try out of {maxtry} failed to connect: {e}',
file=sys.stderr)
if i + 1 < maxtry:
tplog(f'will retry after {try_interval}',
file=sys.stderr)
time.sleep(try_interval)
else:
raise RuntimeError(
f"failed to connect to {host}:{port} in {maxtry} tries"
)
else:
raise RuntimeError(socket.error)
else:
raise RuntimeError(
"neither established_socket nor host_port specified")
if key is None or key == '':
key = None
self.key = key
self.in_coder = tpsup.coder.Coder(key)
self.out_coder = tpsup.coder.Coder(key)
def get_attrs(self,
element: WebElement,
method: str = 'bs4',
verbose: int = 0):
"""
get list of attributes from an element.
https://stackoverflow.com/questions/27307131/selenium-webdriver-how-do-i-find-all-of-an-elements-attributes
somehow, webdriver doesn't have an API for this
:param element:
:param method:
:return:
"""
"""
note: for html '<div class="login-greeting">Hi LCA Editor Tester,</div>'
bs4 will give: {'class': ['login-greeting']}
js will give: {'class': 'login-greeting' }
"""
if method == 'bs4':
html: str = element.get_attribute('outerHTML')
if verbose:
tplog(f"outerHTML={html}")
if html:
attrs = {}
soup = BeautifulSoup(html, 'html.parser')
# https://www.crummy.com/software/BeautifulSoup/bs4/doc/#attributes
for element in soup():
# soup() is a generator
# element.attrs is a dict
attrs.update(element.attrs)
return attrs
else:
return {}
elif method == 'js':
# java script. duplicate attributes will be overwritten
js_script = 'var items = {}; ' \
'for (index = 0; index < arguments[0].attributes.length; ++index) { ' \
' items[arguments[0].attributes[index].name] = arguments[0].attributes[index].value' \
'};' \
'return items;'
attrs = self.driver.execute_script(js_script, element)
return attrs
else:
raise RuntimeError(
f"unsupported method={method}. accepted: bs4 or js")
def main():
print(
'------- test timeout_func_unix(). should work on Unix and fail on windows'
)
try:
sleep_5 = timeout_func_on_unix(2, module_sleep_and_tick, 3)
sleep_5(10)
except Exception as e:
# error on Windows
# AttributeError: module 'signal' has no attribute 'SIGALRM'
print_exception(e)
def main_sleep_and_tick(duration: int, *args, **opt) -> str:
"""
this is local function:
https://stackoverflow.com/questions/36994839/i-can-pickle-local-objects-if-i-use-a-derived-class
"""
print(f"args = {pformat(args)}")
print(f"opt = {pformat(opt)}")
for i in range(duration):
time.sleep(1)
print('tick')
message = f"pid={os.getpid()}"
print(message)
return message
b = main_sleep_and_tick
print(f"type={type(module_sleep_and_tick)}")
print(
'------- test timeout_func(). with function defined in module, should work but will time out'
)
try:
result = timeout_func(2,
module_sleep_and_tick,
10,
message="should timeout")
tplog(f"result={pformat(result)}")
except TimeoutException as e:
tplog_exception(e)
tplog("got expected exception\n\n")
print(
'------- test timeout_func(). with function defined in main, may work on unix but should fail on windows '
'with pickle error')
try:
result = timeout_func(2,
main_sleep_and_tick,
10,
message="should timeout")
tplog(f"result={pformat(result)}")
except AttributeError as e:
# AttributeError: Can't pickle local object 'main.<locals>.main_sleep_and_tick'
tplog_exception(e)
tplog("got expected exception\n\n")
print('\n------- test timeout_func(). child will raise exception')
try:
result = timeout_func(3, test_child_exception)
tplog(f"result={pformat(result)}")
except TimeoutException as e:
tplog_exception(e)
tplog("got expected exception\n\n")
def timeout_func(timeout: int, func: Callable, *args, **kwargs):
"""
time out a func. signal.ALRM not working on Windows. therefore this
https://stackoverflow.com/questions/492519/timeout-on-a-function-call
so we use multiprocessing.Pipe() to do this
https://docs.python.org/3.8/library/multiprocessing.html
use this method because we can
1. multiprocessing is process-based. process is easier to check and kill after timeout
2. p.join(timeout) is easier to set timeout
3. Pipe() is easier to collect return value
there is decorator using a same method. It has covered a lot of edge conidtions. good place to learn
https://github.com/bitranox/wrapt_timeout_decorator/blob/master/wrapt_timeout_decorator/wrap_function_multiprocess.py
"""
# multiprocessing cannot run child with function defined in __main__
# https://stackoverflow.com/questions/43369648/cant-get-attribute-function-inner-on-module-mp-main-from-e-python
# when the multiprocessing library copies your main module, it won't run it as the __main__ script
# and therefore anything defined inside the if __name__ == '__main__' is not defined in the child
# process namespace. Hence, the AttributeError
# I tried to reload module or delay import module to work around this. neither worked.
# re-import failed to work around
# importlib.reload(multiprocessing)
# importlib.reload(multiprocessing.connection)
# delayed import failed to work around
# import multiprocessing
# import multiprocessing.connection
# the parent process can always see the process defined in __main__
# tplog(getattr(sys.modules['__mp_main__'], 'local_sleep_and_tick'))
parent_conn, child_conn = multiprocessing.Pipe(duplex=False)
# how to pass args and kwargs
# https://stackoverflow.com/questions/38908663/python-multiprocessing-how-to-pass-kwargs-to-function
# p = multiprocessing.Process(target=timeout_child, args=(child_conn, func, *args), kwargs=kwargs)
p = multiprocessing.Process(target=timeout_child,
args=(child_conn, func, *args),
kwargs=kwargs)
# kill child after parent exits
# https://stackoverflow.com/questions/25542110/kill-child-process-if-parent-is-killed-in-python
p.daemon = True
p.start()
# Wait for timeout seconds or until process finishes
p.join(timeout)
# If thread is still active
if p.is_alive():
message = f"timed out after {timeout} second(s), terminating pid={p.pid}"
tplog(message)
traceback.print_stack(
) # default to stderr, set file=sys.stdout for stdout
# Terminate
p.terminate()
p.join(0.5)
parent_conn.close()
raise TimeoutException(timeout=timeout, child_pid=p.pid)
else:
result = None
if parent_conn.poll(0.1):
# recv() is a blocked call, therefore, poll() first
result = parent_conn.recv()
parent_conn.close()
if isinstance(result, Exception):
raise result
else:
# child finished and returned anything
return result
else:
# child finished but didn't return anything
parent_conn.close()
def send_and_encode(self,
data: Union[bytes, str],
data_is_file: bool = False,
timeout: int = 6,
**opt) -> int:
""" this is actually use unblocked send() to re-implement a blocked sendall().
The possible benefits:
- add encryption
"""
file = None
fd = None
if data_is_file:
file = data
size = os.path.getsize(file)
fd = os.open(file, os.O_RDWR) # fd is int
data = mmap.mmap(fd, size, access=mmap.ACCESS_READ)
# tplog(f"data type = {type(data)}") #
# https://docs.python.org/3/library/socket.html
# socket.send() vs socket.sendall()
#
# socket.send(bytes[, flags])
# Send data to the socket. Returns the number of bytes sent. Applications are responsible
# for checking that all data has been sent; if only some of the data was transmitted, the application
# needs to attempt delivery of the remaining data.
#
# socket.sendall(bytes[, flags])
# Send data to the socket. Unlike send(), this method continues to
# send data from bytes until either all data has been sent or an error occurs. None is returned on success. On
# error, an exception is raised, and there is no way to determine how much data, if any, was successfully sent.
saved_timeout = self.socket.gettimeout()
# saved_blocking = self.socket.getblocking() this only available when version >= 3.7
# unblock
self.socket.setblocking(
0
) # None: blocking mode; 0: non-blocking mode; positive floating: timeout mode.
# this is the same as self.socket.settimeout(0)
polling_interval = 1
wait_so_far = 0
total_sent = 0
data_length = len(data)
while total_sent < data_length:
buffer_size = 4096
new_end = total_sent + buffer_size
if new_end > data_length:
new_end = data_length
buffer_size = new_end - total_sent
buffer_bytes = self.out_coder.xor(data[total_sent:new_end])
buffer_sent = 0
while buffer_sent < buffer_size:
ready = select.select([], [self.socket], [], polling_interval)
if ready[1]:
sent = self.socket.send(buffer_bytes[buffer_sent:])
# there may be exceptions here
if sent == 0:
tplog(f"Remote connection is closed during our send()")
break
buffer_sent += sent
total_sent += sent
else:
wait_so_far += polling_interval
if wait_so_far >= timeout:
tplog(f"send() timed out after {wait_so_far} seconds")
break
else:
# the above loop did NOT break = the above loop condition was False = the loop ended naturally
# python's way to break nested loop (double loop)
# https://stackoverflow.com/questions/653509/breaking-out-of-nested-loops
continue
break # the above loop DID break = the above loop condition was still True = the loop didn't end naturally
if fd:
os.close(fd)
missing = data_length - total_sent
if missing > 0:
tplog(
f"Sent total {total_sent} bytes. failed to send {missing} bytes"
)
else:
tplog(f"Sent all {total_sent} bytes")
# getblocking()/setblocking() is implemented by gettimeout()/settimeout. no getblocking() before 3.7
self.socket.settimeout(saved_timeout)
# self.socket.setblocking(saved_blocking) # restoring blocking setting
return total_sent
def recv_and_decode(self,
timeout: int = 6,
maxsize=1024 * 1024 * 1024,
file: str = None,
**opt) -> Union[bytes, int]:
""" this is actually use unblocked recv() to re-implement a blocked recv().
The possible benefits:
- sock.recv(buffer_size)'s buffer_size is limited, we can use a loop to recv() bigger file. In this
loop pattern, socket.settimeout() looks awkward. socket.settimeout() seems better for
small data
- decrypt the incoming data with smaller chunks, ie, in buffer_size
"""
if file:
# receive data will write to this file
fh = open(file, 'wb')
else:
fh = None
# https://docs.python.org/3/library/socket.html
# getblocking()/setblocking() is implemented by gettimeout()/settimeout. no getblocking() before 3.7
saved_timeout = self.socket.gettimeout()
# saved_blocking = self.socket.getblocking() this only available when version >= 3.7
# unblock
self.socket.setblocking(
0
) # None: blocking mode; 0: non-blocking mode; positive floating: timeout mode.
polling_interval = 1
wait_so_far = 0
total_size = 0
received_bytearray = bytearray()
while wait_so_far < timeout:
ready = select.select(
[self.socket], [], [],
polling_interval) # last arg is timeout in seconds
if ready[0]:
data = self.socket.recv(4096)
# there may be exceptions here
if data == b'':
tplog(f"Remote connection is closed during our recv()")
break
if fh:
fh.write(self.in_coder.xor(data))
else:
received_bytearray.extend(self.in_coder.xor(data))
total_size += len(data)
if total_size > maxsize:
tplog(
f"Received {total_size} bytes already exceeded maxsize {maxsize}. Stopped receiving."
)
break
else:
# time.sleep(polling_interval) # no need sleep here if we already blocked at select()
wait_so_far += polling_interval
if fh:
fh.close()
if wait_so_far >= timeout:
tplog(f"recv() timed out after {wait_so_far} seconds")
tplog(f"Received total {total_size} bytes")
# getblocking()/setblocking() is implemented by gettimeout()/settimeout. no getblocking() before 3.7
self.socket.settimeout(saved_timeout)
# self.socket.setblocking(saved_blocking) # restoring blocking setting
if file:
return total_size
else:
return bytes(received_bytearray)
def run(seleniumEnv: tpsup.seleniumtools.SeleniumEnv, **opt):
verbose = opt.get('verbose', 0)
mod_file = opt.get('mod_file', 'mod_file')
argList = opt.get('argList', [])
if verbose:
sys.stderr.write(f"{mod_file} argList=\n")
sys.stderr.write(pformat(argList) + "\n")
parser = argparse.ArgumentParser(prog=mod_file, )
parser.add_argument('-js',
dest='use_javascript',
default=False,
action='store_true',
help='use javascript')
parser.add_argument('-wait',
dest='wait',
default=5,
action='store',
type=int,
help='rename the file. default not to rename')
parser.add_argument('-rename',
dest='renamed',
default=None,
action='store',
help='rename the file. default not to rename')
args = vars(parser.parse_args(argList))
if not verbose:
verbose = args.get('verbose', 0)
if verbose:
tplog(f"args = {pformat(args)}")
driver = seleniumEnv.get_driver()
url = f"http://livingstonchinese.org/LCA2/index.php/join-us"
driver.get(url)
# https://stackoverflow.com/questions/46937319/how-to-use-chrome-webdriver-in-selenium-to-download-files-in-python
driver.command_executor._commands["send_command"] = (
"POST", '/session/$sessionId/chromium/send_command')
params = {
'cmd': 'Page.setDownloadBehavior',
'params': {
'behavior': 'allow',
'downloadPath': seleniumEnv.download_dir
}
}
command_result = driver.execute("send_command", params)
# https://dev.to/endtest/a-practical-guide-for-finding-elements-with-selenium-4djf
# <a style="color: #1b57b1; font-weight: normal; text-decoration: none;"
# href="/LCA2/images/docs/public/lca_bylaw_2019_11.pdf">lick to view LCA By Law</a>
#elem = driver.find_element_by_css_selector("#content > div.item-page > div:nth-child(4) > pre:nth-child(15) > span > a")
elem = driver.find_element_by_partial_link_text("view LCA By Law")
src = elem.get_attribute("href")
env = seleniumEnv.env
download_dir = seleniumEnv.download_dir
shortname = "lca.pdf"
urllib.request.urlretrieve(src, f"{download_dir}/{shortname}")
return download_dir
sys.exit(1)
tmpdir = tpsup.tptmp.tptmp().get_nowdir()
# this is client mode if serverHostPort is defined
ensock = tpsup.nettools.encryptedsocket(key, host_port=serverHostPort)
request = {
'mod_file': args['mod_file'],
'args': args['remainingArgs'],
'accept': args['accept'],
}
request_str = json.dumps(request)
request_bytes = bytes(request_str, "utf-8")
tplog(f"Sending {len(request_bytes)} bytes", file=sys.stderr)
ensock.send_and_encode(request_bytes)
# shut down the send channel so that the other side recv() won't wait forever
ensock.send_shutdown()
tplog("Sent. waiting response")
if request['accept'] == 'json':
received_bytes = ensock.recv_and_decode(
timeout=60) # this needs a long wait
tplog(f"received {received_bytes} bytes")
received_str = str(received_bytes, 'utf-8')
received_structure = json.loads(received_str)
tplog(f"Received data structure: {pformat(received_structure)}")
elif request['accept'] == 'tar':
tar_name = os.path.join(tmpdir, "reply.tar")
def run(seleniumEnv: tpsup.seleniumtools.SeleniumEnv, **opt):
username = "******" # change this to the username associated with your account
verbose = opt.get('verbose', 0)
mod_file = opt.get('mod_file', 'mod_file')
argList = opt.get('argList', [])
if verbose:
sys.stderr.write(f"{mod_file}argList=\n")
sys.stderr.write(pformat(argList) + "\n")
parser = argparse.ArgumentParser(prog=mod_file, )
parser.add_argument('-u',
dest='username',
default=username,
action='store',
help='login user')
args = vars(parser.parse_args(argList))
if not verbose:
verbose = args.get('verbose', 0)
if verbose:
tplog(f"args = {pformat(args)}")
username = args['username']
entryBook = EntryBook()
password = entryBook.get_entry_by_key(username).get('decoded')
driver = seleniumEnv.get_driver()
# print(f'driver.title={driver.title}')
url = 'https://livingstonchinese.org/'
driver.get(url)
expected_url = "https://livingstonchinese.org/LCA2/"
actual_url = driver.current_url
assert expected_url == actual_url
seleniumEnv.delay_for_viewer() # give 1 sec to let the tail set up
# https://dev.to/endtest/a-practical-guide-for-finding-elements-with-selenium-4djf
# in chrome browser, find the interested spot, right click -> inspect, this will bring up source code,
# in the source code window, right click -> copy -> ...
# from Edge/Chrome, right click the item -> inspect
login_elem = driver.find_element_by_id('modlgn-username')
login_elem.send_keys(username)
seleniumEnv.delay_for_viewer(1) # delay to mimic humane slowness
password_elem = driver.find_element_by_id('modlgn-passwd')
password_elem.send_keys(password)
# frameinfo = getframeinfo(currentframe())
# print(frameinfo.filename, frameinfo.lineno, file=sys.stderr) # print line number
# from Edge/Chrome, right click the item -> inspect
# because login button has no "id", so I used xpath. xpath is very sensitive to changes in the page
driver.find_element_by_xpath(
'/html/body/div[1]/div/div/div/div[1]/form/div/div[4]/div/button'
).click()
seleniumEnv.delay_for_viewer(1) # delay to mimic humane slowness
# this doesn't work as 'button' is a grand-child of form-login-sutmit
# driver.find_element_by_id('form-login-submit').click()
# frameinfo = getframeinfo(currentframe())
# print(frameinfo.filename, frameinfo.lineno, file=sys.stderr)
xpath = '//*[@id=\"login-form\"]'
elem = driver.find_element_by_xpath('//*[@id=\"login-form\"]')
welcomeText = elem.text
# frameinfo = getframeinfo(currentframe())
# print(frameinfo.filename, frameinfo.lineno, file=sys.stderr)
print(f"We see: {welcomeText}")
# assert re.search("^Hi ", welcomeText)
error = None
pattern = "^Hi "
if not re.search(pattern, welcomeText):
error = f"xpath='{xpath}' failed to match pattern '{pattern}'"
result = {'error': error, 'data': welcomeText}
return result
def run(seleniumEnv: tpsup.seleniumtools.SeleniumEnv, **opt):
verbose = opt.get('verbose', 0)
mod_file = opt.get('mod_file', 'mod_file')
argList = opt.get('argList', [])
if verbose:
sys.stderr.write(f"{mod_file} argList=\n")
sys.stderr.write(pformat(argList) + "\n")
parser = argparse.ArgumentParser(
prog=mod_file,
)
parser.add_argument(
'-js', dest='use_javascript', default=False, action='store_true',
help='use javascript')
parser.add_argument(
'-wait', dest='wait', default=5, action='store', type=int,
help='rename the file. default not to rename')
parser.add_argument(
'-rename', dest='renamed', default=None, action='store',
help='rename the file. default not to rename')
args = vars(parser.parse_args(argList))
if not verbose:
verbose = args.get('verbose',0)
if verbose:
tplog(f"args = {pformat(args)}")
driver = seleniumEnv.get_driver()
url = f"https://metacpan.org/pod/DBI::Log"
tplog(f"going to url={url}")
driver.get(url)
# https://stackoverflow.com/questions/46937319/how-to-use-chrome-webdriver-in-selenium-to-download-files-in-python
driver.command_executor._commands["send_command"] = ("POST", '/session/$sessionId/chromium/send_command')
params = {'cmd': 'Page.setDownloadBehavior',
'params': {'behavior': 'allow', 'downloadPath': seleniumEnv.download_dir}}
command_result = driver.execute("send_command", params)
time.sleep(1)
# https://dev.to/endtest/a-practical-guide-for-finding-elements-with-selenium-4djf
download_text = 'Download ('
elem = driver.find_element_by_partial_link_text(download_text)
elem.click()
tplog(f"clicked '{download_text}'")
wait_time_for_download = float(args['wait'])
time.sleep(wait_time_for_download)
# control file name
# https://stackoverflow.com/questions/34548041/selenium-give-file-name-when-downloading
files: List[str] = [os.path.join(seleniumEnv.download_dir, f) for f in os.listdir(seleniumEnv.download_dir)]
if files:
filename = max(files, key=os.path.getctime)
renamed = args['renamed']
if renamed:
shutil.move(filename, os.path.join(seleniumEnv.download_dir, renamed))
return renamed
else:
return filename
else:
tplog(f"no file downloaded to {seleniumEnv.download_dir}")
raise RuntimeError(f"no file downloaded in {wait_time_for_download} seconds")
def run(seleniumEnv: tpsup.seleniumtools.SeleniumEnv, **opt):
username = "******" # change this to the username associated with your account
verbose = opt.get('verbose', 0)
mod_file = opt.get('mod_file', 'mod_file')
argList = opt.get('argList', [])
if verbose:
sys.stderr.write(f"{mod_file}argList=\n")
sys.stderr.write(pformat(argList) + "\n")
parser = argparse.ArgumentParser(prog=mod_file, )
parser.add_argument('-u',
dest='username',
default=username,
action='store',
help='login user')
args = vars(parser.parse_args(argList))
if not verbose:
verbose = args.get('verbose', 0)
if verbose:
tplog(f"args = {pformat(args)}")
username = args['username']
entryBook = EntryBook()
password = entryBook.get_entry_by_key(username).get('decoded')
driver = seleniumEnv.get_driver()
# print(f'driver.title={driver.title}')
url = 'https://livingstonchinese.org/'
driver.get(url)
expected_url = "https://livingstonchinese.org/LCA2/"
actual_url = driver.current_url
assert expected_url == actual_url
seleniumEnv.delay_for_viewer() # give 1 sec to let the tail set up
# https://dev.to/endtest/a-practical-guide-for-finding-elements-with-selenium-4djf
# in chrome browser, find the interested spot, right click -> inspect, this will bring up source code,
# in the source code window, right click -> copy -> ...
found_logout = True
need_to_login = False
logout_elem = None
logout_css_selector = '#login-form > div.logout-button > input.btn.btn-primary'
try:
logout_elem = driver.find_element_by_css_selector(logout_css_selector)
except NoSuchElementException:
found_logout = False
if found_logout:
tplog(
f"found logout button, meaning someone already logged in. css_selector='{logout_css_selector}'"
)
need_to_logout = True
greeting_text = None
greeting_css = '#login-form > div.login-greeting'
try:
greeting_elem = driver.find_element_by_css_selector(greeting_css)
if verbose:
attrs = seleniumEnv.get_attrs(greeting_elem, method='bs4')
tplog(f"attrs by bs4 = {pformat(attrs)}")
attrs = seleniumEnv.get_attrs(greeting_elem, method='js')
tplog(f"attrs by js = {pformat(attrs)}")
greeting_text = greeting_elem.text
tplog(f"greeting_text='{greeting_text}' at css='{greeting_css}'")
except Exception as e:
print_exception(e)
tplog(
f"cannot find greeting_text at css='{greeting_css}'. need to log out"
)
if greeting_text:
greeting_pattern = "^Hi (.+),"
m = re.search(greeting_pattern, greeting_text)
if m:
username = m.group(1)
expected_username = "******"
if not username == expected_username:
tplog(
f"username='******' did not match expected username='******'. need to log out"
)
else:
need_to_logout = False
tplog(
f"username='******' matched expected username='******'. no need to log in again"
)
else:
tplog(
f"greeting_text='{greeting_text}', not matching expected greeting_pattern='{greeting_pattern}'. "
f"need to log out")
else:
tplog(
f"cannot find greeting_text at css_selector={greeting_css}. need to log out"
)
if need_to_logout:
need_to_login = True
logout_elem.click()
seleniumEnv.delay_for_viewer(1)
else:
tplog(f"seems nobody logged in. we will login.")
need_to_login = True
if need_to_login:
# from Edge/Chrome, right click the item -> inspect
login_elem = driver.find_element_by_id('modlgn-username')
login_elem.send_keys(username)
seleniumEnv.delay_for_viewer(1) # delay to mimic humane slowness
password_elem = driver.find_element_by_id('modlgn-passwd')
password_elem.send_keys(password)
# frameinfo = getframeinfo(currentframe())
# print(frameinfo.filename, frameinfo.lineno, file=sys.stderr) # print line number
# from Edge/Chrome, right click the item -> inspect
# because login button has no "id", so I used xpath. xpath is very sensitive to changes in the page
driver.find_element_by_xpath(
'/html/body/div[1]/div/div/div/div[1]/form/div/div[4]/div/button'
).click()
seleniumEnv.delay_for_viewer(1) # delay to mimic humane slowness
# this doesn't work as 'button' is a grand-child of form-login-sutmit
# driver.find_element_by_id('form-login-submit').click()
# frameinfo = getframeinfo(currentframe())
# print(frameinfo.filename, frameinfo.lineno, file=sys.stderr)
xpath = '//*[@id=\"login-form\"]'
elem = driver.find_element_by_xpath('//*[@id=\"login-form\"]')
welcomeText = elem.text
# frameinfo = getframeinfo(currentframe())
# print(frameinfo.filename, frameinfo.lineno, file=sys.stderr)
tplog(f"We see: {welcomeText}")
# assert re.search("^Hi ", welcomeText)
error = None
pattern = "^Hi "
if not re.search(pattern, welcomeText):
error = f"xpath='{xpath}' failed to match pattern '{pattern}'"
result = {'error': error, 'data': welcomeText}
return result