Looks for a sequence of text (a prompt) inside a set of lines arbitrary cut, with or without newlines

prompt_finder.py

import itertools

tests_passed = {
    'simple end of line': 
        [ b'One\nSSH>', 
          b'no!',
        ],
    'simple begining of line': 
        [ b'SSH>', 
        ],
    'simple empty lines': 
        [ b'\n', 
          b'\n',
          b'SSH>', 
          b'no!',
        ],
    'last character cut': 
        [ b'One\nSSH', 
          b'>',
        ],
    'last two characters cut': 
        [ b'One\nSS', 
          b'H>',
        ],
    'cut after first': 
        [ b'One\nS', 
          b'SH>',
        ],
}
tests_to_do = {
    'cut by a new line': 
        [ b'One\nS\n', 
          b'SH>',
          b'this should show\n',
          b'SSH>',
        ],
    'all cut': 
        [ b'One\nS', 
          b'S',
          b'H',
          b'>',
          b'no!',
        ],
    'false alsert': 
        [ b'One\nSS H>', 
          b'\nSSH>',
        ],
    'false alsert on two lines': 
        [ b'One\nSS', 
          b' H> '
          b'Yes\nSSH>',
        ],
    'false alert all cut': 
        [ b'One\nS', 
          b'S',
          b'H',
          b' >',
          b'SSH>',
        ],
}

tests = {}
tests.update(tests_passed)
tests.update(tests_to_do)

prompts_that_work = [
]
prompts = [
    'SSH>',
    'SS*>',
    'S*H>',
    'S*>',
    '*>',
    '*SH>',
    '*S*>',
    'SSH*',
    'S.H*',
    '....',
    #'...*'  This one does not work...
]

def fake_read(name):
    #print(f'Testing "{name}"')
    for line in tests[name]:
        yield line
    #print(f'Test "{name}" complete!\n')


def position_of_next_character_to_match(prompt, marker):
    for pos in range(marker+1, len(prompt)):
        if prompt[pos] != '*':
            marker = pos
            break
    return min(len(prompt), marker)


def forward_to_character_following(iterator, character):
    for c in iterator:
        if c == ord(character):
            break
    return iterator


def forward_to_begining_of_new_line(iterator):
    return forward_to_character_following(iterator, '\n')


def find_wildcard_stop_character(marker):
    try:
        after_wildcard = next(marker)
        #Collapse a series of ***** into a single star, making SS******> the same as SS*> (both match SSH>)
        while after_wildcard == '*':
            after_wildcard = next(marker)

        return after_wildcard
    except StopIteration:
        #The prompt ends with a wildcard...
        pass


def eat_next_character(iterator):
    loop = True
    byte = None
    try:
        byte = next(iterator)
    except StopIteration:
        loop = False

    return byte, loop
 
def new_test(name, prompt):

    received_chunks = []

    reader = fake_read(name)

    data = next(reader)
    marker = iter(prompt)

    looking_for = next(marker)
    match_anything = looking_for == '.'
    if looking_for == '*':
        #Make a copy of the iterator. Will also server as a flag 
        #that we are in wildcard mode
        marker, saved_marker = itertools.tee(marker)
        #and start looking for the remainder of the prompt
        #as if it was what we were looking for all along
        looking_for = find_wildcard_stop_character(marker)
    else:
        saved_marker = None

    partial_match = False

    while data:
        received_chunks.append(data)

        #For every character in the string
        iterator = iter(data)

        #Possible states:
        #0: Looking for next character after new line
        #1: Looking for the next character in the prompt string
        #2: Looging for the next charcter that would stop the non-greedy wildcard search
        loop = True

        byte = next(iterator)
        #We expect the prompt to start a new line, so go looging for it
        #find_begining_of_new_line(iterator):
        while loop:  #We cannot use a for loop because we might have to test a character twice
            # If tne character at marker the next one we are looking for? It can be
            #   - The exact char
            #   - Any character because the prompt specification had a . in it
            if chr(byte) == looking_for or match_anything:
                partial_match = True
                
                try:
            #       Increment the counter optimistically 
                    looking_for = next(marker)
                    #Will we match the next char as is?
                    match_anything = looking_for == '.'
                    #If we were in wildcard mode, it's over because we found the character
                    #that made us break out of it
                    #but maybe there is another wild card right after it
                    if looking_for == '*':
                        #We start a new search for the remainder of the prompt
                        partial_match = False
                        marker, saved_marker = itertools.tee(marker)
                        looking_for = find_wildcard_stop_character(marker)

        #       If there are no more character to find
                except StopIteration:
        #           found it!
                    data = None   #Signal to get out of the while loop
                    break   #Get out of the for loop

                #If we reach here, we haven't found it all yet, continue
                byte, loop = eat_next_character(iterator)

            elif saved_marker:
                #We've set looking_for to the character that will get us out the wildcard loop
                #so if we get here this is just any other charcter, there is nothing to do but
                #iterate and see. 
                marker, saved_marker = itertools.tee(saved_marker)
                #We already visited the position in saved_marker. We know it is not a special
                #character and that it will not raise. But we could be at end of line of the marker
                try:
                    looking_for = next(marker)
                except StopIteration:
        #           found it!
                    data = None   #Signal to get out of the while loop
                    break   #Get out of the for loop

                #We failed to match this character, but maybe it is the start of
                #a new prompt. Let's *not* advance the iterator and try again, now
                #that we reset the prompt iterator
                if not partial_match:
                    byte, loop = eat_next_character(iterator)
                else:
                    #But that's a one time thing
                    partial_match = False
            else:
                #Start to search from begining of prompt again
                marker = iter(prompt)
                looking_for = next(marker)
                match_anything = looking_for == '.'
                partial_match = False
                #If the character that did not match happens to be a new line
                #then we must let the loop iterate, or else it will skip valid
                #characters after the \n looking for another one. For example,
                #if we parsing through "\nSSH>\n". Without that if, the call to
                #forward_to_begining_of_new_line will skip over the SSH we are 
                #looking for!
                if chr(byte) != '\n':
                    byte,loop = eat_next_character(forward_to_begining_of_new_line(iterator))
                else:
                    byte, loop = eat_next_character(iterator)
        else:
            #We read all the data without finding anything
            data = next(reader)


    lines = b''.join(received_chunks).decode('utf-8').split('\n')

    return lines



for name in tests.keys():
    for prompt in prompts:    
        print(f"Result for [{name},{prompt}]  ", end='')
        print(f"{new_test(name, prompt)}")
    print('---------------------------------')
    print()