parser_params.py

import argparse
import sys


def usage_msg():
    return "\nIntPy's Python command line arguments help:\n\n\
To run your experiment with IntPy use:\n\
$ python "+str(sys.argv[0])+" program_arguments [-h, --help] [-g, --glossary] [-m memory|help, --memory memory|help] [-0, --no-cache] [-H type|help, --hash type|help] [-M method|help, --marshalling method|help] [-s form|help, --storage form|help]\n\n\
To run in the IntPy DEBUG mode use:\n\
$ DEBUG=True python "+str(sys.argv[0])+" program_arguments [-h, --help] [-g, --glossary] [-m memory|help, --memory memory|help] [-0, --no-cache] [-H type|help, --hash type|help] [-M method|help, --marshalling method|help] [-s form|help, --storage form|help]\n\n\
"
def glossary_msg():
    return hashes_msg() + marshalling_msg() + storage_msg() + memory_msg()

def hashes_msg():
    return "\nHashes: \n\
    =>md5   : is a cryptographic hash fuction with a better collision resistence and lower performance compared to the others.\n\
    =>murmur: is a modern non-cryptographic hash function with a low collision rate and high performance.\n\
    =>xxhash: is a modern non-cryptographic hash function with a lower collision resistence and better performance compered to murmur.\n\
    usage: $ python "+str(sys.argv[0])+" program_arguments -H|--hash options\
    \n"
def marshalling_msg():
    return "Marshalling:\n\
    =>Pickle:\n\
    usage: $ python "+str(sys.argv[0])+" program_arguments -M|--marshalling options\n\
    \n"

def storage_msg():
    return "Storage:\n\
    =>db-file: use database and file to store data.\n\
    =>db     : use database to store data\n\
    =>file   : use file to store data.\n\
    usage: $ python "+str(sys.argv[0])+" program_arguments -s|--storage options\
    \n"

def memory_msg():
    return "Memory forms:\n\
    =>ad      : original version with some bug fixes and instrumentation, all data are stored directly in the database.\n\
    =>1d-ow   : one dicionary (1d), only write (ow), 1st implementation of dictionary: new data is added to the dictionary only when cache miss occur and the function decorated with @deterministic is executed.\n\
    =>1d-ad   : one dicionary (1d), all data loaded at the begining (ad), 2nd implementation of dictionary (uses 1 dictionary): at the begining of the execution all the data cached is loaded to the dictonary before the user script starts to run.\n\
    =>2d-ad   : two dicionaries (2d), all data loaded at the begining (ad), 3rd implementation of dictionary (uses 2 dictionaries): at the begining of the execution all the data cached is loaded to the dictionary DATA_DICTIONARY before the user script starts to run. When cache miss occurs and a function decorated with @deterministic is processed, its result is stored in NEW_DATA_DICTIONARY. This way, only the elements of NEW_DATA_DICTIONARY are added to the database at the end of the execution.\n\
    =>2d-ad-t : two dicionaries (2d), all data loaded at the begining with a thread (ad-t), 4th implementation of dictionary (uses 2 dictionaries): at the begining of the execution a thread is started to load all the data cached in the database to the dictionary DATA_DICTIONARY. When cache miss occurs and a function decorated with @deterministic is processed, its result is stored in NEW_DATA_DICTIONARY. Only the elements of NEW_DATA_DICTIONARY are added to the database at the end of the execution but it is possible that some elements in NEW_DATA_DICTIONARY are already in the database due to the concurrent execution of the experiment and the thread that populates DATA_DICTIONARY.\n\
    =>2d-ad-f : two dicionaries (2d), all data loaded at the begining of a function(ad-f), 5th implementation of dictionary (uses 2 dictionaries): when @deterministic is executed a select query is created to the database to bring all results of the function decorated with @deterministic stored in the cache. A list of functions already inserted to the dictionary is maintained to avoid unecessary querys to the database. The results are then stored in the dictionary DATA_DICTIONARY. When cache miss occurs and a function decorated with @deterministic is processed, its result is stored in NEW_DATA_DICTIONARY. This way, only the elements of NEW_DATA_DICTIONARY are added to the database at the end of the execution.\n\
    =>2d-ad-ft: two dicionaries (2d), all data loaded at the begining of a function with a thread (ad-ft), 6th implementation of dictionary (uses 2 dictionaries): when @deterministic is executed a select query is created to the database to bring all results of the function decorated with @deterministic stored in the cache. A list of functions already inserted to the dictionary is maintained to avoid unecessary querys to the database. The results of the query are stored in the dictionary DATA_DICTIONARY by a thread. When cache miss occurs and a function decorated with @deterministic is processed, its result is stored in NEW_DATA_DICTIONARY. This way, only the elements of NEW_DATA_DICTIONARY are added to the database at the end of the execution.\n\
    =>2d-lz   : two dicionaries (2d), lazy mode (lz), 7th implementation of dictionary (uses 2 dictionaries): new data is added to DATA_DICTIONARY when cache hit occurs (LAZY approach) and new data is added to NEW_DATA_DICTIONARY when cache miss occur and the function decorated with @deterministic is executed.\n\
    usage: $ python "+str(sys.argv[0])+" program_arguments -m|--memory options\n\
    \n"
    

def get_params():
    memories = ['help','ad', '1d-ow', '1d-ad', '2d-ad', '2d-ad-t', '2d-ad-f', '2d-ad-ft', '2d-lz']

    hashes = ['help','md5', 'murmur', 'xxhash']

    marshals = ['help','pickle']

    storageOptions = ['help','db-file','db','file']

    intpy_arg_parser = argparse.ArgumentParser(usage=usage_msg())

    
    intpy_arg_parser.add_argument('args',
                                   metavar='program arguments',
                                   nargs='*',
                                   type=str, 
                                   help='program arguments')
    
    intpy_arg_parser.add_argument('-m',
                                  '--memory',
                                   choices=memories,
                                   metavar='',
                                   nargs=1,
                                   type=str,
                                   default=['2d-ad'],
                                   help='IntPy\'s mechanism of persistence: choose one of the following options: '+', '.join(memories))
    
    intpy_arg_parser.add_argument('-H',
                                  '--hash',
                                   choices=hashes,
                                   metavar='',
                                   nargs=1,
                                   default=['md5'],
                                   help='SpeedUpy\'s mechanism of hashes: choose one of the following options: '+', '.join(hashes))
    
    intpy_arg_parser.add_argument('-g',
                                  '--glossary',
                                  default=False,
                                  action='store_true',
                                  help='show details of SpeedUpy versions')
   
    intpy_arg_parser.add_argument('-0',
                                  '--no-cache',
                                  default=False,
                                  action="store_true",
                                  help='SpeedUpy\'s disable cache')

    intpy_arg_parser.add_argument('-M',
                                  '--marshalling',
                                   choices=marshals,
                                   metavar='',
                                   nargs=1,
                                   default=['pickle'],
                                   help='SpeedUpy\'s mechanism of marshalling: choose one of the following options: '+', '.join(marshals))
    
    intpy_arg_parser.add_argument('-s',
                                  '--storage',
                                   choices=storageOptions,
                                   metavar='',
                                   nargs=1,
                                   default=['db-file'],
                                   help='SpeedUpy\'s mechanism of storage: choose one of the following options: '+', '.join(storageOptions))

    
    args = intpy_arg_parser.parse_args()

    
    if args.glossary:
        print(glossary_msg())
        sys.exit()
    
    argsp_m = args.memory

    argsp_s = args.storage

    argsp_M = args.marshalling

    argsp_no_cache = args.no_cache

    argsp_hash = args.hash


    if str(argsp_m[0]) == 'help' or str(argsp_M[0]) == 'help' or str(argsp_hash[0]) == 'help' or str(argsp_s[0]) == 'help':
        if str(argsp_m[0]) == 'help':
            print(memory_msg())
            
        if str(argsp_M[0]) == 'help':
            print(marshalling_msg())

        if str(argsp_hash[0]) == 'help':
            print(hashes_msg())

        if str(argsp_s[0]) == 'help':
            print(storage_msg())
        sys.exit()

    return argsp_m, argsp_M, argsp_s, argsp_no_cache, argsp_hash


"""
if argsp.version == ['1d-ow'] or argsp.version == ['v021x']:
    v_data_access = ".data_access_v021x_1d-ow"
elif argsp.version == ['1d-ad'] or argsp.version == ['v022x']:
    v_data_access = ".data_access_v022x_1d-ad"
elif argsp.version == ['2d-ad'] or argsp.version == ['v023x']:
    v_data_access = ".data_access_v023x_2d-ad"
elif argsp.version == ['2d-ad-t'] or argsp.version == ['v024x']:
    v_data_access = ".data_access_v024x_2d-ad-t"
elif argsp.version == ['2d-ad-f'] or argsp.version == ['v025x']:
    v_data_access = ".data_access_v025x_2d-ad-f"
elif argsp.version == ['2d-ad-ft'] or argsp.version == ['v026x']:
    v_data_access = ".data_access_v026x_2d-ad-ft"
elif argsp.version == ['2d-lz'] or argsp.version == ['v027x']:
    v_data_access = ".data_access_v027x_2d-lz"
else:
    v_data_access = ".data_access_v021x_1d-ow"
"""