def fix_sort_slash(med, words, medDict):
if "/" not in med:
return _format_return(med, words, medDict)
slashWords = med.split('/')
slashWords = clean_words(slashWords)
slashWords = sorted(slashWords)
med1 = '/'.join(slashWords)
return _format_return(med1, slashWords, medDict)
def try_with(med, words, medDict):
if "/" in med:
tmpMed = med.replace("/", " with ")
tmpWords = clean_words(tmpMed.split(' '))
tmpMed = ' '.join(tmpWords)
if tmpMed in medDict:
return _format_return(tmpMed, tmpWords, medDict)
return _format_return(med, words, medDict)
def fix_dash(med, words, medDict):
if "-" not in med:
return _format_return(med, words, medDict)
noDash = med.replace('-', ' ')
words = clean_words(noDash.split(' '))
tmpWords = sorted(words)
tmpMed = '/'.join(tmpWords)
if tmpMed in medDict:
return _format_return(tmpMed, tmpWords, medDict)
return _format_return(' '.join(words), words, medDict)
def slash_subclass(med, words, medDict):
if "/" in med:
slashWords = med.split('/')
slashWords = clean_words(slashWords)
slashCat = map(lambda x: x in medDict, slashWords)
if all(slashCat):
slashCat = map(lambda x: medDict[x], slashWords)
if all(x == slashCat[0] for x in slashCat):
return _format_return(slashWords[0], words, medDict)
return _format_return(med, words, medDict)
def check_eyes(med, words, medDict):
if 'ophthalmic' in med:
return _remove_word(words, 'ophthalmic', medDict)
if 'eye drop' in med:
cleanWords = set(words).difference(set(['eye', 'drops', 'drop']))
cleanWords = sorted(cleanWords, key=words.index)
medWords = clean_words(cleanWords)
medWords = filter(not_decimal, cleanWords)
return _format_return(' '.join(medWords), medWords, medDict)
return _format_return(med, words, medDict)
def fix_words(med, medWords, medDict):
medWords = map(lambda x: x.replace(';', ''), medWords)
medWords = map(lambda x: x.replace('\'', ''), medWords)
badWords = set([
'inj.', 'inj', 'injection', 'cream', 'hcl', 'oral', 'concentrate',
'ointment', 'liquid', 'suspension', "patch", 'chewable', '.', 'neb',
'flush', 'protocol', 'initiated', 'patient', 'uses', 'intra-muscular',
'intramuscular', 'transdermal', 'valerate', 'sublingual', 'syrup',
'extract', 'capsule', 'cap', 'dissolve', 'tablet', 'tab', 'mdi',
'soln', 'solution', 'infusion', '%', 'sustained', 'elixir', 'powder',
'inhaler', 'hfa', 'suppository', 'topical', 'disintegrating',
'delayed', 'jelly', 'gel', 'monohydrate', 'macrocrystals',
'macrocrystal', 'otic', 'immediate', 'releas', 'release', 'pca', 'h2o',
'human', 'sliding', 'scale', 'scal', 'bolus', 'take', 'home', 'pack',
'coated', 'parenteral', 'vaginal', "#", 'supp', 'supplements', 'er'
])
words = sorted(set(medWords).difference(badWords), key=medWords.index)
words = map(lambda x: x.replace('%', ''), words)
words = map(_check_num_slash, words)
words = clean_words(words)
return _format_return(' '.join(words), words, medDict)
def get_med_approx(med, medDict):
med = med.replace(' - ', '-')
# search by splitting either on : or ()
medWords = try_splits(med)
medSuccess = map(lambda x: x in medDict, medWords)
if any(medSuccess):
idx = medSuccess.index(True)
return medWords[idx], medDict[medWords[idx]]
tmpMed = med.replace("w/", "/ ")
words = tmpMed.split(' ')
cleanFuncs = [
remove_parens, quick_match, fix_synonyms, fix_words, fix_vitamin,
quick_match, try_with, truncate_for, check_strength, fix_sort_slash,
try_no_slash, slash_subclass, fix_dash, check_eyes, fuzzy_match,
quick_lookup, remove_extra, quick_match
]
for func in cleanFuncs:
# print str(func), tmpMed, words
tmpMed, words, medType = func(tmpMed, words, medDict)
if medType is not None:
print "Med:" + str(med) + " = " + str(tmpMed)
return tmpMed, medType
if not len(words):
return med, None
if "/" in med:
slashWords = med.split('/')
slashWords = clean_words(slashWords)
slashCat = map(lambda x: x in medDict, slashWords)
if all(slashCat):
slashCat = map(lambda x: medDict[x], slashWords)
grpCat = set(list(itertools.chain(*slashCat)))
print "Concatenated together:" + med
return ' '.join(slashWords), list(grpCat)
medType = check_rxnorm(med, medDict)
if medType is not None:
return med, medType
medType = check_rxnorm(tmpMed, medDict)
return tmpMed, medType
def _remove_word(words, txt, medDict):
medWords = filter(lambda x: x not in set(['hr', 'ns', txt, '/']), words)
medWords = map(lambda x: decimal_and_pattern(x, txt), medWords)
medWords = clean_words(medWords)
return _format_return(' '.join(medWords), medWords, medDict)
