Expand Connaught lexicon

dictionary = open("/home/jim/Playing/mfa_alignments/snc.dict")
dictionary2 = open("/home/jim/Playing/mfa_alignments/snc.dict.exp", "w")

alts = {
    "acu": "acub",
    "againn": "ainn",
    "agam": "am",
    "agamsa": "amsa",
    "agat": "ad",
    "agatsa": "adsa",
    "arís": "aríst",
    "bóthar": "bór",
    "ceistigh": "ceisnigh",
    "claí": "cladh",
    "cluiche": "cluife",
    "contae": "condae",
    "croitheadh": "crathadh",
    "dada": "tada",
    "daoibh": "dhaoib",
    "de": "dhe",
    "de": "ge",
    "deartháir": "driotháir",
    "dheirfiúr": "dhrifiúr",
    "dheirfiúracha": "dhrifiúracha",
    "dom": "dhom",
    "domsa": "dhomsa",
    "droichead": ["draed", "draighead"],
    "duit": "dhuit",
    "duitse": "dhuitse",
    "díbh": "díofa",
    "díobh": "díob",
    "dócha": "dóiche",
    "dóibh": "dóib",
    "dóigh": "dóiche",
    "dúinn": "dhúinn",
    "féin": "fhéin",
    "foighne": "foighid",
    "folach": "falach",
    "foscadh": "fascadh",
    "gnaithe": "gnaíthe",
    "iúdás": "iúdas",
    "leo": ["leob", "leofa"],
    "léi": ["léithe", "léí"],
    "litir": "leitir",
    "luigh": "loigh",
    "luí": "loighe",
    "mé": "me",
    "naimhdeach": "náimhdeach",
    "namhaid": "náimhid",
    "nuacht": "nuaíocht",
    "nuachta": "nuaíochta",
    "nóiméad": ["móiméad", "mhóiméad"],
    "nóiméid": ["móiméid", "mhóiméid"],
    "orthu": "orthub",
    "scafánta": "scufánta",
    "scornach": "scórnach",
    "sé": "se",
    "sibh": "sib",
    "taispeáin": "taspáin",
    "taispeánfaidh": "taspánfaidh",
    "teacht": "tíocht",
    "theacht": "thíocht",    
}
nonwords = {
    "bór": "b oo r",
    "draed": "d r ee d",
    "draighead": "d r ai d",
    "ge": "g @",
    "léí": "lj ee ii",
}
maybe_missing = {
    "duit": "d i tj",
    "nuaíocht": "n uu i@ x t",
    "nuaíochta": "n uu i@ x t @",
    "am": "a m", # a'm
    "móiméad": "m oo mj ee d",
    "móiméid": "m oo mj ee dj",
    "mhóiméad": "v oo mj ee d",
    "mhóiméid": "v oo mj ee dj",
    "taspánfaidh": "t @ s p aa nn h @", # 0 t @ s . 1 p aa nn . 0 h @
}
_ALTS = {**nonwords, **maybe_missing}

_SOUGHT = []
_SKIP_ALTS = []
for (a, b) in alts.items():
    if type(b) == list:
        for x in b:
            if x not in _ALTS.keys():
                _SOUGHT.append(x)
    else:
        if b not in _ALTS.keys():
            _SOUGHT.append(b)
_REVERSE_ALTS = {}
for item in alts.items():
    if type(item[1]) == list:
        items = item[1]
    else:
        items = [item[1]]
    for sitem in items:
        if sitem not in _REVERSE_ALTS.keys():
            _REVERSE_ALTS[sitem] = set()
        _REVERSE_ALTS[sitem].add(item[0])

def deletable_schwa_single(word, phones):
    out = []
    out.append((word, phones))
    if len(phones) == 1 and phones[0] == '@':
        out.append((word, ["sil"]))
    else:
        if phones[0] == '@':
            out.append((word, phones[1:]))
            if phones[-1] == '@':
                out.append((word, phones[1:-1]))
        if phones[-1] == '@':
            out.append((word, phones[:-1]))
    return out
def deletable_schwa(wordlist):
    out = []
    for item in wordlist:
        out += deletable_schwa_single(item[0], item[1])
    return out

def nasal_o(item):
    # FIXME: way too simplistic
    word = item[0]
    phones = item[1]
    phonestr = " ".join(phones)
    out = [item]
    if "mó" in word and "m oo" in phonestr:
        outph = phonestr.replace("m oo", "m uu").split(" ")
        out.append((word, outph))
    if "mhó" in word and "v oo" in phonestr:
        outph = phonestr.replace("v oo", "v uu").split(" ")
        out.append((word, outph))
    if "nó" in word and "n oo" in phonestr:
        outph = phonestr.replace("n oo", "n uu").split(" ")
        out.append((word, outph))
    return out

def endswith_list(text, endings):
    for ending in endings:
        if text.endswith(ending):
            return True
    return False

def handle_igh(item):
    if type(item) != tuple:
        raise Exception("item is not a tuple: " + item)
    word = item[0]
    phones = item[1]
    out = [item]
    if word.endswith("igh") and phones[-1] == "@":
        out.append((word, phones[0:-1] + ["ii"]))
    if word.endswith("igh") and phones[-1] == "ii":
        if not endswith_list(word, ["uigh", "aoigh"]):
            out.append((word, phones[0:-1] + ["@"]))
    if word.endswith("dh") and phones[-1] == "@":
        out.append((word, phones + ["x"]))
        out.append((word, phones + ["tj"]))
        out.append((word, phones + ["v"]))
    if word.endswith("dh") and phones[-1] == "x":
        out.append((word, phones[0:-1]))
        out.append((word, phones[0:-1] + ["tj"]))
        out.append((word, phones[0:-1] + ["v"]))
    return out

for line in dictionary.readlines():
    line = line.strip()
    pieces = line.split("\t")
    word = pieces[0]

    phones = pieces[1].split(" ")
    entries = list()
    tmptup = (word, phones)
    entries.append(tmptup)
            
    if word in _SOUGHT:
        for replacement_word in _REVERSE_ALTS[word]:
            tmp_replace = [(replacement_word, b) for (a, b) in entries]
        entries.extend(tmp_replace)
    elif word in alts.keys():
        if type(alts[word]) == list:
            tmp_words = alts[word]
        else:
            tmp_words = [alts[word]]
        for tmp_word in tmp_words:
            if tmp_word not in _SOUGHT:
                entries.append((word, _ALTS[tmp_word].split(" ")))

    if word.endswith("acha") or word.endswith("anna"):
        entries.append((word, phones[:-1] + ["ii"]))
    if word.endswith("igh") or word.endswith("dh"):
        tmp_igh = []
        for entry in entries:
            tmp_igh.extend(handle_igh(entry))
        entries.extend(tmp_igh)
    tmp_nasal = []
    for entry in entries:
        tmp_nasal.extend(nasal_o(entry))
    entries.extend(tmp_nasal)
    tmp_schwa = deletable_schwa(entries)
    entries.extend(tmp_schwa)

    joined = [" ".join([a] + b) for (a, b) in entries]
    sort_join = sorted(joined)
    for entry in set(sort_join):
        dictionary2.write(entry + "\n")

dictionary.close()
dictionary2.close()