Perseus Greek Isopsephy Project

isopsephy.py

# -*- coding: utf-8 -*-

import re

data = {}

"""

Data mapping between roman and greek letters, isopsephy values and linguistic components

Resources:

- http://www.perseus.tufts.edu/img/keyCaps.gif for source
- http://www.class.uh.edu/mcl/faculty/pozzi/grnl1/intr/0.2.1.pract.vow.htm
- http://en.wikipedia.org/wiki/Greek_alphabet
- http://www.chlt.org/FirstGreekBook/JWW_FGB1.html
- http://www.webtopos.gr/eng/languages/greek/alphabet/alpha.htm

Segments:
- vowel
- consonant

Subsegments:
- semivowel (liquid, siblant and γ-nasal not specified on data table)
- double
- mute

Mutes (not specified on data table):

{class-order} {letter}

labial-smooth π
labial-middle β
labial-rought φ

palatal-smooth κ
palatal-middle γ
palatal-rought χ

lingual-smooth τ
lingual-middle δ
lingual-rought θ


Seven vowels: e h i o w u

"""

# letters from 1 to 9
#alpha:http://en.wiktionary.org/wiki/ἄλφα
data[1] = {'greek': 'α', 
           'capital': 'Α',
           'name': 'αλφα',
           'segment': 'vowel',
           'subsegment': 'short',
           'roman': 'a',
           'value': 1}
#beta:http://en.wiktionary.org/wiki/βῆτα
data[2] = {'greek': 'β',
           'capital': 'Β',
           'name': 'βητα',
           'segment': 'consonant',
           'subsegment': 'mute',
           'roman': 'b',
           'value': 2}
#gamma:http://en.wiktionary.org/wiki/γάμμα
data[3] = {'greek': 'γ',
           'capital': 'Γ',
           'name': 'γαμμα',
           'segment': 'consonant',
           'subsegment': 'mute',
           'roman': 'g',
           'value': 3}
#delta:http://en.wiktionary.org/wiki/δέλτα
data[4] = {'greek': 'δ',
           'capital': 'Δ',
           'name': 'δελτα',
           'segment': 'consonant',
           'subsegment': 'mute',
           'roman': 'd',
           'value': 4}
#epsilon:http://en.wiktionary.org/wiki/epsilon
data[5] = {'greek': 'ε',
           'capital': 'Ε',
           'name': 'ε ψιλον',
           'segment': 'vowel',
           'subsegment': 'short',
           'roman': 'e',
           'value': 5}
#digamma/stigma/episemon/wau
#http://en.wikipedia.org/wiki/Digamma
data[6] = {'greek': 'ϛ', 'small2': 'ϝ',
           'capital': 'Ϛ', 'capital2': 'Ϝ',
           'name': 'διγαμμα', 'name2': 'στιγμα', 'name3': 'επισημον', 'name4': 'βαυ',
           'segment': 'numeral',
           #'subsegment': '',
           #'roman': '_',
           'value': 6}
#zeta:http://en.wiktionary.org/wiki/ζῆτα
data[7] = {'greek': 'ζ',
           'capital': 'Ζ',
           'name': 'ζητα',
           'segment': 'consonant',
           'subsegment': 'double',
           'roman': 'z',
           'value': 7}
#eta:http://en.wiktionary.org/wiki/ἦτα
data[8] = {'greek': 'η',
           'capital': 'Η',
           'name': 'ητα',
           'segment': 'vowel',
           'subsegment': 'long',
           'roman': 'h',
           'value': 8}
#theta:http://en.wiktionary.org/wiki/θῆτα
data[9] = {'greek': 'θ',
           'capital': 'Θ',
           'name': 'θητα',
           'segment': 'consonant',
           'subsegment': 'mute',
           'roman': 'q',
           'value': 9}

# letters from 10 to 90
#iota:http://en.wiktionary.org/wiki/ἰῶτα
data[10] = {'greek': 'ι',
            'capital': 'Ι',
            'name': 'ιωτα',
            'segment': 'vowel',
            'subsegment': 'short',
            'roman': 'i',
            'value': 10}
#kappa:http://en.wiktionary.org/wiki/κάππα
data[20] = {'greek': 'κ',
            'capital': 'Κ',
            'name': 'καππα',
            'segment': 'consonant',
            'subsegment': 'mute',
            'roman': 'k',
            'value': 20}
#lambda:http://en.wiktionary.org/wiki/λάμβδα
data[30] = {'greek': 'λ',
            'capital': 'Λ',
            'name': 'λαμβδα',
            'segment': 'consonant',
            'subsegment': 'semivowel',
            'roman': 'l',
            'value': 30}
#mu:http://en.wiktionary.org/wiki/mu
data[40] = {'greek': 'μ',
            'capital': 'Μ',
            'name': 'μυ',
            'segment': 'consonant',
            'subsegment': 'semivowel',
            'roman': 'm',
            'value': 40}
#nu:http://en.wiktionary.org/wiki/νῦ
data[50] = {'greek': 'ν',
            'capital': 'Ν',
            'name': 'νυ',
            'segment': 'consonant',
            'subsegment': 'semivowel',
            'roman': 'n',
            'value': 50}
#xi:http://en.wiktionary.org/wiki/ξεῖ
data[60] = {'greek': 'ξ',
            'capital': 'Ξ',
            'name': 'ξει',
            'segment': 'consonant',
            'subsegment': 'double',
            'roman': 'c',
            'value': 60}
#omicron:http://en.wiktionary.org/wiki/omicron
data[70] = {'greek': 'ο',
            'capital': 'Ο',
            'name': 'ο μικρον',
            'segment': 'vowel',
            'subsegment': 'short',
            'roman': 'o',
            'value': 70}
#pi:http://en.wiktionary.org/wiki/πεῖ
data[80] = {'greek': 'π',
            'capital': 'Π',
            'name': 'πει',
            'segment': 'consonant',
            'subsegment': 'mute',
            'roman': 'p',
            'value': 80}
#koppa:http://en.wikipedia.org/wiki/Koppa_(letter)
#http://www.webtopos.gr/eng/languages/greek/alphabet/earlyletters.htm
data[90] = {'greek': 'ϙ', 'small2': 'ϟ',
            'capital': 'Ϙ', 'capital2': 'Ϟ',
            'name': 'κοππα',
            'segment': 'numeral',
            #'subsegment': '',
            #'roman': '_',
            'value': 90}

# letters from 100 to 900
#rho:http://en.wiktionary.org/wiki/ῥῶ
data[100] = {'greek': 'ρ',
             'capital': 'Ρ',
             'name': 'ρω',
             'segment': 'consonant',
             'subsegment': 'semivowel',
             'roman': 'r',
             'value': 100}
#sigma:http://en.wiktionary.org/wiki/σίγμα
data[200] = {'greek': 'σ', 'small2': 'ϲ', 'small3': 'ς',
             'capital': 'Σ', 'capital2': 'Ϲ', 'capital3': 'Σ',
             'name': 'σιγμα',
             'segment': 'consonant',
             'subsegment': 'semivowel',
             'roman': 's',
             'value': 200}
#tau:http://en.wiktionary.org/wiki/tau
data[300] = {'greek': 'τ',
             'capital': 'Τ',
             'name': 'ταυ',
             'segment': 'consonant',
             'subsegment': 'mute',
             'roman': 't',
             'value': 300}
#upsilon:http://en.wiktionary.org/wiki/upsilon
data[400] = {'greek': 'υ',
             'capital': 'Υ',
             'name': 'υ ψιλον',
             'segment': 'vowel',
             'subsegment': 'short',
             'roman': 'u',
             'value': 400}
#phi:http://en.wiktionary.org/wiki/phi
data[500] = {'greek': 'φ',
             'capital': 'Φ',
             'name': 'φει',
             'segment': 'consonant',
             'subsegment': 'mute',
             'roman': 'f',
             'value': 500}
#khi, chi:http://en.wiktionary.org/wiki/chi
data[600] = {'greek': 'χ',
             'capital': 'Χ',
             'name': 'χει',
             'segment': 'consonant',
             'subsegment': 'mute',
             'roman': 'x',
             'value': 600}
#psi:http://en.wiktionary.org/wiki/psi
data[700] = {'greek': 'ψ',
             'capital': 'Ψ',
             'name': 'ψει',
             'segment': 'consonant',
             'subsegment': 'double',
             'roman': 'y',
             'value': 700}
#omega:http://en.wiktionary.org/wiki/omega
data[800] = {'greek': 'ω',
             'capital': 'Ω',
             'name': 'ω μεγα',
             'segment': 'vowel',
             'subsegment': 'long',
             'roman': 'w',
             'value': 800}
#sampi/disigma
#http://en.wikipedia.org/wiki/Sampi
#http://www.tlg.uci.edu/~opoudjis/unicode/other_nonattic.html#sampi
#http://www.parthia.com/fonts/sampi.htm
#http://www.jstor.org/stable/636031
data[900] = {'greek': 'ϡ', 'small2': 'ͳ',
             'capital': 'Ϡ', 'capital2': 'Ͳ',
             'name': 'σαμπι', 'name2': 'δισιγμα',
             'segment': 'numeral',
             #'subsegment': '',
             #'roman': '_',
             'value': 900}

greek_roman_values = {}
greek_roman_letters = {}
roman_greek_letters = {}

keys = ['roman', 'greek', 'capital', 'capital2', 'small2', 'small3', 'small4']
for num, d in data.items():
    for k in keys:
        if d.has_key(k):
            greek_roman_values[d[k]] = num
            if k == 'roman':
                greek_roman_letters[d[k]] = d['greek']
                greek_roman_letters[d[k].upper()] = d['capital']
                greek_roman_values[d[k].upper()] = num
            else:
                if d.has_key('roman'):
                    if k == 'capital' or k == 'capital2':
                        roman_greek_letters[d[k]] = d['roman'].upper()
                    else:
                        roman_greek_letters[d[k]] = d['roman']

regex_greek_roman_values = re.compile('|'.join(greek_roman_values.keys()))
regex_greek_to_roman_letters = re.compile('|'.join(roman_greek_letters.keys()))
regex_roman_to_greek_letters = re.compile('|'.join(greek_roman_letters.keys()))

def isopsephy(str):
    """
    Str is a roman letter representation (transliteration) of the greek word or sentence 
    that will be converted to the numerical value letter by letter
    """
    str = regex_greek_roman_values.sub(lambda x: '%s ' % greek_roman_values[x.group()], str)
    return sum([int(i) for i in str.split()])

def to_roman(word):
    """
    Create a roman letter version of the greek word.
    This will change all greek (primary), capital, capital2, small2, small3, and small4
    letters to roman letter. Capital letters are honored.
    """
    return regex_greek_to_roman_letters.sub(lambda x: roman_greek_letters[x.group()], word)

def to_greek(word):
    """
    Create a greek version of the roman letter word.
    This will change a-zA-Z except j, J, v & V to the corresponding greek letters
    Capital letters are honored.
    """
    return regex_roman_to_greek_letters.sub(lambda x: greek_roman_letters[x.group()], word)

names = {'name': 'name_value', 'name2': 'name_value2', 'name3': 'name_value3', 'name4': 'name_value4'}
for num, d in data.items():
    for k, v in names.items():
        if d.has_key(k):
            d[v] = isopsephy(d[k])

# accents / diacritics for simplified greek letters
accents = {}

accents['Ὑ'] = 'υ'
accents['Ὕ'] = 'υ'
accents['ὖ'] = 'υ'
accents['ῦ'] = 'υ'
accents['ύ'] = 'υ'
accents['ὗ'] = 'υ'
accents['ὐ'] = 'υ'
accents['ὑ'] = 'υ'
accents['ϋ'] = 'υ'
accents['ὔ'] = 'υ'
accents['ὺ'] = 'υ'
accents['ὕ'] = 'υ'
accents['ὓ'] = 'υ'

accents['Ἠ'] = 'η'
accents['Ἦ'] = 'η'
accents['Ἤ'] = 'η'
accents['Ἡ'] = 'η'
accents['ἠ'] = 'η'
accents['ἦ'] = 'η'
accents['ὴ'] = 'η'
accents['ῇ'] = 'η'
accents['ἡ'] = 'η'
accents['ή'] = 'η'
accents['ῃ'] = 'η'
accents['ῆ'] = 'η'
accents['ἥ'] = 'η'
accents['ἢ'] = 'η'
accents['ᾖ'] = 'η'
accents['ἤ'] = 'η'
accents['ῄ'] = 'η'
accents['ᾗ'] = 'η'
accents['ᾔ'] = 'η'
accents['ἣ'] = 'η'
accents['ἧ'] = 'η'
accents['ᾐ'] = 'η'
accents['ή'] = 'η'
accents['ῇ'] = 'η'

accents['Ἰ'] = 'ι'
accents['Ἴ'] = 'ι'
accents['Ἱ'] = 'ι'
accents['ῖ'] = 'ι'
accents['ί'] = 'ι'
accents['ῖ'] = 'ι'
accents['ὶ'] = 'ι'
accents['ἰ'] = 'ι'
accents['ἵ'] = 'ι'
accents['ἴ'] = 'ι'
accents['ἱ'] = 'ι'
accents['ἶ'] = 'ι'
accents['ΐ'] = 'ι'
accents['ἷ'] = 'ι'
accents['ϊ'] = 'ι'
accents['ῒ'] = 'ι'
accents['ἳ'] = 'ι'
accents['ἳ'] = 'ι'
accents['ί'] = 'ι'

accents['Ὁ'] = 'ο'
accents['Ὃ'] = 'ο'
accents['Ὅ'] = 'ο'
accents['Ὅ'] = 'ο'
accents[' Ὄ'] = 'ο'
accents['Ὀ'] = 'ο'
accents['Ὄ'] = 'ο'
accents['ὁ'] = 'ο'
accents['ὃ'] = 'ο'
accents['ό'] = 'ο'
accents['ὸ'] = 'ο'
accents['ὅ'] = 'ο'
accents['ὄ'] = 'ο'
accents['ὀ'] = 'ο'

accents['ὰ'] = 'α'
accents['ά'] = 'α'
accents['ἀ'] = 'α'
accents['ᾳ'] = 'α'
accents['ἄ'] = 'α'
accents['ἂ'] = 'α'
accents['ἃ'] = 'α'
accents['ᾶ'] = 'α'
accents['ᾷ'] = 'α'
accents['ἅ'] = 'α'
accents['ἁ'] = 'α'
accents['ἆ'] = 'α'
accents['ά'] = 'α'
accents['ά'] = 'α'

accents['Ὡ'] = 'ω'
accents['ῷ'] = 'ω'
accents['ῳ'] = 'ω'
accents['ώ'] = 'ω'
accents['ῶ'] = 'ω'
accents['ώ'] = 'ω'
accents['ὡ'] = 'ω'
accents['ᾧ'] = 'ω'
accents['ὥ'] = 'ω'
accents['ὢ'] = 'ω'
accents['ὼ'] = 'ω'
accents['ὦ'] = 'ω'
accents['ὧ'] = 'ω'
accents['ὤ'] = 'ω'
accents['ὠ'] = 'ω'
accents['ῴ'] = 'ω'

accents['Ἐ'] = 'ε'
accents['Ἔ'] = 'ε'
accents['Ἑ'] = 'ε'
accents['Ἓ'] = 'ε'
accents['Ἕ'] = 'ε'
accents['έ'] = 'ε'
accents['ἐ'] = 'ε'
accents['ἔ'] = 'ε'
accents['ἑ'] = 'ε'
accents['ὲ'] = 'ε'
accents['ἕ'] = 'ε'
accents['ἐ'] = 'ε'
accents['ἓ'] = 'ε'

accents['Ἀ'] = 'α'
accents['Ἁ'] = 'α'
accents['Ἆ'] = 'α'
accents['Ἄ'] = 'α'
accents['Ἅ'] = 'α'
accents['ά'] = 'α'

accents['Ῥ'] = 'ρ'
accents['ῥ'] = 'ρ'

regex_roman = re.compile(r'[^a-z ]+')
def preprocess_roman(str):
    # regex to remove all special characters leaving only a-z and empty scape
    # for example: a)a/atos tau1 -> aaatos tau
    return regex_roman.sub('', str)

regex_greek = re.compile('|'.join(accents.keys()))
def preprocess_greek(str):
    # handle diacritics
    return regex_greek.sub(lambda x: accents[x.group()], str)

Perseus Greek Isopsephy Project.ipynb

perseus.py

# -*- coding: utf-8 -*-

import re
import pandas as pd
from os.path import isfile
from urllib import urlretrieve
from lxml import etree
from isopsephy import isopsephy, data, to_roman, to_greek, preprocess_roman, preprocess_greek

#fileName = "Perseus_text_1999.04.0058.xml"

def load_dataframe(fileName):

    csvFileName = "%s.csv" % fileName
    xmlFileName = "%s.xml" % fileName

    if isfile(csvFileName):
        print "Retrieving data from local csv copy..."
        return pd.read_csv(csvFileName)
    else:
        return load_dataframe_xml(fileName)

def load_dataframe_xml(fileName):

    xmlFileName = "%s.xml" % fileName

    if isfile(xmlFileName):
        print "Retrieving data from local xml copy..."
    else:
        print "Downloading and saving data to the local copy..."
        urlretrieve("http://www.perseus.tufts.edu/hopper/dltext?doc=%s" % xmlFileName, xmlFileName)

    try:
        tree = etree.parse(xmlFileName)
        print "XML tree loaded for pursuit!"
    except Exception as e:
        print(e.message)

    ids = []
    words = []
    transliterations = []
    translations = []
    isopsephies = []
    charcounts = []

    # perseus dictionary words are divided to alphabetic sections
    for div in tree.xpath("text/body/div0"):
        # entry is the main element
        for entry in div.xpath("entry"):
            # collects word ids
            ids.append(entry.attrib['id'])
            # sanitize transliterated word and add to the collection
            word = preprocess_roman(entry.attrib['key'])
            transliterations.append(word)
            # calculate characters
            charcounts.append(len(word))
            # get isopsephy value of the word (still translitered)
            val = isopsephy(word)
            isopsephies.append(val)
            # create greek version of the word and add to the collection
            words.append(to_greek(word))
            # get translations for the word and add to the collection
            for sense in entry.xpath("sense"):
                trs = []
                for tr in sense.xpath("trans/tr//text()"):
                    tr = tr.encode('utf-8')
                    trs.append(tr)
            translations.append('|'.join(trs))

    df = pd.DataFrame({'Id': ids,
                       'Word': words,
                       'Transliteration': transliterations,
                       'Translation': translations,
                       'Isopsephy': isopsephies,
                       'CharCount': charcounts
                        })

    df.to_csv("%s.csv" % fileName, index=False)

    print "Data collected from xml file and saved to csv"

    return df