Searching BDAG by Semantic Domain

[Λύχνις Δαν]

Hi ya,

 

I was looking over a link concerning the NIDNTTE and ran into mention of Louw and Nida's lexicon again. I wondered if one could search a regular lexicon like say LSJ or BDAG using those classifications. Could I for example pick a L&N domain and then look up all those words in BDAG. It seems I cannot do things like [HITS ?] between tools searching on different fields and in any case its tricky to get the entries to highlight though getting a single domain is easy enough.

 

So I did this by doing the search in L & N and saving the pane contents as a text file in UTF-16. Then processed it with Python to extract the entries and construct a search string which I could paste into BDAG. Laborious ? Well the first time, but perhaps I'll do more of it.

 

If anyone has a better way to do this sort of thing I would love hear about it. Below is the code if anyone wants it.

 

This is the workspace:

 

 

Here is the Python:

 

#!/usr/bin/python
# coding: utf-8
"""
Extract the Greek words from the semantic domain list saved from Accordance
Louw & Nida Greek-English Lexicon of the New Testament Based on Semantic Domains
and construct a query for use in a Tool tab to find hits for all the words.

Requires Python 2.7

Usage:

extractDomainWords.py <TXT file saved from Accordance>

Prerequisites:

You have a text file which is a TXT file save from a search tab on the L & N
lexicon above.

After you run this cut and paste the search string into a search tab. Bear in
mind that you may have to correct some word forms to match what the tool has.
"""

# Copyright Daniel Semler Jan 2014

import codecs, re, sys

def process_args():
"""Return a dictionary with the cmdline arguments
named in a specific way. Keys are :
exe : the name of the executable
filename : the file to process
"""
args = {}
args['exe'] = sys.argv[0]
if len(sys.argv) > 1:
args['filename'] = sys.argv[1]
return args

def find_words(filename):
"""Find all initial words in Greek from each entry. Entries are of the form

27.1 εὑρίσκωb: to learn something previously not known, frequently involving an element of surprise — ‘to learn, to find out, to discover.’ [p. 326] ζητεῖν τὸν θεὸν εἰ ἄρα γε ψηλαφήσειαν αὐτὸν καὶ εὕροιεν ‘to seek God and perhaps find him as they were groping around for him’ Ac 17:27; συνεψήφισαν τὰς τιμὰς αὐτῶν καὶ εὗρον ἀργυρίου μυριάδας πέντε ‘they calculated their value and discovered that they had been worth fifty thousand (drachmas)’ Ac 19:19; οὐχ εὕρισκον τὸ τί ποιήσωσιν ‘but they could not find out how to do it’ Lk 19:48.

It is encoded as UTF-16.

Extract the words up to the first ':' and remove the initial semantic
domain, in this case "27.1 ".
"""
wds = []
for l in codecs.open(filename, encoding = 'utf-16'):
if l is not None:

# Find the words up to the first ':'
sem_word_re = re.compile(u'^\d+.\d+\s+(.*?):.*')
if sem_word_re.search(l) is not None:

# Find the first word up to \s ; ,
w = sem_word_re.search(l).group(1)
trim_word_re = re.compile(u'(.*?)[\s;,].*')
if trim_word_re.search(w) is not None:
sword = trim_word_re.search(w).group(1)
else:
sword = w

# Trim off any trailing a-z footnote indicator
trim_az_re = re.compile(u'(.*?)[a-z]$')
if trim_az_re.search(sword) is not None:
tword = trim_az_re.search(sword).group(1)
else:
tword = sword
wds.append(tword)

return wds

cmd_args = process_args()
if (cmd_args.get('filename') is None):

print __doc__
sys.exit(1)

words = find_words(cmd_args['filename'])

# Construct the query list
print u'(',
num = len(words)
i = 0
for w in words:
print w,
if i < num-1:
print ", ",
i = i + 1
print u')'

 

Thx

D

[dandennison]

This makes me think it would be great to have a search operator like [FUNC foo args] would call out to foo with args.

 

To avoid the inevitable cross-platform bugs, this could make use of Google Cloud Functions or AWS Lambdas.

[Λύχνις Δαν]

I've actually written an AWS Lambda based query against Hebrew and Greek texts. Of course you need a licensed or open source of data but it can be done.

 

Thx

D

[dandennison]

all the more reason to have built-in support in Accordance.

[Λύχνις Δαν]

Yep.

 

thx

D