Repository files navigation

A command-line sound change applier written in Python.

To begin using this sound change applier, you just need to download the executable (.exe) file in theReleases page and use it through the command-line prompt:

$ sca sca-options

You can also clone this repo and use thesca.bat orsca.sh file:

$ git clone https://github.com/erickcan/sound-change-applier.git$ sca sca-options# if Windows$ ./sca.sh sca-options# if Linux

sca (-f rules-file words-file | -n named-rules-json named-rule words)    [--csv-output] [-s sound-classes-json | --no-sound-classes]

• --file-based-sound-change |-f to apply a set of rules to a set of words, each defined in a separate file, creating a text file with the changed words;
  • rules-file: text file with the rules
  • words-file: text file with the words
• --named-sound-change |-n to apply a named sound change to words passed on the command line, and then prints the words after the change;
  • named-rules-json: JSON file where sound classes are defined
  • named-rule: name of the rule to apply (defined innamed-rules-json)
  • words: words to apply the sound change
• --csv-output to create a before and after of the changed words;
• --sound-classes-file |-s to specify the JSON file where the sound classes are defined (see alsoDefault sound classes);
• --no-sound-classes to not make use of sound classes.

Suppose the following files:

named_rules.json

{"terminal-devoicing":"[bdgz] > [ptks] / _#","h-dropping":"h -> _ / _","z-rhotacization":"z -> r / V_V","l-vocalization":"l => w / _!V","/æ/-raising":"æ => eə / _N","th-fronting":"[θð] -> [fv] / _"}

rules.txt

d => ð / V_V[gk] > _ / #_na -> ə / _#l -> w / _!V[aou]e > [æøy] / _sw > s / _[ou][ao]N -> [ãõ] / _s => z / [bdg]_e -> _ / CC_#

words.txt

aeonbedscluedaemonknowledgelateralsmediasandswords

If this were passed to the command line,

$ sca -f rules.txt words.txt --csv-output

the following file would be created:

sound-change-YYYY-MM-DD-HH-MM-SS.csv

aeon,æõbeds,bedzclue,clydaemon,dæmõknowledge,nowledglaterals,laterawsmedia,meðiəsand,sãdswords,sordz

And if the following were passed into the command line,

$ sca -n named_rules.json h-dropping "here he had hallucinated"

it would print:

ere e ad allucinated

The rules can be written in any of the following forms:

x  > y / a_bx -> y / a_bx => y / a_b

wherex becomesy whenx is betweena andb. (Spaces between>,->,=> and/ are insignificant.)

If neither--sound-classes-file nor--no-sound-classes are used, the following sound classes are used:

{"V":"aeiou","C":"bcdfghjklmnpqrstvwxyz","P":"pbtdkg","F":"fvsz","N":"mn","S":"sz"}