Cross-platform ENIAM with useful bindings 🚀

This project provides only a wrapper for ENIAM (see online source code) Author of original code is Wojciech Jaworski.

Using eniam library

First, you have to install eniam by calling pip install eniam
Then you can try parsing some example phrases (see examples/example.ipynb):

# !pip install opencv-python
# !pip install camelot-py
# !pip install wand
from eniam import *

dom_result = Eniam(['KOT', 'MYSZ'], {
    rule(lemma='kot',pos=subst,case=nom):     gram('KOT'),
    rule(lemma='gonić',pos=fin,person=ter):   (ip<gram('KOT'))>gram('MYSZ'),
    rule(lemma='mysz',pos=subst,case=acc):    gram('MYSZ'),
    root_rule():                                  s % ip,
}, ['KOT', 'MYSZ', 'ZDARZENIE'], {
    valence_rule('kot', 'noun'): 'KOT',
    valence_rule('mysz', 'noun'): 'MYSZ',
    valence_rule('gonić', 'verb'): 'ZDARZENIE',
}).dom("Kot goni mysz.")

dom_result.show()

Eniam syntax

The eniam library uses custom primitive AST the most of the operations looks exactly the same as in normal ENIAM convensions except for those modifications:

/ is replaced with >
\ is replaced with <
ip{ |x1, |x2, |x3... |xn} is replaced with ip[ x1 | x2 | ... | xn]
?x is replaced with function optional(x)
a\?(x) is replaced with a % x
the grammar can contain inline literals created using gram() function for example rule(lemma='kot',pos=subst): gram('np/np\\np'), would be a valid rule.

Useful methods

Showing the result in Jupyter notebook Eniam(...).dom("Kot goni mysz.").show()
Getting the html code Eniam(...).dom("Kot goni mysz.").html()
Saving the html code Eniam(...).dom("Kot goni mysz.").save_html('output_file')
Multiple input sentences
Showing the result in Jupyter notebook Eniam(...).dom(["Kot goni mysz.", "Mysz goni kota."]).show()

Using raw ENIAM interface

Running inside docker

To run ENIAM docker please use the following commands:

    # Run subsyntax tool
    $ docker run -it styczynski/eniam:1.0 /root/subsyntax --help
    # Run lexicon printer
    $ docker run -it styczynski/eniam:1.0 /root/print_lexicon --help
    # Run DOM parser
    $ docker run -it styczynski/eniam:1.0 /root/domparser --help

Running with Python wrapper

You can install Python wrapper to get a nice wrapper around the docker container (this requires Python +3 and Docker installed):

    $ pip install eniam
    $ eniam-cli sub --help
    $ eniam-cli lex --help
    $ eniam-cli dom --help

Note: Before installing you may wish to add python /bin/ directory to the PATH variable. On MacOS for Python 3.7 that would be:

    $ export PATH="$PATH:/usr/local/Cellar/python/3.7.5/Frameworks/Python.framework/Versions/3.7/bin"

See setting up Python PATH

Metadata

Release files for eniam 1.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for eniam 1.0
File	Size	Uploaded
eniam-1.0.tar.gz	96.1 kB	May 16, 2020	Details

Built distribution (wheel)

Table of built distributions (wheels) for eniam 1.0
File	Interpreter	ABI	Platform	Reset
eniam-1.0-py3-none-any.whl 211.6 kB May 16, 2020	Python 3	none	any	Details

Total release size: 307.6 kB

Release files / eniam-1.0.tar.gz

Download URL	eniam-1.0.tar.gz
Size	96.1 kB
Tags	Source
SHA-256 checksum How to use checksums	`9b76640c0c9fe19198e0633f4beb8c16cdf35b6148b539e028b3ce0f27a7cde5`
BLAKE2b-256 checksum How to use checksums	`8cb8766dcf47eb0f8bfb9c390d196e73745794e82a462f5d557d94e027648a40`
Upload date	May 16, 2020
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/2.0.0 pkginfo/1.5.0.1 requests/2.22.0 setuptools/41.6.0 requests-toolbelt/0.9.1 tqdm/4.38.0 CPython/3.7.5`

Release files / eniam-1.0-py3-none-any.whl

Download URL	eniam-1.0-py3-none-any.whl
Size	211.6 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`f54ab07194f3844292a156f4c4d5f7d83699b3d3a948a90c9875c9a773f4f5bf`
BLAKE2b-256 checksum How to use checksums	`c77244612140ace533edeb646b89c303db2ce494d3ef57e316ff5ad48a80e20d`
Upload date	May 16, 2020
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/2.0.0 pkginfo/1.5.0.1 requests/2.22.0 setuptools/41.6.0 requests-toolbelt/0.9.1 tqdm/4.38.0 CPython/3.7.5`