Skip to content

Latest commit

 

History

History
72 lines (50 loc) · 1.7 KB

README.md

File metadata and controls

72 lines (50 loc) · 1.7 KB

Python Pattern Matching

Inspired by pampy and core.match, this is a pattern matching library for dynamic list of patterns.

Right now it only supports matching lists of same sizes.

Usage

Import the lib

import PatternMatchStore from py_matching_pattern

First you will add initialize a PatternMatchStore with a fixed size of keys to match (size of the list).

pmdb = PatternMatchStore(keysize=3)

After that you can add the patterns. Every value is valid match, including None.

The "catch all" value is default attribute of the instance.

_ = pmdb.default
pmdb.put(keys=["a","b","c"],value=1)
pmdb.put(keys=["a","b","b"],value=2)
pmdb.put(keys=["a","b",_],value=3)
pmdb.put(keys=["a",_,_],value=4)
pmdb.put(keys=["a",None,"c"],value=5)
pmdb.put(keys=["a",_,"d"],value=6)

These patterns are staged for the DB. This is useful if you reload the patterns from another thread.

To make the pattern live, you commit it.

pmdb.commit()

Now you can finally start matching values:

pmdb.get(keys=["a","b","c"])==1
pmdb.get(keys=["a","b","b"])==2
pmdb.get(keys=["a","b","d"])==3
pmdb.get(keys=["a","c","d"])==4
pmdb.get(keys=["a",None,"c"])==5
pmdb.get(keys=["a","e","d"])==6
pmdb.get(keys=["e","e","e"])==None

The staged patterns are kept between commits, so you can add or override to it later.

But in case you want to clean it:

pmdb.clean()

It can also raise an exception if not found, but by default it will return None:

pmdb = core.PatternMatchStore(keysize=3,raise_on_notfound=True)
pmdb.get(keys=["e","e","e"]) # raise KeyNotFound

License

MIT