kawadi
kawadi (કવાડિ in Gujarati) (Axe in English) is a versatile tool that used as a form of weapon and is used to cut, shape and split wood.
kawadi is collection of small tools that I found useful for me more often. Currently it contains text search which searches a string inside another string.
Text Search
Text search in kawadi uses sliding window technique to search for a word or phrase in another text. The step size in the sliding window is 1 and the window size is the size of the word/phrase we are interested in.
For example, if the text we are interested in searching is "The big brown fox jumped over the lazy dog" and the word that we want to search is "brown fox".
text = "The big brown fox jumped over the lazy dog"
interested_word = "brown fox"
window_size = len(interested.split()) -> len(["brown", ["fox"]])
slides = sliding_window(text, window_size) -> ['The', 'big']['big', 'brown']['brown', 'fox']['fox', 'jumped']['jumped', 'over']['over', 'the']['the', 'lazy']['lazy', 'dog']
for each slide in slides
score(" ".join(slide), interested_word)
if score >= threshold then
select slide
else
continue
Currently, there are 3 similarity scores are calculated and averaged to calculate the final score. These similarity scores are Cosine, JaroWinkler and Normalized Levinstine similarities.
In development
- Add functionality to accept custom user similarity metrics.
- [] Generate documentation.
- [] Write the custom counter
You can follow the project development in the Projects tab.
Quick Start
from kawadi.text_search import SearchInText
search = SearchInText()
text_to_find = "String distance algorithm"
text_to_search = """SIFT4 is a general purpose string distance algorithm inspired by JaroWinkler and Longest Common Subsequence. It was developed to produce a distance measure that matches as close as possible to the human perception of string distance. Hence it takes into account elements like character substitution, character distance, longest common subsequence etc. It was developed using experimental testing, and without theoretical background."""
result = search.find_in_text(text_to_find, text_to_search)
print(result)
[
{
"sim_score": 1.0,
"searched_text": "string distance algorithm",
"to_find": "string distance algorithm",
"start": 27,
"end": 52,
}
]
If the text that needs to be searched is big, SearchInText can utilize multiprocessing to make the search fast.
from kawadi.text_search import SearchInText
search = SearchInText(multiprocessing=True, max_workers=8)
Custom user defined score calculation.
Its often the case that the provided string similarity score is not enough for the use case that you may have. For this very case, you can add, your own score calculation.
from kawadi.text_search import SearchInText
def my_custom_fun(**kwargs):
slide_of_text:str = kwargs["slide_of_text"]
text_to_find:str = kwargs["text_to_find"]
# Here you can then go on to do preprocessing if you like,
# or use them to count char based n-gram string matching scores.
return score: float
search = SearchInText(search_threshold=0.9, custom_score_func= your custom func)
This custom score function will have access to two things slide_of_text for every slide in text (From the example above, "The big", "big brown" and so on...) and text_to_find.
Note: The return type of this custom function should be same as the type of
search_thresholdas you can see from the above example.
Installation
Stable Release: pip install kawadi
Development Head: pip install git+https://github.com/jdvala/kawadi.git
Development
See CONTRIBUTING.md for information related to developing the code.
Free software: MIT license
10 Oct 13, 2022
3 Dec 5, 2021
1 Mar 31, 2022
8 Mar 30, 2022
6 Nov 9, 2021
6 Aug 6, 2021
13 Sep 19, 2022
24 Aug 28, 2022
158 Dec 22, 2022
3 Nov 03, 2022
53 Sep 07, 2022
4 May 01, 2022
26 Jun 26, 2022
223 Dec 24, 2022
3 Aug 06, 2022
1 Nov 17, 2021
7 May 21, 2022
2 Feb 03, 2022
0 Dec 11, 2022
1 Oct 19, 2021
19 Sep 13, 2022
2 Jan 29, 2022
1 Dec 12, 2021
11 Jul 30, 2021
3 Sep 15, 2021
9 Sep 30, 2022
7 Apr 19, 2022
7 Aug 27, 2021