keytext 0.3

Keyword based text extraction Pacakage (keytext)

keyword based text extraction toolkit (keytext)

What is it?

keytext is an all-in-one versatile and efficient Python package designed for keyword-based text search, manipulation, and data cleansing. Whether you need to extract contextual information around specific keywords, remove unwanted terms from texts and dataframes, or precisely locate the positions of keywords within a Pandas DataFrame, keytext is your indispensable toolkit for advanced robust toolkit text analysis and data management.

Main Features

Here are just a few of the things that keytext does well:

• Keyword Positioning: Locate the exact start and end positions of a keyword within a given text, facilitating precise information retrieval.
• Contextual Extraction: Extract left and right texts, characters, words, and sentences surrounding a specified keyword as well as words between .
• Flexible Configuration: Customize the number of left and right characters, words, or sentences to tailor the extraction to your specific requirements.
• Text Between Keywords: Extract the text between two occurrences of the same keyword, offering deeper insights into the context of your data.
• Word Removal: Efficiently remove a list of specified words from texts, enhancing text cleanliness and relevance.
• Dataframe Cleansing: Seamlessly remove unwanted words from text columns in Pandas DataFrames, ensuring data integrity.
• Cell Positioning in DataFrame: Identify the row and column positions of a keyword within a Pandas DataFrame, enabling precise data manipulation.
• Random Pattern Search: Check for arbitrary patterns or regular expressions within the text data of a DataFrame, uncovering hidden insights and potential correlations.
• Easy Integration: Integrate keytext into your Python projects effortlessly, enhancing your text processing and data cleansing workflows.

Installation Procedure

PyPI
pip install keytext==0.3

Dependencies:

• Regex - Adds support to itterating and finding keywords from the text and dataframe
• Pandas - Adds support to deal with dataframe

Functionalities (with parameters description):

keytext.keypos_text(keyword, text)

- Return all starting and ending position of the keyword from a giuven text
- Output will be in list of tuples

keytext.extract_sents(keyword, text, format)

- This function extract all the sentences from a giuven text that contain the keyword
- By default format is l, that means list of sentences. If we pass p then the outpt format will be paragraph.

keytext.extract_words(keyword, text, left, right)

- This function extract the neighbourhood words of the keyword from a given text.
- In case of left_w = 0, right_w = n it will provide n number of words from the right side of the keyword, n should be an integer
- In case of left_w = m, right_w = 0 it will provide m number of words from the left side of the keyword, m should be an integer
- In case of left_w = m, right_w = n it will provide m number of words from the left side of the keyword, n number of words from the right side of the keyword

keytext.extract_chr(keyword, text, left_chr, right_chr)

- This function extract the neighbourhood charecters of the keyword from a given text.
- In case of left_chr = 0, right_chr = n it will provide n number of charecters from the right side of the keyword, n should be an integer
- In case of left_chr = m, right_chr = 0 it will provide m number of charecters from the left side of the keyword, m should be an integer
- In case of left_chr = m, right_chr = n it will provide m number of charecters from the left side of the keyword, n number of charecters from the right side of the keyword

keytext.left_texts(keyword, text, occurrence)

- This function will return the left side of the keyword i.e. from the keyword to beginning of the text based on all occurence of keyword
- If we pass the 1 or 2 in occurence then it will return the left side text of 1st or 2nd occurence of the keyword from a text, Occurene should be 1,2,...,n,'all'
- Provid ethe output in list format if occurence is "all"

keytext.right_texts(keyword, text, occurrence)

- occurence means the repeation of the keyword in  text
- This function will return the right side of the keyword i.e. from the keyword to ending of the text based on all occurence of keyword
- If we pass the 1 in occurence then it will return the right side text of 1st occurence of the keyword from a text, Occurene should be 1,2,...,n,'all'
- Provid ethe output in list format if occurence is "all"

keytext.between_fixed_keyword(keyword, text)

- Provide the part of the text between two same keyword
- Output will come in list format

keytext.between_distinct_keywords(keyword_start, keyword_end, text, keyword_start_occurence, keyword_end_occurence)

- keyword_start_occurence indicates the the repeatition of the starting keyword in given string
- keyword_end_occurence indicates the the repeatition of the starting  keyword in given string
- Provide the part of the text between two distinct keyword
- Output will come in list format
- For getting all snap texts in list format pass keyword_start_occurence = 0 and keyword_end_occurence = 0

keytext.text_keyword_remover(remover_list, text, replaced_by)

- This function remove the keyword from the text
- Non alphanumeric charecters need to be write in regex format

keytext.keypos_df(keyword, dataframe)

- Return all cells position of the keyword from a giuven dataframe
- Output will be in list of tuples

keytext.dataframe_keyword_remover(remover_list, dataframe, replaced_by)

- This function remove the keyword from the dataframe
- Non alphanumeric charecters need to be write in regex format

keytext.dataframe_pattern_finder(pattern, dataframe)

- This function find the regex patterns from the dataframe the keyword from the dataframe
- It will return the matched word with cell identity

Contributing to pandas

All contributions, bug reports, bug fixes, documentation improvements, enhancements, and ideas are welcome. Feel free to ask questions on the mailing list

Change Log

0.1 (03/01/2024)

• First Release

0.2 (03/01/2024)

• Second Release

0.3 (03/01/2024)

• First Release