The repository includes shared library (located in /lib directory for various platforms) that can splitting UTF-8 text into paragraphes and sentences.
Additional, the repository includes two projects with safe FFI-wrappers for the library written in Rust and C# languages. There are own readme files for more information.
There is a cli utility that can splits text into several parts about equals length. You can find source here. Here is a help of the utility:
USAGE:
book-parse-shell.exe [FLAGS] [OPTIONS] <source>
ARGS:
<source>
Source file that will be parsed.
FLAGS:
-c
Starts count a parsed part's number from 0 instead of 1.
-h, --help
Prints help information
--verbose-splitting
Show verbose info when splitting stady is active
-V, --version
Prints version information
--view-index-sentence
Inserts sentences' index into output. Sentences' index format: (P:N:U) where P: current
`part` count number, N: sentence's count number inside paragraph, U: sentence's count
number inside the book (it's also an unique index).
OPTIONS:
--max <max>
Recommended maximum size of splitted parts. [default: 600]
--min <min>
Every splitted part should contains at least this count symbols. [default: 200]
-o, --output <output>
Path to output file to save parsed text. If this options is empty, parsed data will be
written into stdout.
-p, --parts-separator <parts-separator>...
Sets template for parts separator to inserting its instead of empty rows. Default is:
`## {} ##` if this argument was passed without a value.
-s, --split-by-paragraph <split-by-paragraph>
Parsed text will be splitted by paragraphes that contain a single sentence with given
string.