Add ability to load wider range of EOL separated text files

[yruslan]

Background

Currently, .option("is_text", "true") is limited to ASCII UTF-8 files. EOL separators allow us to treat a file as a fully variable-length file.

Feature

Implement is_text through variable-length readers to

• [ ] Allow other charsets for ASCII
• [ ] Allow specifying EOL character and escape
• [ ] Allow EBCDIC with code pages
• [ ] Allow multisegment and hierarchical records