Output text with style in a variety of formats. This is useful for highlighting target text to improve reviewing experience in NLP scenarios.

Styled output formats include:

• html for jupyter notebooks
• excel workbooks

Within a notebook cell:

>>> from styled_text import StyledText
>>> txt = "This is a long sentence."
>>> StyledText.text_to_notebook_output(txt,[(5,7),(8,9),(15,23)], True)

<z style="color: gray;" >This <b style="color: black;">is</b> <b style="color: black;">a</b> long <b style="color: black;">sentence</b>.</z>
This is a long sentence.

Using Doccano text labeling application:

• group individual messages into single dialogues
• upload dialogues into Doccano
• download jsonl, then process with StyledText.df_to_excel()

For detailed instructions, see ./tests/doccano_workflow.md.

pip install ./styled_text/.

pytest --collect-only

• overlapping highlights
• condense code to improve maintainability
• make more robust
• functionality for additional output mediums