rtfcre

rtfcre is a Python library for reading and writing steno dictionaries in the RTF/CRE (Rich Text Format with Court Reporting Extensions) format. The library provides an API similar to that of the json module for reading and writing dictionaries.

Features

• Speed: The parsing logic is written in Rust using parser combinators, making it much faster than practically any pure-Python implementation.

• Comments: Rather than just exposing translations, rtfcre also reads the comments embedded in each entry ({\*\cxcomment like this}).

• Unicode: Full Unicode support -- while the dictionary files are not encoded in UTF-8, Unicode characters in translations are still fully supported. Translations can be in any language and they will seamlessly be converted to escapes when writing.

• Plover support: Translations are converted automatically to Plover's native syntax (e.g. fingerspelling is represented with {&a} rather than {\cxfing a}) and converted back when writing.

Installation

pip install rtfcre

Usage

To read an RTF dictionary:

import rtfcre

# Reading directly from a file (make sure to open binary)
with open("dict.rtf", "rb") as file:
  dic = rtfcre.load(file)

# Reading from a string
rtf = r"""
{\rtf1\ansi{\*\cxrev100}\cxdict{\*\cxsystem KittyCAT}
{\*\cxs KAT}cat
{\*\cxs KOU}cow
}
""".lstrip()
dic = rtfcre.loads(rtf)

To write the RTF dictionary:

# Writing to a file (make sure to open binary)
with open("dict.rtf", "wb") as file:
  dic.dump(file)

# Writing to a string
rtf = dic.dumps()

The dictionary object itself also supports the standard dict API:

dic["KAT"] = "cat"

"KAT" in dic  # True
dic["KAT"]  # "cat"

del dic["KAT"]

dic["TKOG"]  # KeyError
dic["TKOG"] = "dog"
dic["TKOG"]  # "dog"

as well as a reverse lookup API for mapping from translations to steno strokes:

dic.reverse_lookup("cat")  # ["KAT"]

To access comments:

dic.lookup("TKOG")  # ("dog", None)

dic.add_comment("TKOG", "TK means D")
dic.lookup("TKOG")  # ("dog", "TK means D")

dic.remove_comment("TKOG")