sammdot / rtfcre

Python library for Rich Text Format with Court Reporting Extensions (RTF/CRE) dictionaries

Sammi De Guzman 陳慧珊 🇵🇭 🇭🇰 🇨🇦

Hi! I'm Sammi (she/her), a software engineer at Twitter, and currently based in San Francisco, California. I studied Computer Science at the University of Toronto, where I previously worked as a teaching assistant.

I like to cook, write code, design pretty websites, take transit, fly virtual planes, and run around a deserted island in my spare time.

rtfcre's Issues

Line breaks and tabs are not escaped properly

\cxds\n\cxds gets translated as {^}\\n{^} rather than {^}\n{^}, which results in writing a literal "\n".

May need to implement \cxplvrnl and \cxplvrtab in order to properly support whitespace characters.

pip install rtfcre NOT WORKING!!!

it says this
pip install rtfcre

ERROR: Could not find a version that satisfies the requirement rtfcre (from versions: none)
ERROR: No matching distribution found for rtfcre

ii tried to clode the repository but still not working
git clone https://github.com/sammdot/rtfcre.git

remote: Enumerating objects: 403, done.
Receiving objects: 99% (399/403)ed 0 (delta 0), pack-reused 403
Receiving objects: 100% (403/403), 65.79 KiB | 455.00 KiB/s, done.
Resolving deltas: 100% (236/236), done.
PS G:\OneDrive - University of Engineering and Technology Taxila\python Freelancig\transcriptions> cd rtfcre

PS G:\OneDrive - University of Engineering and Technology Taxila\python Freelancig\transcriptions\rtfcre> pip install .

Support Plover-specific control words without ignored groups

For control words like \cxplvrcancel or \cxplvrrtfcw that don't have anything coming after them that might be interpreted as content, allow them to be included in the rest of the translation code without being in an ignored group.

For example: {\*\cxplvrortho}\cxds ish may be written simply as \cxplvrortho\cxds ish.

Escape Plover control characters to prevent misparsing

{:^}, {=^}, {#^} get parsed by Plover as a meta, a macro and a key combo respectively.

Recognize non-English steno strokes

Steno groups currently do allow Unicode text, but they remain escaped. Recognizing non-English keys would allow for using theories like plover-korean.

Example:

{\*\cxs \u12615\u12636\u12593\u12599}\u50864\u47532\u44032

translates to the following JSON entry:

{
  "ㅇㅜㄱㄷ": "우리가",
}

Support lookahead translations

Regular expression lookahead translations are now supported in Plover 4.0.0-dev8, allowing translations to change depending on the following text. Add a new control word \cxplvrlap (LookAhead Pattern) to specify the lookahead pattern, and then \cxplvrlam (LookAhead Match) for the text that will be used as the translation when the lookahead matches.

Plover syntax:

{=(?i)t/true/false}
{:if_next_matches:(?i)t/true/false}

Proposed RTF syntax:

false{\*\cxplvrlap (?i)t}{\*\cxplvrlam true}

This would allow steno engines that do not support lookaheads to automatically default to the false (i.e. lookahead does not match) translation.

Dictionary loading fails when last entry contains a command

Example:

{\rtf1\ansi{\*\cxrev100}\cxdict{\*\cxsystem Test}
    {\*\cxs PHROLG}{\*\cxplvrcmd lookup}
    {\*\cxs PHREUT}{\*\cxplvrcmd quit}
  }

Support prefix strokes

Plover 4.0.0-dev8 added support for prefix strokes, which are indicated in the steno part of the dictionary, not the translation. This is incompatible with RTF, so we introduce a new control word \cxplvrprfx to support this behavior, and \cxplvrwdend for the corresponding {$} operator to indicate the end of a word.

Recognize non-Plover control words

Unable to convert Stan Sakai's main.json to rtf

Platform: Windows 10
rtfcre ver: 2.0.1

Steps:

1. Download Stan Sakai's main.json
2. Try to convert to rtf via CLI
3. "thread 'main' has overflowed its stack"

New ortho-attach control word

For specifying whether an attachment should be orthographically aware or just a raw attach. Would solve issues with punctuation attachment especially.

Fix Unicode support

– All characters represented with Unicode escapes should also use placeholder characters for non-supporting RTF readers: e.g. my Luminex II translates "San José" San Jos\u233 as "San Jos", ignoring the e entirely
– Unicode characters from U+8000 to U+FFFF should use negative numbers
– Unicode characters above U+FFFF should be represented in UTF-16 encoding

Reference: https://www.zopatista.com/python/2012/06/06/rtf-and-unicode/

RTF files encoded in ISO-8859-1 (or Windows-1252?) can't be parsed

Emily (EPL/HREU) on the Plover Discord:

i've got a bit of an issue with dictionaries and looking for support
so i just got the magnum steno dictionary in RTF/CRE format all good
and i can load it into plover as RTF and that seems to work, it just takes a long time to do so as it's not native .json
I'd like to convert it to json - but i'm running into issues around UTF8 and such
there are a few entries (such as the one shown above) that have UTF8 codes in them, and rtfcre doesn't like that and I can't work out how to get around this

magnum.rtf: text/rtf; charset=iso-8859-1