The abbreviation "cf." is not properly read in French

[FiableDotBiz]

Steps to reproduce

spd-say -l fr cf.

Obtained behavior

The abbreviation "cf." is read letter by letter "c. f." ([seɛf] in the international phonetic alphabet).

Expected behavior

The abbreviation "cf." should be read letter "confer" ([kõˈfɛʁ] in the international phonetic alphabet). This is because this abbreviation is very used, well-known (it appears even in the dictionary of the French academy ( https://www.dictionnaire-academie.fr/article/A9C3488-A ) and unambiguous.

Behavior information

Please follow the next step, to provide us with precious information about how things went wrong on your machine:

• Set LogLevel to 5 in your /etc/speech-dispacher/speechd.conf
• Restart your computer to make sure it gets taken into account
• Reproduce the issue
• Make a copy of the speech-dispacher.log file, it is usually found in /run/user/1000/speech-dispatcher/log. Also make a copy of yourspeechmodule.log (e.g. espeak-ng.log).
• Set LogLevel back to 3 in your /etc/speech-dispacher/speechd.conf
• Attach the speech-dispatcher.log and yourspeechmodule.log file to your bug report.
• Also attach your speechd.conf file and yourspeechdmodule.conf file.

Distribution

updated Fedora workstation 38

speech-dispatcher.log espeak-ng.log espeak-ng-mbrola.log

Version of Speech-dispatcher

0.11.4

[sthibaul]

Mmmm, in this case we can probably introduce a generic rule that all synthesis will benefit from, indeed.

[FiableDotBiz]

I don't know if there is enough artificial intelligence in speech-dispatcher or in Orca, but the reading of other abbreviations could be contextualized. For instance if it's detected that the text is a Christian one, the Bible book abbreviations should be read fully: "Gn" → "Genèse", "Ex" → "Exode" etc., as a human reader would do.

[sthibaul]

There is no AI in the speech stack :) I actually don't think we should put any, as in: we'd rather have a deterministic behavior.