bug#17742: Acknowledgement (Support for enchant?)

[Reuben Thomas]

On 19 December 2016 at 17:37, Agustin Martin <agustin6martin@gmail.com> wrote:

About Enchant, last time I looked at it it was mostly intented for use
through libenchant, not through the standalone enchant binary, which was
more like some kind of testing tool. As a matter of fact its list of
options is quite short and it seems to lack support for personal
dictionaries.

​I have a patch currently under consideration by enchant's maintainer to upgrade the functionality of the standalone enchant binary to be similar to ispell, hunspell etc.​

See https://github.com/rrthomas/enchant/

 

Also, there is no easy way to know which particular spellchecking engine is
being used. Enchant uses $(datadir)/enchant and ~/.enchant config files to
define preferences, but I see no way to make enchant tell which one is being
used. So, it is not easy to parse dictionary info.

​enchant-lsmod lists the engine used for each dictionary, so this is certainly possible, as I said earlier. However, it's not a good idea, I think, to directly search for dictionary files and parse them, because this breaks the abstraction of using the binary and/or library.​

Sorry if I have missed some things. Gmail tags some of Reuben mails as spam
and puts them out of my usual workflow.

​This is caused by an unfortunate interaction between my domain's mail configuration (which uses strict DMARC, so that if email does not match some cryptographic signature, it should be discarded by receiving domains) and the gnu.org listserv (which changes outgoing mail but keeps the original sender, so that messages appear to come from me but do not match the signature).

I have raised this before with the listserv admins, but they've not yet been able to update their configuration (I know that sufficiently recent mailman can be configured to work).​

--

http://rrt.sc3d.org