lingdb / Sound-Comparisons

Exploring phonetic diversity across language families —
http://www.soundcomparisons.com
Other
13 stars 8 forks source link

Phonetic transcriptions without sound files #351

Open PaulHeggarty opened 8 years ago

PaulHeggarty commented 8 years ago
runjak commented 8 years ago

Can you give me an example where to find current entries with -- or ... in them?

PaulHeggarty commented 8 years ago

I don’t think there are any (or perhaps a few in Romance), I had thought we would do it in the reverse order: you get this ready, and then we apply it to the transcriptions. If you want a test, I can send you an SQL with some examples.

PaulHeggarty commented 8 years ago

Here is an sql file with about ten words that now have .. (TWO dots, better in fact) as the transcription. in this language: (http://www.soundcomparisons.com/#/en/Malakula/language/Taute%3A%20Alpalak) Currently that page shows 'soon' as the transcription, but ifor you upload the sql it will show ".." for these words. Ideally your code should recognise ".." as a special case, and display -- (two hyphens), and without without the enclosing phonetic [ ] brackets, to show that there's nothing phonetic, just no data. 2016_05_18__Transcriptions_Malakula215-_Rgn1-_North_W-Nth.zip

runjak commented 8 years ago

Thanks - this should make it easier for me to get the desired results.

runjak commented 8 years ago

So I know where to look for such cases, I'm only a bit uncertain as to how you would like this to be displayed on the site. The current way this would be displayed: screenshot_2016-05-25_14-43-55 Possible alternatives: screenshot_2016-05-25_14-44-35 screenshot_2016-05-25_14-45-18

PaulHeggarty commented 8 years ago

The first alternative is fine. There are pros and cons of actually not even having these words showing for that language, but for now let's leave them showing. One other consideration is what might affect the search/filter function on transcriptions. Is there a way of telling it that these fields are to be completely ignored? The reason is that some transcriptions will include hyphens and perhaps even dots, and/or they may have values in regular expressions. We also don’t want them to be included in search results for ^$ or ., I suppose.

runjak commented 8 years ago

I've deployed the changes from this issue and think that they should work as intended. It's possible that the usual caching problems apply.

PaulHeggarty commented 8 years ago

I actually put a confusing title on this issue, and there are two cases, one of which still remains. This is that there are now lots of the Malakula transcriptions that are given as .. in the database. On the map, the corresponding language dot appears, but without any transcription bubble. This is fine. On the table views, however, it would be nice to have the two dots showing, as per the second of Jakob's images above. In fact, the cell appears but the two dots do not. Could you put them in (I hope it won't be too complex to do this only in the tables, and not on the map.) screenshot - 27_05_2016 18_31_17

PaulHeggarty commented 8 years ago

I actually put a confusing title on this issue, and there are two cases, one of which still remains. This is that there are now lots of the Malakula transcriptions that are given as .. in the database. On the map, the corresponding language dot appears, but without any transcription bubble. This is fine. On the table views, however, it would be nice to have the two dots showing, as per the second of Jakob's images above. In fact, the cell appears but the two dots do not. Could you put them in (I hope it won't be too complex to do this only in the tables, and not on the map.) screenshot - 27_05_2016 18_31_17

PaulHeggarty commented 8 years ago

Sorry, I didn’t realise that I'd never actually hit the comment button on the above, so this issue is still actually live, albeit with only a very minor tweak left to fix. Shame I didn’t realise this before you did the last deployment.

PaulHeggarty commented 8 years ago

Cells that are entirely blank in the 1 Word all Languages view (i.e. without any language, just the spare empty cells at the end of the last row in a language region group) nonetheless have the two dots showing, when they should be just blank.
screenshot - 15_06_2016 16_29_46