search

desilinguist / wordalignui

Java application for creating bilingual word alignments
3 stars 2 forks source link

alignUI: Indian Language support with utf-8 encoding #2

Open indiclinguist opened 7 years ago

indiclinguist commented 7 years ago

I find this tool very useful. So I compiled the jar as per the instruction given in manula.html. However, I am unable to work on Bangla (e.g., vrinda in MS) and Oriya (e.g., Kalinga in MS) Scripts. Sample parallel files are attached. It works correctly for test.hn (hindi) and English. The exact issue is that the alignUI does not display test.mn or test.or sentences. It shows blank boxes as shown in the image file. But, when I link the words blindly it creates desired output file as expected in UTF-8 (aligned.2 file in the sample-utf8-data.zip). Neither the sentence pane nor the Wordbuttons displays the test.mn or test.or text properly.

The other minor issues are: (1) There is no warning pop-up as described in the manual, which could prevent incomplete alignment of any pair. (2) The current sentence number is displayed. It could be really useful for the annotator to have the Total number# and the completed sentence number# displayed along side the current Sentence ID (pair_id) at the top frame.

Can anyone help to resolve the above issues? Thanks in Advance.

sample-utf8-data.zip

desilinguist commented 7 years ago

Thanks for filing the issue. However, I'm going to be away at a conference starting Saturday for a couple of weeks so I won't have a chance to look at this until then. On Thu, Aug 31, 2017 at 12:11 AM indiclinguist notifications@github.com wrote:

I find this tool very useful. So I compiled the jar as per the instruction given in manula.html. However, I am unable to work on Bangla (e.g., vrinda in MS) and Oriya (e.g., Kalinga in MS) Scripts. Sample parallel files are attached. It works correctly for test.hn (hindi) and English. The exact issue is that the alignUI does not display test.mn or test.or sentences. It shows blank boxes as shown in the image file. But, when I link the words blindly it creates desired output file as expected in UTF-8 (aligned.2 file in the sample-utf8-data.zip). Neither the sentence pane nor the Wordbuttons displays the test.mn or test.or text properly.

The other minor issues are: (1) There is no warning pop-up as described in the manual, which could prevent incomplete alignment of any pair. (2) The current sentence number is displayed. It could be really useful for the annotator to have the Total number# and the completed sentence number# displayed along side the current Sentence ID (pair_id) at the top frame.

Can anyone help to resolve the above issues? Thanks in Advance.

sample-utf8-data.zip https://github.com/desilinguist/wordalignui/files/1265728/sample-utf8-data.zip

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/desilinguist/wordalignui/issues/2, or mute the thread https://github.com/notifications/unsubscribe-auth/AAJ3kFF_uU8y_dJPmQGVXVevojAhS7gPks5sdjJggaJpZM4PIR4C .

indiclinguist commented 7 years ago

OK. You can look at it when you are back. To add a few observation, I find that it works fine in Fedora23. In Mac, it works but fonts are not rendered in the panel and button. In windows 7 and 10, except English the fonts are displayed as white boxes. Hope this helps in resolving the issue.