GNOME Bugzilla – Bug 651217
[blocked] When braille contraction is enabled, after typing a second capitalised word the cursor has the wrong ofset.
Last modified: 2018-02-08 12:49:50 UTC
Created attachment 188729 [details] Debug.out with possible show why happening this problem Dear Developers, I found a possible other bug with Orca contraction braille feature: If a translation table containing capsign indicator, and typed two word with capital beginning, Orca puts the cursor with wrong place after the second capital beginning word begin typing task. Reproducation example with I always possible reproducing this issue: If you for example choosing american english grade1 braille, and typed Papa Papa words, in braille display the cursor will be have wrong place after you begin typing second capital beginning Papa word. Orca wrong offsetting the cursor after second capital beginning word typing. When I testing this problem, and typed the first Papa word and press space, the cursor right have after the space character. But when I begin typing the second Papa word, the cursor will be last p letter, not after the last a letter after I typed last a letter. If you would like reproducing this issue, after first Papa word typing enough to typing the second capital p letter, the cursor will be have the capital sign indicator, not after the capital P letter. Attila
The problem is happening if src/orca/braille.py source file the contractLine function using louis.compbrlAtCursor mode. If this mode is used, the cursor highlighted word is translating with computer braille format with louis.translate function. If the mode is 0, always positioned right the braille cursor. Attila
Hi Attila, This is really a liblouis bug. I have added it to the todo file in liblouis svn (rev 766), can you please look at it and send a patch with a modified harness file? Thanks.
Unfortunately I don't understand this, of course looked the TODO file. What input text you would like the testing harness file, what table and what entries you would like to define? I need with you wanted mode, cursorPos, brlCursorPos field values. Unfortunately this day I have got lot of other works, but if I have little time I will be sending you a test harness file to test Liblouis side this problem with harness system. Attila
Created attachment 222384 [details] [review] Test patch, don't commit to Liblouis Mesar, runHarness.py sure testing cursor failures? Possible I am not right add the test harness source file the cursor test, but runHarness.py file not showing failures. How can possible testing to cursor land wrong place if mode is compbrlAtCursor with capital texts? Attila
Thanks Attila, I managed to get my lou_harnessGenerator to work. One day we should really find out why liblouis sometimes work and sometimes doesn't. Even if you install everything in their places, restart etc. Anyway, I have committed a testcase which illustrates the problem.
Closing blocked bugs because we need the issue fixed in the blocker.