Re: Sanskrit input (Devanagari and transliteration)
@ amerikkanu
Are you using your method of input also for hindi writing or only for sanskrit? Both are written in devanagari, but the behaviour with ligatures and vowels is a little bit different. For example:
The following word written in Hindi is transcribed as follow:
खिड़की = khiṛkī = Window
But when you read this word as it would be written in Sanskrit, the transcription would look different:
खिड़की = khiṛakī
Because hindi is only using ligatures for sanskrit words and some other exceptions, most of the hindi words do not have ligatures at all. Additionaly, the vowel a is always added to a consonant in sanskrit, but not in hindi. It's a little bit confusing, but I hope that you know what mean.
Is there a method to configure SCIM for sanskrit and hindi with their special writing rules? How are you handling this problem?
PS: I am using a spaced repetition program to learn vocabulary for hindi and sanskrit. When typing an answer in devanagari, i see dotted circles after signs that have to be ligatures and signs with vowels. How do I get rid of this circles, so that my answer has the exact look of the requested word?
See pictures for clarification:
http://www.box.net/shared/s2m5dxpl65
http://www.box.net/shared/jkpm0xz3e0
http://www.box.net/shared/czaz2sj9k9
Maybe this link will help you: http://groups.google.com/group/ankis...3989cda4fdb0f3
Thanks in advance!
Re: Sanskrit input (Devanagari and transliteration)
I only type in Hindi infrequently, so I've never bothered to use anything but the Sanskrit input. It's surprising that such a popular spoken language as Hindi doesn't have a more intuitive keyboard. Being a linux/computing newb, I found it difficult to come up with a perfectly phonetic keyboard for Hindi. Juggling the rules for halant dropping, consonant conjoining and cons-vowel combination left me with two alternatives: typing the word-internal (but not word-final) 'a' vowels, but having consonants combine automatically; or not typing 'a' vowels in a word, but needing a dead key to signal a combination of consonants. Either method is more phonetic than Itrans as it stands (and much better than the rest), but neither is perfect. The first method, though, is nevertheless pretty similar to the popular IME on *******, Baraha, so hopefully it won't disappoint too much. If it's approved, I'll post a link to the howto here. Thanks for the feedback.
Re: Sanskrit input (Devanagari and transliteration)
Ajit, glad to hear this tweak has been helpful to you. Unfortunately, it seems that a recent update of SCIM is interrupting our input method. I just tweaked another system with this method and found that SCIM/m17n now wouldn't allow me to create a new file, e.g., hi-sanskrit.mim, so I had to overwrite hi-itrans.mim. Ubuntu now takes care of a lot of SCIM configuration, and it's conceivable that that may be disrupting some of our previous settings. You might want to check the environment variables against the SCIM settings specified here. Also, if you're not already overwriting the hi-itrans file, that would be best (as well as to delete hi-sanskrit or whatever other entry you've added to the m17n database). Also, I've found that rebooting rather than just restarting X is often necessary to test a configuration, and of course, it's easier to start with the clean slate of a fresh install, if that's possible. Let us know how the troubleshooting goes and best of luck! Perhaps the SCIM-walas can help us out....
Re: Sanskrit input (Devanagari and transliteration)
@ amerikannu
Do you know something about this issue?
Re: Sanskrit input (Devanagari and transliteration)
That looks to me to be a Unicode font issue since those circles are part of those characters, indicating that they are dependent ones, allowing IMEs to easily combine characters correctly. You might want to consult unicode font sites to inquire about that.
Re: Sanskrit input (Devanagari and transliteration)
Hmm, I can't really help myself with this unicode font issue.
Would you mind to install this program and have a quick look yourself.
Download a .deb package here: http://ichi2.net/anki/download/index.html
A deck: http://www.box.net/shared/cqtdf7ou89
Simply load the deck with this program and type the answer in devanagari.
Thank you.
Re: Sanskrit input (Devanagari and transliteration)
From what I can see, this is not a SCIM/m17n issue, both of which do their job in inputting the Devanagari properly. Your program, Anki, however, is then separating characters without regard for the way in which the script is written -- which is syllabically. The Devanagari unicode fonts are smartly designed to produce syllables from individual characters, which is why you see circles where the Anki program has broken a syllable. "Syllable" for the purposes of the script = either (1) an independent Vowel, or (2) Consonant(s) + a dependent Vowel. Hence, when Anki breaks up conjunct consonants or separates dependent vowels from the preceding consonants on which they depend, it is breaking up the unified syllable recognized by the unicode font. This may be an issue that the Anki developer(s) can configure specifically for Devanagari languages like Hindi/Sanskrit. If I were you, I would email the developer(s) directly or post your issue on the Anki's forums. Best of luck!
Re: Sanskrit input (Devanagari and transliteration)
Thanks for your reply.
The new version of Anki, 0.9.9.7., solved this problem.
Re: Sanskrit input (Devanagari and transliteration)
Hellow I followed the old instructions for 8.04 and they worked, but now I'm using 8.10 and the new instructions, and nothing works.
I see the option "t-hi-sanskrit" on scim, but after choosing it, it stays stuck at "en-ispell", and I can't in fact use any of the options except "eng/keyboard". This is disturbing because I rely on Chinese input for my job, and I can't even use that now.
edit: undid what was done in tutorial, normal functioning scim is back.
Re: Sanskrit input (Devanagari and transliteration)
I exited scim and did this, as described here.
"Install the Devanagari fonts package and the packages for the input method editor (SCIM):
Code:
sudo aptitude install ttf-devanagari-fonts scim-qtimm scim-tables-additional scim-bridge scim-m17n m17n-lib-bin
In scim, under Hindi input, choose hi-itrans."
Without the extra configuration, there's no conflicts and scim-m17n works normally in Intrepid. Just that more awkward system that amerikannu had changed before, but its usable.