I want to convert unicode to monotonic and ascii

Stephen Hughes
Posts: 3323
Joined: February 26th, 2013, 7:12 am

I want to convert unicode to monotonic and ascii

Post by Stephen Hughes »

This is a side question from my Memrise post from a while ago. Viz. http://www.ibiblio.org/bgreek/forum/vie ... ise#p11400.

I am working on a better Memrise list for (Sinophone initially and later Anglophone) beginners that addresses the concerns / suggestions raised by JR in the other thread.

The work-in-progress lists are available at;
Chinese - "通用希腊语高频词汇" http://www.memrise.com/course/136782/
English - "Frequent NT Greek words" http://www.memrise.com/course/135672/
Jonathan Robie wrote:It requires you to get your accents and breathing marks right, which is good practice, but might be overwhelming for beginners.
To be flexible for beginners (and people using school / public computers where they can't install a Greek IME) I am giving the vocab learning lists maximum flexibility for input methods. That has raised a number of issues:

(1) What are the most common keyboard layouts for inputing Greek? Is there a de facto industry standard? My end goal is that I want to let users from a number of typing backgrounds input "transliterated" Greek on computers where the Greek IME is not installed (and they don't have the right to install an IME).
(2) Is there a non-labourious way to go from "ἀγαθός" to "ἀγαθός _ἀγαθός _ἀγαθος _αγαθός _αγαθός _αγαθος"? And from "ἀγαπάω ἀγαπῶ ἀγαπᾶν" to "ἀγαπάω _ἀγαπάω _ἀγαπαω _αγαπάω _αγαπάω _αγαπαω _ἀγαπῶ _ἀγαπω _αγαπῶ _αγαπω _ἀγαπᾶν _ἀγαπαν _αγαπᾶν _αγαπαν"
Jonathan Robie wrote:Near synonyms and synonyms may have English glosses that are indistinguishable. I find it aggravating when it asks for a word that means 'send' and I give it one, but not the one it was thinking about. This is only a problem for a relatively small number of words.
Also I am arranging the glosses to have a reasonable degree of flexibility and a choice of common variants that can be inputed severally or in a number of combinations.

Is there a readily available way to get from "sin, crime" to "sin, crime _sin _crime _crime, sin"?
Γελᾷ δ' ὁ μωρός, κἄν τι μὴ γέλοιον ᾖ
(Menander, Γνῶμαι μονόστιχοι 108)
Jonathan Robie
Posts: 4158
Joined: May 5th, 2011, 5:34 pm
Location: Durham, NC
Contact:

Re: I want to convert unicode to monotonic and ascii

Post by Jonathan Robie »

First off, if you are doing this, may I suggest starting with Dodson's glosses? I can send them to you.

I don't think you need to eliminate the accents and breathing marks, as long as you allow the monotonic or ascii forms as alternate answers, which memrise lets you do (I have played with this a little).

What operating system do you have? These days, you can usually just install the right keyboard and start typing once you know the layout. I'm a Linux guy, but there are others here who can help with just about any operating system. These lists already exist in machine readable form though, so if you are typing it all in from scratch, you may be working harder than you need to.

I think you can use memrise for more than just vocabulary lists, incidentally. The Mexican Spanish course is a good example that actually teaches some language.

This isn't ready for prime time at all, but here's an experiment I did with simple noun phrases:

http://www.memrise.com/list/136539/goetchius/

This was for my own use, so breathing marks and accents are required ...

Ideally, I would like to have an authoring system that combines lessons with spaced repetition, imagine a combination of Rosetta Stone or LiveMocha with Anki or Memrise. Can anyone tell me if Moodle would fit the bill?
ἐξίσταντο δὲ πάντες καὶ διηποροῦντο, ἄλλος πρὸς ἄλλον λέγοντες, τί θέλει τοῦτο εἶναι;
http://jonathanrobie.biblicalhumanities.org/
Stephen Hughes
Posts: 3323
Joined: February 26th, 2013, 7:12 am

Re: I want to convert unicode to monotonic and ascii

Post by Stephen Hughes »

Jonathan Robie wrote:if you are doing this, may I suggest starting with Dodson's glosses?
I can at least include Dodson's glosses as alternatives - there are sure to be people who know them. Personally I have been working from the same copy of Holly (1978) since I started learning New Testament Greek. I could add the other glosses from various books later, but for now to give the project managability and an end, I will just stick with Z&G's list for now. For the Chinese, I have a number of lists (all very different), and my aim is useablity rather than exhaustiveness - just the list that I have been using with my students (90% of which I can recite for / with them).

My office is XP SP3 (Chinese Simplified) across the board. I'm working to accomodate people who use public / office computers who do not have the right to install a Greek IME or GreekKeys.
Jonathan Robie wrote:These lists already exist in machine readable form though, so if you are typing it all in from scratch, you may be working harder than you need to.
Do you mean lists like the
Stephen Hughes wrote:(2) Is there a non-labourious way to go from "ἀγαθός" to "ἀγαθός _ἀγαθός _ἀγαθος _αγαθός _αγαθός _αγαθος"? And from "ἀγαπάω ἀγαπῶ ἀγαπᾶν" to "ἀγαπάω _ἀγαπάω _ἀγαπαω _αγαπάω _αγαπάω _αγαπαω _ἀγαπῶ _ἀγαπω _αγαπῶ _αγαπω _ἀγαπᾶν _ἀγαπαν _αγαπᾶν _αγαπαν"
that I want to include in the variants?
Jonathan Robie wrote:I don't think you need to eliminate the accents and breathing marks, as long as you allow the monotonic or ascii forms as alternate answers, which memrise lets you d
That's the plan. Entering the words and glosses is easy, putting alternatives into the system is laborious!!!
Γελᾷ δ' ὁ μωρός, κἄν τι μὴ γέλοιον ᾖ
(Menander, Γνῶμαι μονόστιχοι 108)
NathanSmith
Posts: 62
Joined: June 10th, 2011, 12:38 am
Location: Portland, OR, USA
Contact:

Re: I want to convert unicode to monotonic and ascii

Post by NathanSmith »

The non-laborious way would of course to have a computer do that for you. :-)

The task of converting a polytonic Unicode word into all the permutations down to monotonic Unicode can be done by a script. Here is an example of removing all diacritics using Python. It would require a bit more work to find and return all possible combinations as you want in your example, but it would probably be quicker than doing so by hand.

James Tauber has written a script which can be used to convert Betacode to Unicode, and that could be reversed if desired to convert words to ASCII (but I must admit, I cringe at the thought ;) ). There is also a Unidecode Python package which claims to do simple Unicode->ASCII conversions.

However, rather than accepting all of those permutations, it might be best to accept ASCII or monotonic Unicode as answers only.
Stephen Hughes
Posts: 3323
Joined: February 26th, 2013, 7:12 am

Re: I want to convert unicode to monotonic and ascii

Post by Stephen Hughes »

NathanSmith wrote:The non-laborious way would of course to have a computer do that for you. :-)

The task of converting a polytonic Unicode word into all the permutations down to monotonic Unicode can be done by a script. Here is an example of removing all diacritics using Python. It would require a bit more work to find and return all possible combinations as you want in your example, but it would probably be quicker than doing so by hand.

James Tauber has written a script which can be used to convert Betacode to Unicode, and that could be reversed if desired to convert words to ASCII (but I must admit, I cringe at the thought ;) ). There is also a Unidecode Python package which claims to do simple Unicode->ASCII conversions.

However, rather than accepting all of those permutations, it might be best to accept ASCII or monotonic Unicode as answers only.
I've had a look at those and they look useful if I could use Python. I need something more for idiot-level computer users, like http://textmechanic.com/

As far as I know Python is a SSP language, so if you are talking about a computer doing it for me as the memrise programme is running then that looks like a conversion step that could ideally happen between the user input and the memrise checking the words, i.e. polytonic, monotonic and English letters would all be converted to a basic Greek alphabet from, that could then be checked by memrise. But it is a generic sort of site, and list / course makers don't have such a high degree of dynamic control over input processing, but only over what discrete froms will be accepted after input. The many forms that I have listed are all possibly valid and expected (beginner) user inputs for Greek. [The valid Chinese inputs that I have allowed for are simplified Chinese character, pinyin with tones marked and pinyin without tones marked, which will suit a variety of users in a variety of situations].
Γελᾷ δ' ὁ μωρός, κἄν τι μὴ γέλοιον ᾖ
(Menander, Γνῶμαι μονόστιχοι 108)
Louis L Sorenson
Posts: 711
Joined: May 5th, 2011, 9:21 pm
Location: Burnsville, MN, USA
Contact:

Re: I want to convert unicode to monotonic and ascii

Post by Louis L Sorenson »

Jonathan Robie
Posts: 4158
Joined: May 5th, 2011, 5:34 pm
Location: Durham, NC
Contact:

Re: I want to convert unicode to monotonic and ascii

Post by Jonathan Robie »

For my money ... I would probably support Greek with or without accents, but not with incorrect accents. If someone types the accents in, they probably want to get it right.

Looks like there is a bulk uploader for memrise:

http://www.memrise.com/thread/1300033/

One nice aspect of that - if you have the data on your hard disk, you own it. If it's only on memrise, you can only use it if you manage to get it onto your hard disk.
ἐξίσταντο δὲ πάντες καὶ διηποροῦντο, ἄλλος πρὸς ἄλλον λέγοντες, τί θέλει τοῦτο εἶναι;
http://jonathanrobie.biblicalhumanities.org/
Stephen Hughes
Posts: 3323
Joined: February 26th, 2013, 7:12 am

Re: I want to convert unicode to monotonic and ascii

Post by Stephen Hughes »

Jonathan Robie wrote:For my money ... I would probably support Greek with or without accents, but not with incorrect accents. If someone types the accents in, they probably want to get it right.
The issues there are not so easy to arrange into a stand-back-and-let-it-run learning system. The micromanagement and step-by-step correction that we have in a classroom is not there. The consequence of that lack of control and given that the memrise system doesn't allow me to pop in a error contexualised comment or explanation when a mistake is made means that in some cases even when a student is chopping wood with the blunt side of the axe, it could still be a valid path to a warm winter's evening.

Considering the issue of error v. inaccuracy, your suggestion, not only relates to accents, but also raises questions about breathings... According to my earlier training, an omitted breathing was a mistake, but an omitted accent was a pardonable offence. But if I use the Living Koine pronunciation as the spoken form of the language in the memrise list, then the spiritus asper and spiritus lenis are not effectively differentiated in pronunciation are not so important as the accents, which determine word-stress and are therefore necessary for effective pronunciation.

Another question of correctness or incorrectness is the oxytone / barytone distinction on the final syllable. A beginner who is not familiar with the rule differentiating the two, may have seen a form in writing with the barytone and copy it, but should that be classed as wrong? I don't think it is "wrong" (in the context of the language) because it is a possible form of the word - and in the post classical period the stress accent is the same no matter what the accent. But it could be "out of place" in a vocabulary learning context.

While we could say that an oxytone for a perispomenon is in the bigger scheme of things "wrong", the oxytone is never-the-less a valid accent mark, so it is 50% right (function - right, form - wrong). For Modern Greek, from the Koine period up to the 1970's it was always necessary to make the correct distinction between the 3 accents, but now of course in Modern Greek that is not necessary - now only the position of the stress is important, and a single undifferentiated stress marker is used - often a simple verticle stroke.

If later, cratic froms are listed, could the coronis be allowed to be omitted? I think so. But should it also be allowed to be written as an apostrophe? Perhaps. Iota subscript will probably want to be accepted as Iota adscript, because there will be students who are not working with a keyboad layout that can handle the subscript. The diaresis is another matter. It seems that it is universally necessary for correct pronunciation, but (again) not everyone (in non-first world countries or those using it on smart phones) will be able to type it on all computers at all times.

I think all students need to learn the alphabet near the beginning of their study, but perhaps not all teachers want to stress learning the accents so clearly. And some will not want to give much attention to breathings either.

In my understand that having the wrong breathing is wrong (no bones about it), but the evidence from the ellided prepositions in the papyri and spelling in the Coptic dialects suggest that there was variability in pronunciation in some places / at some periods. But the NT text that we read now has a standard for writing breathings.

Another issue (out of place in the context of this thread but relevent to what I am doing) is first person present indicative active or infinitive? Most of my learning was with the later, but the more I teach and the more I think about it, the infinitive is really the basic (almost atomic) unit of the language (and possibly the accusative not the nominative). Do any courses teach from the infinitive up for verbal conjugational structures (or from the accusative out for nominal decensions)?

Does anyone have a thought about what they consider the lines for rightness and wrongness start and end for beginners on the road from inaccuracy to accuracy?
Γελᾷ δ' ὁ μωρός, κἄν τι μὴ γέλοιον ᾖ
(Menander, Γνῶμαι μονόστιχοι 108)
Shirley Rollinson
Posts: 415
Joined: June 4th, 2011, 6:19 pm
Location: New Mexico
Contact:

Re: I want to convert unicode to monotonic and ascii

Post by Shirley Rollinson »

Stephen Hughes wrote:
Jonathan Robie wrote:For my money ... I would probably support Greek with or without accents, but not with incorrect accents. If someone types the accents in, they probably want to get it right.
- - -
- - -
I think all students need to learn the alphabet near the beginning of their study, but perhaps not all teachers want to stress learning the accents so clearly. And some will not want to give much attention to breathings either.

In my understand that having the wrong breathing is wrong (no bones about it), but the evidence from the ellided prepositions in the papyri and spelling in the Coptic dialects suggest that there was variability in pronunciation in some places / at some periods. But the NT text that we read now has a standard for writing breathings.

Another issue (out of place in the context of this thread but relevent to what I am doing) is first person present indicative active or infinitive? Most of my learning was with the later, but the more I teach and the more I think about it, the infinitive is really the basic (almost atomic) unit of the language (and possibly the accusative not the nominative). Do any courses teach from the infinitive up for verbal conjugational structures (or from the accusative out for nominal decensions)?

Does anyone have a thought about what they consider the lines for rightness and wrongness start and end for beginners on the road from inaccuracy to accuracy?
I consider the rough breathing to be equivalent to the letter "h", so is part of the spelling of a word, and needs to be learned from the beginning.
However, starting verbs from the Present Infinitive Active rather than the First Person Present Indicative might cause a bit of confusion when dealing with contract verbs.
For Declensions - if I did not start with the Nominative (which I do) - I would choose to start with the Genitive rather than the Accusative, because of what happens with the Third Declension.
Learning Nominative, then Accusative, then Genitive and Dative means that one can work up the complexity of a sentence from "Peter speaks.", "Peter speaks and Paul writes.", "Peter writes the words.", "Peter writes the words in the book", "Peter speaks the words and Paul writes them in his book." etc. which probably fits best with the way students learn to produce sentences.
Louis L Sorenson
Posts: 711
Joined: May 5th, 2011, 9:21 pm
Location: Burnsville, MN, USA
Contact:

Re: I want to convert unicode to monotonic and ascii

Post by Louis L Sorenson »

Stephen,

Has your question been answered? For polytonic to monotonic, you can use the online http://www.translatum.gr/converter/p2m/ ... otonic.php. It would not be too hard to have the hopperizer at Katabiblon add that transformation. You mentioned that there are some limitations on online software availability in your area. Are you looking for a downloadable application or an online tool?
Post Reply

Return to “Unicode and Fonts”