Announcing: Persian Input Methods

Dear Fellow Persian Speaker,

Towards improving capabilities for writing in Persian, we want to introduce to you a new input method that permits intuitive use of a standard latin keyboard for transliteration. Use of this composition input method can be learned in a few minutes as an alternative to use of the Iranian (isiri-9147) standard keyboard.

Use and development of multilingual software in Persian has 3 main dimensions:

1. Character Set, Fonts and Shaping
2. Bi-directionality
3. Input Methods

Some aspects of these are inherently more static. Our character set has converged on isiri-6219. Our shaping software has converged on libm17n. And bidirectionality has converged on UAX #9.

Other aspects of these are inherently more dynamic. It is highly desirable for us to have many good fonts and many good input methods.

In a document titled:

Persian Input Methods
For Emacs And More Broadly Speaking
شیوه‌هایِ درج به فارسی‌
http://www.persoarabic.org/PLPC/120036

We take on the topic of design of input methods for Persian/Farsi.

Broadly speaking, three types of inputs method are applicable to Persian/Farsi.

Traditional Simple Keyboard Input Methods:

This is the traditional one-to-one mapping of keys on a computer keyboard to Persian letters. ISIRI-9147 has become the convergence point for this.

Composite Keyboard Input Methods:

This is a more powerful method which converts sequences of characters into one letter. For example “kh” becomes خ. The obvious phonetic one-to-one mapping of keys to characters (e.g., “p” to پ, “b” to ب, “f” to ف) combined with the multi-character composite transliterations, provides for an intuitive composite mapping that is precise and that can be learned very quickly. We propose that farsi-transliterate-banan become a point of convergence for composite Persian keyboard mappings.

Pinglish Phonetic Word Input Methods:

Pinglish/Finglish is an informal and loose transliteration of Persian for human-to-human communication. Pinglish is word oriented. Various Pinglish style web services such as behnevis, permit producing Persian text using Pinglish. We don’t consider this input method suitable for extensive formal writing in Persian. Pinglish’s strength comes from its loose and informal nature and pinglish does not lend itself to convergence and standardization.

We have chosen Emacs as our multilingual editor-centered user experience platform of choice and we have implemented two Persian input methods for Emacs.

The National Iranian Keyboard (isiri-9147)

Presently the main Persian input method is the national Iranian keyboard. ISIRI-9147 defines the layout of Iran’s Persian keyboard. Its full title is:

فنّاوریِ اطلاعات - چیدمان حروف و علائم فارسی بر صفحه کلید رایانه
ستاندارد ملی ایران ۹۱۴۷ − چاپ اول
Institute of Standards and Industrial Research of Iran
Information Technology – Layout of Persian Letters and Symbols on Computer Keyboards

The ISIRI-9147, is fully implemented as an input method for Emacs.

But, generally speaking, the ISIRI-9147 Persian keyboard is not well suited to Iranian expatriates living in the West. And others who are familiar with qwerty, and who don’t wish to learn isiri-9147.

farsi-transliterate-banan: A Multi-Character Composition Persian Input Method

Persian-speaking expatriates are usually already completely familiar and accustomed to the standard qwerty keyboard, and they don’t want to have to learn and adapt to ISIRI-9147. Rather, they expect software to adapt to them.

This is what the farsi-transliterate-banan – “Banan Multi-Character (Reverse) Transliteration Persian Input Method” – accomplishes. This input method addresses the needs of a user who:

• Can write in Farsi (not just speak it).
• Is familiar with and accustomed to the qwerty Latin keyboard.
• Is unfamiliar with ISIRI-9147 and does not wish to learn it.
• Writes and otherwise communicates in mixed Globish/Persian, not pure Persian.
• Is intuitively familiar with the transliteration of Farsi/Persian into Latin based on two-letter phonetic mapping to Persian characters. (For example: gh ق – kh خ – sh ش – ch چ – zh ژ ).

The transliteration keyboard is intuitive in design, so that the mappings are natural and easy to remember for a Persian writer. It provides equivalent capability to farsi-isiri-9147, allowing input of all characters enumerated in ISIRI-6219.

farsi-transliterate-banan is phonetically oriented. But it is very different from Pinglish. Pinglish is word-oriented, where you sound out the word using Latin letters, including the vowels. farsi-transliterate-banan is letter-oriented, where you type the Latin letter(s) closest to the Persian letter, and usually omit vowels.

Getting Started With Persian Emacs

Summary information for obtaining Emacs, installing Emacs and getting you started with these Persian input methods is available at: http://www.persoarabic.org/emacs. Complete documentation for these input methods is included in PLPC-120036. Emacs is halaal software and its copying and usage is not restricted and is free of charge.

A sample farsi editing session demonstrating intuitiveness of use of farsi-transliterate-banan is available at:
http://www.persoarabic.org/emacs/#SampleFarsiEditingSession

Broader Scope Of These Emacs Persian Input Methods

In the “Persian Input Methods” document we do several things. First we fully document two input methods for Emacs. Based on that, we next point out various improvements to relevant specifications, namely: isiri-6219 and isiri-9147. Third we suggest and encourage adoption of Emacs’s multi-character reverse transliteration input method (farsi-transliterate-banan) as a convergence point for Persian composition input methods in other environments such as Gnome.

As you use these Persian input methods, we are interested in your feedback. You can send us your comments, criticisms and corrections via the URL http://mohsen.1.banan.byname.net/contact, or by email to feedback@ mohsen.1.banan.byname .net.

—

We hope you find these tools helpful.

At your service,