# Copyright © 2018 Ikrami Ahmad
#
# Copying and distribution of this file, with or without modification,
# are permitted in any medium without royalty provided the copyright
# notice and this notice are preserved.  This file is offered as-is,
# without any warranty.

display: unicode-without-blank.dis
table:
  locale: ar
  grade: 2
  __assert-match: ar-ar-g2.ctb
flags: {testmode: bothDirections}
tests:
  # 1) first rule: these should be abbreviated everywhere, whether as stand alone word or part of a
  # bigger word.
  
  # stand alone
  - [بعيد, ⠘⠃]
  - [برنامج, ⠸⠃]
  - [تعليم, ⠈⠞]
  - [ثاني, ⠘⠹]
  - [جميع, ⠐⠚]
  - [مكفوف, ⠈⠍]
  - [اللائي, ⠈⠉]
  - [كان, ⠛]
  - [كريم, ⠘⠛]
  - [آفاق, ⠠⠜]
  - [أدنى, ⠠⠕]
  - [إنسان, ⠐⠪]
  # some are contractions for two words
  - - لم يكن
    - ⠐⠑
  - - ما زال
    - ⠈⠏
  - - لا يزال
    - ⠸⠧
  # they can be part of a word
  - - ثقافة
    - ⠈⠹⠡
  - - ومرة
    - ⠺⠐⠍
  - - وتقطعت
    - ⠺⠞⠈⠟⠞
  # a word can have two contractions
  - - الثالث
    - ⠉⠸⠹
  - - كبيرهم
    - ⠘⠅⠻
  # they can be used alongside punctuation marks
  - - "ثانيا:"
    - ⠘⠹⠁⠐⠂
  - - (صغير)
    - ⠦⠘⠯⠴

  # 2) second rule: these should be abbreviated if they come as a single word only, they can't be
  # abbreviated if part of bigger word.
  
  # abbreviated
  - - أو
    - ⠁
  - - قد
    - ⠄
  # not abbreviated
  - - لقد
    - ⠇⠟⠙
  - - كلهم
    - ⠅⠇⠻
  - - ثمرة
    - ⠹⠐⠍

  # 3) third rule: these should be abbreviated if they are a word or at the beginning of a word
  # only, because of their similarity with diacritic symbols.
  
  # abbreviated
  - - لم
    - ⠑
  - - إلى
    - ⠢
  - - عنوان
    - ⠔⠺⠁⠝
  - - معمعة
    - ⠆⠍⠷⠡
  - - حيثما
    - ⠥⠏
  # not abbreviated
  - - ملل
    - ⠍⠇⠇
  - - لعنة
    - ⠇⠷⠝⠡

  # 4) fourth rule: these should be abbreviated if they came as a seperate word and if they came in
  # the middle of words only, because of their similarity to punctuation marks.
  
  # abbreviated
  - - هو
    - ⠦
  - - متى
    - ⠶
  - - مهووس
    - ⠍⠦⠺⠎
  - - هو الذي جاء
    - ⠦ ⠐⠉ ⠚⠁⠄
  # not abbreviated
  - - ومتى
    - ⠺⠍⠞⠕
  - - هوشة
    - ⠓⠺⠩⠡
  - - من هو؟
    - ⠴ ⠓⠺⠦
  - - (متى هذا
    - ⠦⠍⠞⠕ ⠐⠓
  
  # 5) fifth rule: these should be abbreviated if they come as a word or part of a word at the
  # beginning or the middle, because of their similarity with punctuation marks that close
  # sentences.
  
  # abbreviated
  - - من
    - ⠴
  - - في
    - ⠖
  - - منتهى
    - ⠴⠞⠓⠕
  - - مستنقع
    - ⠍⠲⠝⠟⠷
  # not abbreviated
  - - وفي
    - ⠺⠋⠊
  - - لبست
    - ⠇⠃⠎⠞
  - - مدمن
    - ⠍⠙⠍⠝
  - - (ثم من)
    - ⠦⠹ ⠍⠝⠴
  
  # 6) sixth rule: if the following abbriviations come as single words, they should be abbreviated
  # normally. but if they come as a part of bigger word, dots 36 must be placed before the character
  # that represents the abbreviated word.
  
  # as a seperate word
  - - ليس
    - ⠇
  - - مثل
    - ⠍
  - - ذاكر
    - ⠮
  - - شيء نحو الهدف
    - ⠩ ⠝ ⠉⠓⠙⠋
  # due to the use of "lowword", they will not be translated next to punctuations.
  - - قريب جدا
    - ⠘⠟ ⠚
  # as part of a word:
  - - الخير
    - ⠉⠤⠭
  - - فإذا
    - ⠋⠤⠪
  - - مثلما
    - ⠤⠍⠏
  - - ونحوه
    - ⠺⠤⠝⠓
  
  # 7) seventh rule: if the letter that procedes any of these abbreviations consists of dots 1 2 and
  # 3 only, they should not be abbreviated.
  - - علية
    - ⠷⠇⠊⠡
  - - تائبون
    - ⠞⠁⠽⠃⠺⠝
  - - مرتبات
    - ⠍⠗⠞⠃⠁⠞
  
  # 8) eighth rule: those should be abbreviated only at the end of word.
  - - بنات
    - ⠃⠝⠨
  - - قائمون
    - ⠘⠽⠸
  - - قالوا
    - ⠟⠉⠼