QurASCII-Transliteration - Development History

The idea for the QurAscii transliteration came up in course of the transformation of the Tanzil Quran (which was published on the internet in 2007) into a relational Access database during Ramadan 2008.

The goal was, to only use ASCII characters ('Basic Latin' Block '0000' to '007F' of Unicode), in order to create a fully reversible transliteration of the holy quran. As the chracters '0000' trough '0020' and '007F' are used for control characters and '0030' through '0039' for numbers, 84 ASCII characters remained for this task.

In a first approach the Buckwalter Transliteration was used. But as the Buckwalter transliteration had some defizits in relation to modern html/xml representation of arabic text and a lot of special characters/diacrits, which are used in the holy Quran, were missing, a slightly different approach had to be selected.

Basic idea was to use only ASCII consonants (and no special characters) to transliterate the arabic consonants (except alif which is treated as a carrier and vocal extension). For diacritic vowels also the related ASCII vowels were used. The remaining vowels were distributed to 'Alif Madda', 'Alif Wasla' und 'Alif Maksura'. For plain 'Alif' the exclamation mark was used as it is of similar shape.

The resulting deviations from the Buckwalter Transliteration are listed in the following table:

As manual transliteration in a database of course was not feasible, I decided to teach myself the open source programming language Perl which seemed to be the best choice for the task of textual analysis and conversion.

After longer testing and optimizing the handling of the Quran data, through object oriented programming, in 2011 I finally was able to launch the first version of my 'Quran-Explorer' comprising a full cross-reference of used words and a rudimentary search function.

In order to optimize the search, the following modification was done to the original transliteration:

Additionally also the 'Combining Diacritical Marks' (Unicode range '0653' to '0655') and the 'Quranic Annotation Sings' (Unicode range '06D6' to '06ED') were assigned as per following table:

Arabic	QurASCII	Explanation
ٓ	~	same shape
ٔ	^
ٕ	=
ۖ	<	pause mark => used HTML reserved character
ۗ	>	pause mark => used HTML reserved character
ۘ	;	pause mark => used special character
ۙ	&	pause mark => used HTML reserved character
ۚ	'	pause mark => used HTML reserved character
ۛ	"	pause mark => used HTML reserved character
ۜ	J
۝	.	verse mark => used special character
۞	#	verse mark => used special character
۟	R
۠	Q
ۡ	*
ۢ	M
ۣ	K
ۤ	?
ۥ	V
ۦ	X
ۧ	+
ۨ	N
۩	P	prostration mark
۪	B
۫	C
۬	F
ۭ	L

In 2013 the fulltext seach for the Quran was optimized. During that time I identified problems with the letters '?' and '^' which are metacharacters in Perl's regular expressions. To enable searching for all characters which are used in Tanzil Quran the following additional modifications had to be done:

In the following the resulting overall QurASCII-table for the assignment of all ASCII characters '0000' through '007F' to the related Arabic characters is given: