Persian

Updated 4 October, 2021

This page brings together basic information about the Arabic script and its use for the Iranian Persian language. It aims to provide a brief, descriptive summary of the modern, printed orthography and typographic features, and to advise how to write Persian using Unicode.

Phonetic transcriptions on this page should be treated as an approximate guide, only. Many are more phonemic than phonetic, and there may be variations depending on the source of the transcription.

More about using this page
Related pages.
Other script summaries.

Sample (Persian)

Select part of this sample text to show a list of characters, with links to more details.
Change size:   28px

تمام افراد بشر آزاد بدنیا میایند و از لحاظ حیثیت و حقوق با هم برابرند. همه دارای عقل و وجدان میباشند و باید نسبت بیکدیگر با روح برادری رفتار کنند.

هر کس میتواند بدون هیچگونه تمایز مخصوصا از حیث نژاد، رنگ، جنس، زبان، مذهب، عقیدهٔ سیاسی یا هر عقیده دیگر و همچنین ملیت، وضع اجتماعی، ثروت، ولادت یا هر موقعیت دیگر، از تمام حقوق و کلیهٔ آزادی‌هائیکه در اعلامیه ذکر حاضر شده است، بهره‌مند گردد.

Usage & history

The Persian alphabet is a variant of the Arabic script, adapted to meet the needs of the different sounds in the Persian language. It is used to write the Persian language in Iran and Afghanistan.

الفبای فارسی‎ ælefbɒːje fɒːrsiː Farsi alphabet

The former Persian Pahlavi scripts were replaced by the Arabic script, but not until some time after the Arab conquest of Persia. It was introduced by the Saffarid and Samanid dynasties in 9th-century Greater Khorasan.

Basic features

Persian uses the Arabic script, with extensions to covers its wider repertoire of sounds. The Arabic script is an abjad. This means that in normal use the script represents only consonant and long vowel sounds. See the table to the right for a brief overview of features for the modern Persian orthography.

Persian text runs right-to-left in horizontal lines, but numbers and embedded Latin text are read left-to-right.

It is sometimes written using the nasta'liq style of Arabic writing. Glyphs are more drawn out, and the baseline tends to be sloping from word to word.

The script is cursive, and some basic letter shapes change radically, depending on what they join to.

There is no case distinction.

Words are separated by spaces.

Modern Persian uses 36 basic letters to write consonants, some of which are hangovers used to spell words loaned Arabic. For example, there are 2 ways to write t, 3 ways to write s, and 4 ways to write z. Although it is not always easy to guess the vowel sounds in a word, the consonants are largely reliable phonetically. There is mostly a one-to-one correspondance between letters and sounds.

Although the vowel diacritics are not normally used, Persian has 3 diacritic code points that can be used to indicate the 'short' vowels, if needed. Additional diacritics are available to indicate the absence of a vowel in consonant clusters, and gemination in vowelled text. Long vowels, and word initial and final short vowels are represented using one of 4 consonant letters, although there can occasionally be some ambiguity as to which sounds they represent.

To indicate vowel cluster boundaries and the ezafe conjunction, Persian uses a combining hamza above carrier letters. (A standalone hamza is only used occasionally.) The choice between precomposed and decomposed realisations of characters used for these features has a few complications.

A mandatory ligature is used for combinations of lam + alif.

Persian uses native digits, though the code points are different from those used for the Arabic language, and Arabic code points are used for several of the more common punctuation marks.

Joining forms

Because the Arabic script is 'cursive' (ie. joined-up) writing, letters tend to have different shapes depending on whether or not they join with adjacent letters (see cursive). However, several characters have no left-joining form, and this has an effect on the following letter shape.

Here and in the literature there is sometimes ambiguity around the use of the terms 'initial', 'medial', 'final', and 'isolated' – in particular, they may be confused with word-initial and word-final. Here we describe the usage in this document.

'Initial' generally refers to glyphs that are only joined to the left. These forms tend to occur word-initially but also occur word-medially following a letter that doesn't join to the left. We try to be clear when we mean word-initial. Word-initial vowels are always attached to or preceded by ا [U+0627 ARABIC LETTER ALEF], eg. اَ or ای‍ـ, and those combinations may be listed in tables as (word-)initial forms.

'Medial' forms are usually those that join on both sides to an adjacent letter.

'Final' forms are those that join only to the right. These forms often occur at the end of a word, but not after a letter that only joins to the right. Especially in Persian, these forms can also be found in word-medial position, usually at morphological boundaries, eg. گرگ‌ها . The left joining behaviour is prevented using [U+200C ZERO WIDTH NON-JOINER].

'Isolated' forms occur when a letter is used on its own, or when a letter in a word is unjoined. Unjoined forms occur after a letter that doesn't join to the left, most commonly in word-final position, but also in sequences of characters that only join to the right, eg. the following word has two unjoined forms at the end: بسیاری. When a vowel is written on its own or word-initially it is always attached to, represented by, or preceded by ا [U+0627 ARABIC LETTER ALEF], eg. this helps to distinguish between ای i and ی j.

Character index

Letters

Show

Consonants & matres lectionis

ء␣آ␣أ␣ؤ␣ا␣ب␣ت␣ث␣ج␣ح␣خ␣د␣ذ␣ر␣ز␣س␣ش␣ص␣ض␣ط␣ظ␣ع␣غ␣ف␣ق␣ل␣م␣ن␣ه␣و␣پ␣چ␣ژ␣ک␣گ␣ی
ة

Modifier letter

ـ

Not used for Persian

ئ␣ك␣ي␣ۀ␣ہ␣ۂ␣ۃ␣ە␣ࢨ

Combining marks

Show

Vowels

ً␣ٌ␣ٍ␣َ␣ُ␣ِ␣ٓ

Other

ٔ
ّ␣ْ

Numbers

Show
۰␣۱␣۲␣۳␣۴␣۵␣۶␣۷␣۸␣۹

Punctuation

Show

General punctuation

،␣؛␣؟␣٪

ASCII

!␣.␣:

Quotes

«␣»

Opening/closing

(␣)

Other

Show
‌␣‍␣⁧␣‫␣⁦␣‪␣⁨␣⁩␣‬␣‏␣‎
Character lists show:

Phonology

These are sounds of the Iranian Persian language, and more particularly tend to reflect the Tehran dialect.

Click on the sounds to reveal locations in this document where they are mentioned.

Phones in a lighter colour are non-native or allophones. Source Wikipedia.

Vowel sounds

There are 6 vowel sounds, though there are also allophonic variants. The basic phonemes are as follows:

i i u u e e o o æ æ ɒ

i, u, and ɒ are regarded as long vowels. e, o, and æ are short.

Consonant sounds

labial dental alveolar post-
alveolar
palatal velar uvular glottal
stop p b t d       k ɡ ɢ ʔ
affricate       t͡ʃ d͡ʒ        
fricative f v   s z ʃ ʒ   x ɣ   h
nasal m   n        
approximant     l   j    
trill/flap     r    

In Iranian Persian ɣ and q have merged into ɣ~ɢ, as a voiced velar fricative ɣ when positioned intervocalically and unstressed, and as a voiced uvular stop ɢ otherwise.wpl,#Phonology

Vowels

Click on the characters in the lists for detailed information. For a mapping of sounds to graphemes see vowel_mappings.

Matres lectionis

ا␣آ␣ی␣و␣ه

Persian is normally written without vowel diacritics. As a general rule, only long vowels are represented by matres lectionis, except that all vowels in word-initial position are written with or preceded by ا [U+0627 ARABIC LETTER ALEF]. Another exception is a final e, which is written with ه [U+0647 ARABIC LETTER HEH]. (Final a and o are also represented by a consonant, but are rarely found.)

initial   medial   final  
ا o e æ ‍ا ɑː ‍ا ɑː
آ ɑː
ایـ ـیـ ـی
او o ‍و ‍و o ow
        ـه e
Letters mapped to vowel sounds. (The table should be read right-to-left.)

The vowels e oand æ are not marked in medial position, and with the exception of e generally do not occur in word-final position. However, one common word in the Tehran dialect of Persian that does end with æ is نه no

آ [U+0622 ARABIC LETTER ALEF WITH MADDA ABOVE] is included in the table because it is normally represented by a single, precomposed character.

See also vowel_mappings.

Vowel diacritics

In situations where it is necessary to unambiguously indicate the underlying vowel sounds, the following diacritics can be added to base letters.

ِ␣ُ␣َ

The following table summarises the main vowel to character assigments. Note that some sounds are distinguished in vowelled text by an absence of diacritics. More information can be found by clicking on the characters above, or in the section vowel_mappings.

Each table cell shows word-initial, word-medial, and word-final forms from right to left. Click/tap on items to see a list of the components for that cell.

ای‍‍ی‍‍ی او‍و‍و
e o
اِ◌ِ◌ِ‍ه اُ◌ُ◌ُو
  æ  
  اَ◌َ◌َ‍ه  
ɑː
آ‍ا‍ا

Initial u is rare, as are final æ and o.

Final ه [U+0647 ARABIC LETTER HEH] and و [U+0648 ARABIC LETTER WAW] are only treated as vowels if they follow a consonant sound. If they follow a vowel sound, they revert to their normal consonant value. Unfortunately, since short vowels are only rarely shown, it can sometimes be difficult to tell from the written text whether to pronounce these letters as vowel or consonant.

Other diacritics

ً␣ٌ␣ٍ␣ٓ

The doubled vowel diacritic, ◌ً [U+064B ARABIC FATHATAN] is used at the ends of certain Arabic-derived adverbs in vowelled text. It is usually written over an alif, although the vowel sound is short. Examples, لزوماً اصلاً

Other doubled vowel diacritics, ◌ٌ [U+064C ARABIC DAMMATAN​] and ◌ٍ [U+064D ARABIC KASRATAN​] are not used in Persian, but are taught to support education in the Qur'anwpa,#Tanvin_(nunation).

ٓ [U+0653 ARABIC MADDAH ABOVE] is only found in decomposed text, and is associated only with alef. See آ [U+0622 ARABIC LETTER ALEF WITH MADDA ABOVE].

Standalone vowels

Word-medial Persian vowels are sometimes pronounced without an intervening consonant, but they are normally written as if they are separated by a glottal stop. The glottal stop is written using a hamza and its carrier (see hamza).

Word-initial standalone vowels are always attached to or preceded by ا [U+0627 ARABIC LETTER ALEF], eg. اَ or ای‍ـ.

Ezafe

ِ␣ٔ␣ی

Ezāfe is a grammatical particle used to link words together. It is used between adjectives and nouns. Between a sequence of nouns it is similar in use to the word 'de' in French. It is pronounced ɛ~e or (after a vowel) .

Ezafe can be written in three ways in vowelled text, however in normal Persian text only the third is visible, because the diacritics are omitted.pm,41

  ِ   [U+0650 ARABIC KASRA] After a word-final consonant. گازِ طبیعی
  ٔ   [U+0654 ARABIC HAMZA ABOVE] After a word that ends with a short vowel (usually in the combination ‍هٔ [U+0647 ARABIC LETTER HEH + U+0654 ARABIC HAMZA ABOVE]). نشانهٔ سجاوندی
یِ [U+06CC ARABIC LETTER FARSI YEH + U+0650 ARABIC KASRA] After a word that ends with a long vowel. نشانه‌هایِ سجاوندی

Vowel to script mapping

The following tables show how vowel sounds map to common characters or sequences of characters.

Persian follows Arabic in using diacritics to express short vowel sounds, but also rarely uses them in normal text. Vowel diacritics are shown here, but can be ignored to reflect normal usage in Persian text

The table shows initial (i), medial (m), and final (f) forms. Click on the character names for examples of usage.

e
i

ا [U+0627 ARABIC LETTER ALEF] in normal text. 

اِ [U+0627 ARABIC LETTER ALEF + U+0650 ARABIC KASRA] in vowelled text.

 
m

Not shown in normal text.

◌ِ [U+0650 ARABIC KASRA] in vowelled text.

 
f

ه [U+0647 ARABIC LETTER HEH] in normal text

◌ِ‍ه [U+0650 ARABIC KASRA + U+0647 ARABIC LETTER HEH] in vowelled text.

o
i

ا [U+0627 ARABIC LETTER ALEF] in normal text. 

اُ [U+0627 ARABIC LETTER ALEF + U+064F ARABIC DAMMA] in vowelled text.

 
m

Not shown in normal text.

◌ُ [U+064F ARABIC DAMMA]

 
f

و [U+0648 ARABIC LETTER WAW] in normal text.

◌ُو [U+064F ARABIC DAMMA + U+0648 ARABIC LETTER WAW] in vowelled text.

 
m

Not shown in normal text.

◌َ [U+064E ARABIC FATHA]

 
f

ه [U+0647 ARABIC LETTER HEH] in normal text (not common)

◌َ‍ه [U+064E ARABIC FATHA + U+0647 ARABIC LETTER HEH] in vowelled text.

Consonants

Click on the characters in the lists for detailed information. For a mapping of sounds to graphemes see consonant_mappings.

Basic letters

Stops

پ␣ب␣ت␣ط␣د␣ک␣گ␣ق␣ء

Affricates

چ␣ج

Fricatives

ف␣و␣س␣ث␣ص␣ز␣ذ␣ض␣ظ␣ش␣ژ␣خ␣غ␣ه␣ح␣ع

Nasals

م␣ن

Liquids

ر␣ل␣ی

و [U+0648 ARABIC LETTER WAW] and ی [U+06CC ARABIC LETTER FARSI YEH] represent both consonants and vowels. See mapToVowels.

Due to the influence of Arabic spelling in loan words, Persian has 2 letters for t, 3 letters for s, 4 letters for z, and 2 letters for h. The most common letter for s is س [U+0633 ARABIC LETTER SEEN], and for z is ز [U+0632 ARABIC LETTER ZAIN].

Glottal stop

In Persian, a glottal stop is commonly written using ع [U+0639 ARABIC LETTER AIN], eg. معنوی عربی شیعه

In other places, Persian uses a hamza (see hamza).

Hamza

أ␣ء␣ٔ␣ؤ

The hamza (called hamze in Persian) is used to represent a glottal stop, however rather than being written as a single letter it is normally written as a combination of a diacritic and a base letter.

In most cases, that is the combination ـیٔـ [U+06CC ARABIC LETTER FARSI YEH + U+0654 ARABIC HAMZA ABOVE], eg. مسئول

However, if the glottal stop is preceded by the sound o, the combination is ـؤـ [U+0624 ARABIC LETTER WAW WITH HAMZA ABOVE], eg. مؤمن

Although it may not be pronounced, the hamza and its carrier also appears between vowels. For example, فائده زئوس مؤثر سؤال سوئیس

In the examples above you can see that the hamza+carrier provides a place to put a vowel diacritic for short vowels in vowelled text. Between two long vowels it represents a nominal glottal stop.

In the i sound of the indefinite ending, the hamza may also be used, or alternatively the ending may double the YEH after a long vowel, eg. compare: پایٔی پایی مویٔی مویی

On more rare occasions the hamza may appear over ا [U+0627 ARABIC LETTER ALEF], eg. رأی تأکید

It may also appear as an isolated form, ء [U+0621 ARABIC LETTER HAMZA], eg. علاءالدین امضاء

The hamza is also used over short, word-final vowels for ezafe.

A number of precomposed combinations of base letter and hamza are encoded in Unicode. Many of these decompose and recompose under normalisation as canonical alternatives, but a few do not and need to be treated with care. For information about which precomposed characters are used or not used here see hamza_choices.

Consonant clusters

ْ

The absence of a vowel sound can be indicated with the diacritic  ْ [U+0652 ARABIC SUKUN], although this diacritic is not normally shown in text, eg. نَسْتَعْلیق

Consonant lengthening & gemination

ّ

In vowelled text, which is rare, geminated consonants are shown using the diacritic  ّ [U+0651 ARABIC SHADDA​], eg. تپّه اوّلی

Consonant sounds to characters

The following maps Persian consonant sounds to common graphemes.

Encoding choices

In the Persian orthography different sequences of Unicode characters may produce the same visual result. Here we look at those, and make notes on usage.

Hamza & precomposed characters

Unicode support for the various uses of the hamza are complicated.u,384 For notes on the usage of the hamza in Persian, see hamza and ezafe.

Canonically-equivalent alternatives

A number of combinations with the hamza diacritic can be represented as either a precomposed character or a decomposed sequence, where the parts are separated in Unicode Normalisation Form D (NFD) and recomposed in Unicode Normalisation Form C (NFC), so both approaches are canonically equivalent. These include the following:

Precomposed Decomposed
أ [U+0623 ARABIC LETTER ALEF WITH HAMZA ABOVE] أ [U+0627 ARABIC LETTER ALEF + U+0654 ARABIC HAMZA ABOVE]
آ [U+0622 ARABIC LETTER ALEF WITH MADDA ABOVE] آ [U+0627 ARABIC LETTER ALEF + U+0653 ARABIC MADDAH ABOVE]
ؤ [U+0624 ARABIC LETTER WAW WITH HAMZA ABOVE] ؤ [U+0648 ARABIC LETTER WAW + U+0654 ARABIC HAMZA ABOVE]

The single code point per vowel-sign is the form preferred by the Unicode Standard and the form in common use for Persian, but either could be used.

Alternatives that are not canonically equivalent

These cases involve precomposed characters that look identical to the sequences used in Persian, however there is a catch because the precomposed characters have canonical decompositions to letters that are not used in Persian.

Recommended Not recommended  
هٔ [U+0647 ARABIC LETTER HEH + U+0654 ARABIC HAMZA ABOVE]  ۀ [U+06C0 ARABIC LETTER HEH WITH YEH ABOVE] decomposes to:
ۀ [U+06D5 ARABIC LETTER AE + U+0654 ARABIC HAMZA ABOVE]
ۂ [U+06C2 ARABIC LETTER HEH GOAL WITH HAMZA ABOVE] decomposes to:
ۂ [U+06C1 ARABIC LETTER HEH GOAL + U+0654 ARABIC HAMZA ABOVE]
یٔ [U+06CC ARABIC LETTER FARSI YEH + U+0654 ARABIC HAMZA ABOVE] 

ئ [U+0626 ARABIC LETTER YEH WITH HAMZA ABOVE] decomposes to:

ئ [U+064A ARABIC LETTER YEH + U+0654 ARABIC HAMZA ABOVE] 
[U+08A8 ARABIC LETTER YEH WITH TWO DOTS BELOW AND HAMZA ABOVE] (no decomposition)

① The Unicode Standard explicitly recommends use of the decomposed sequence when combining a hamza with HEH (for ezafe). The two precomposed characters on the right are problematic because they decompose to sequences containing ە [U+06D5 ARABIC LETTER AE] or ہ [U+06C1 ARABIC LETTER HEH GOAL], neither of which are appropriate for Persian.

② The precomposed YEH with hamza appears, on the face of it, to be a likely alternative for the sequence on the left. However, the decomposition is to hamza over ي [U+064A ARABIC LETTER YEH] rather than ی [U+06CC ARABIC LETTER FARSI YEH], and that letter is not used for Persian (it doesn't represent the dots in the way FARSI YEH does). In fact, the Unicode Standard requires the precomposed and decomposed forms of the Arabic YEH combined with hamza to never have dots below in any of the joining forms. For this reason, this page uses and recommends the decomposed sequence for yeh+hamza. That said, however, in the wild the precomposed character appears to be widely used.

[U+08A8 ARABIC LETTER YEH WITH TWO DOTS BELOW AND HAMZA ABOVE] is a character used for an African orthography which retains the dots for all joining forms. It is therefore also an inappropriate precomposed character for Persian.

Content authors should use a separate hamza for these sequences in Persian, even though the precomposed characters look the same visually, because they don't represent the same semantics, and may introduce problems if text is decomposed. However, because approaches may yield exactly the same result when displayed, applications will need to recognise the precomposed characters and treat them or map them to the appropriate sequence. Input mechanisms, on the other hand, can produce one rather than the other, and that choice should be made with advisement.

Confusables & spelling errors

The following lists some common errors found in Persian text due to the similarity of Unicode characters, or perhaps sometimes due to problems inputting the correct character.

Correct Incorrect  
ی [U+06CC ARABIC LETTER FARSI YEH] ي [U+064A ARABIC LETTER YEH]
ه [U+0647 ARABIC LETTER HEH]

ہ [U+06C1 ARABIC LETTER HEH GOAL]

ە [U+06D5 ARABIC LETTER AE] 

ة [U+0629 ARABIC LETTER TEH MARBUTA]

ۃ [U+06C3 ARABIC LETTER TEH MARBUTA GOAL]

ک [U+06A9 ARABIC LETTER KEHEH] ك [U+0643 ARABIC LETTER KAF]

① As mentioned in the previous section, Arabic YEH doesn't drop the dots below the letter in isolate and final positions.

② Although they look the same, these are all different characters. The HEH GOAL is used for languages that include Urdu and Kashmiri, whereas the LETTER AE is used for Uighur and central Asian languages.

③ & ④ Again, characters that a font may render exactly the same, but that are based on different base letters.

Although these characters look the same (at least in certain joining forms), they are all different characters, and should not be used interchangeably. If they are used, the ability to search and compare text is impaired unless the application is aware of and takes counter-measures against this substitution.

Numbers

Persian uses the extended arabic-indic digits in the Arabic block.

۰␣۱␣۲␣۳␣۴␣۵␣۶␣۷␣۸␣۹

This is a separate set of characters from those used for Arabic, to accommodate different shaping and directional behaviour. Shapes differ from those of Arabic for the digits 4, 5, and 6.

Persian۰۱۲۳۴۵۶۷۸۹
Arabic٠١٢٣٤٥٦٧٨٩
Comparison of digit shapes in Persian and Arabic.

See expressions for a discussion of how to handle numeric ranges.

Percentages

٪

Persian may use the Arabic percent sign, ٪ [U+066A ARABIC PERCENT SIGN].

؜۵٬۴۳۲٫۱٪

The figure 5,432.1% using Persian characters.

The percent sign is typed and stored after the numbers. Like the numeric sequences using the ASCII hyphen (mentioned in expressions), it will appear to the left of a number if that number is preceded by Persian characters. However, if the percentage appears alone or at the beginning of a line it is necessary to use an ALM formatting character just before it to prevent the sign appearing on the right.

Observation: Wikipedia uses an ASCII percent sign with ASCII digits

Dates

TBD

Currency

The name of the currency is usually spelled out: ریال rial

The Unicode Standard does have a symbol code point, [U+FDFC RIAL SIGN], but it is only a compatibility character for use when converting from Iranian standards, and should not be used in normal Unicode textu14,379.

Formatting characters

The Arabic script uses a number of Unicode characters that affect the way that other characters are rendered.

Persian text makes use of a relatively large set of invisible formatting characters, especially in plain text, many of which are used to manage text direction (see directioncontrols), and others are used to control cursive shaping behaviour (see shapingcontrols).

Text direction

Persian is written horizontally and right-to-left in the main, but (as with most RTL scripts) numbers and embedded LTR script text are written left-to-right (producing 'bidirectional' text).

کنسرسیوم یونیکد برای اولین بار Unicode Standard را در سال 1991 منتشر کرد (نسخه 1.0)
Persian words are read right-to-left, starting from the right of the line, but numbers and Latin text (highlighted) are read left-to-right.

The Unicode Bidirectional Algorithm automatically takes care of the ordering for all the text in fig_bidirectional, as long as the 'base direction' is set to RTL. In HTML this can be set using the dir attribute, or in plain text using formatting controls.

If the base direction is not set appropriately, the directional runs will be ordered incorrectly as shown in fig_base_direction.

کنسرسیوم یونیکد برای اولین بار Unicode Standard را در سال 1991 منتشر کرد (نسخه 1.0)
The exact same sequence of characters with the base direction set to RTL (top), and with no base direction set on a LTR page (bottom).

Show default bidi_class properties for characters in the Persian orthography described here.

For more information about how directionality and base direction work, see Unicode Bidirectional Algorithm basics. For information about plain text formatting characters see How to use Unicode controls for bidi text. And for working with markup in HTML, see Creating HTML Pages in Arabic, Hebrew and Other Right-to-left Scripts.

On this page, see also expressions and breaking_latin for additional features related to direction.

Managing text direction

Unicode provides a set of 10 formatting characters that can be used to control the direction of text when displayed. These characters have no visual form in the rendered text, however text editing applications may have a way to show their location.

RLE [U+202B RIGHT-TO-LEFT EMBEDDING] (RLE), LRE [U+202A LEFT-TO-RIGHT EMBEDDING] (LRE), and PDF [U+202C POP DIRECTIONAL FORMATTING] (PDF) are in widespread use to set the base direction of a range of characters. RLE/LRE come at the start, and PDF at the end of a range of characters for which the base direction is to be set.

More recently, the Unicode Standard added a set of characters which do the same thing but also isolate the content from surrounding characters, in order to avoid spillover effects. They are RLI [U+2067 RIGHT-TO-LEFT ISOLATE] (RLI), LRI [U+2066 LEFT-TO-RIGHT ISOLATE] (LRI), and PDI [U+2069 POP DIRECTIONAL ISOLATE] (PDI). The Unicode Standard recommends that these be used instead.

There is also PDI [U+2068 FIRST STRONG ISOLATE] (FSI), used at the start of a range to set the base direction according to the first recognised strongly-directional character.

RLM [U+200F RIGHT-TO-LEFT MARK] (RLM) and LRM [U+200E LEFT-TO-RIGHT MARK] (LRM) are invisible characters with strong directional properties that are also sometimes used to produce the correct ordering of text.

For more information about how to use these formatting characters see How to use Unicode controls for bidi text. Note, however, that when writing HTML you should generally use markup rather than these control codes. For information about that, see Creating HTML Pages in Arabic, Hebrew and Other Right-to-left Scripts.

Expressions & sequences

A sequence of numbers separated by hyphens (for example a range) runs from left to right in Persian (unlike Arabic language text).

fig_bidi_range shows some Persian text, which is right-to-left overall, containing a numeric range that is ordered LTR, ie. it starts with ۱۱۶۹ (1169) and ends with ۱۱۷۰ (1170).

۱۱۶۹-۱۱۷۰ 	آغاز نگارش هیستوریا به دستور امالریک یکم
A numeric range in Persian language text.

Certain types of hyphen or other characters can affect the way expressions run, and you need to be aware of that when writing Persian text. For more details see the Expressions & sequences section in the description of the Arabic language orthography.

Glyph shaping & positioning

This section brings together information about the following topics: writing styles; cursive text; context-based shaping; context-based positioning; baselines, line height, etc.; font styles; case & other character transforms.

You can experiment with examples using the Persian character app.

The orthography has no case distinction, and no special transforms are needed to convert between characters.

Writing styles

Persian may be written in a nasta'liq writing style. Key features include a sloping baseline for joined letters, and overall complex shaping and positioning for base letters and diacritics alike. There are also distinctive shapes for many glyphs and ligatures.

مستحق  •  شخص  •  کیفیت
Sloping baselines and complex joining behaviours in Persian nastaliq text.

This is achieved in Unicode by applying the correct font – the underlying characters used are not different for nasta'liq vs. other styles.

The Wikipedia Persian home page, showing the title in nasta'liq script.

Persian may also be written in several other styles, especially in artistic and historical writing.

Cursive script

Arabic script joins letters together. Fonts need to produce the appropriate joining form for a code point, according to its visual context. This results in four different shapes for most letters (including an isolated shape). The highlights in fig_cursive below show the same letter, ع [U+0639 ARABIC LETTER AIN], with two different joining forms. 

عقل ودیعت
The letter ع [U+0639 ARABIC LETTER AIN] in 2 different joining contexts.

A few Arabic script letters only join on the right-hand side.

There are 2 Unicode blocks containing Arabic presentation forms: these contain individual characters corresponding to the various joining forms and ligatures. With only a handful of exceptions, characters in those blocks should not be used for text content; they are only for managing legacy encodings. Instead, characters in the main Arabic block should be used, and the font will manage the necessary cursive shaping.

Cursive joining forms

Most dual-joining characters add or become a swash when they don't join to the left. A number of characters, however, undergo additional shape changes across the joining forms. fig_joining_forms and fig_right_joining_forms show the basic shapes in Persian and what their joining forms look like.

Two pairs of characters in the first table have base shapes that are identical, but they manage the dots differently in different joining forms. These have been put onto separate rows.

isolatedright-joineddual-joinleft-joinedPersian letters
ب ـب ـبـ بـ
ب␣ت␣ث␣پ
ن ـن ـنـ نـ
ن
ق ـق ـقـ قـ
ق
ف ـف ـفـ فـ
ف
س ـس ـسـ سـ
س␣ش
ص ـص ـصـ صـ
ص␣ض
ط ـط ـطـ طـ
ط␣ظ
ک ـک ـکـ کـ
ک␣گ
ل ـل ـلـ لـ
ل
ه ـه ـهـ هـ
ه
م ـم ـمـ مـ
م
ع ـع ـعـ عـ
ع␣غ
ح ـح ـحـ حـ
ح␣خ␣ج␣چ
ی ـی ـیـ یـ
ی
Joining forms for shapes that join on both sides.
isolatedright-joined Persian letters
ا ـا
ا␣آ␣أ
ر ـر
ر␣ز␣ژ
د ـد
د␣ذ
و ـو
و␣ؤ
Joining forms for shapes that join on the right only.

Managing glyph shaping

ZWJ [U+200D ZERO WIDTH JOINER] (ZWJ) and ZWNJ [U+200C ZERO WIDTH NON-JOINER] (ZWNJ) are used to control the joining behaviour of cursive glyphs. They are particularly useful in educational contexts, but also have real world applications.

ZWJ permits a letter to form a cursive connection without a visible neighbour. It can be used for illustrating cursive joining forms, eg. ان‍‍   ‍س‍‍   ‍ان Characters from the Presentation Forms blocks in Unicode should not be used in such cases.

ZWNJ prevents two adjacent letters forming a cursive connection with each other when rendered, eg. ان‌س‌ان

This is particularly useful for Persian, since certain Persian suffixes don't join with word-final letters when they appear finally in a morpheme, eg. خانه‌ها تکیه‌گاه

Click on the words above to see the composition.

To achieve this, you need to use ZWNJ [U+200C ZERO WIDTH NON-JOINER]. It's also possible to sometimes see text where the suffix is written after a space, or simply joined to the end of the word. However, those alternatives are not available when the word ends with ه [U+0647 ARABIC LETTER HEH].§

͏MVS [U+034F COMBINING GRAPHEME JOINER] is used in Arabic to produce special ordering of diacritics. The name is a misnomer, as it is generally used to break the normal sequence of diacritics.

Context-based shaping

Context-based shaping is everwhere in Persian due to the combination of the cursive behaviour of the script plus the strong tendency to arrange joined characters in cascades or vertical arrangements.

As in Arabic, lam followed by alef ligates, eg. اسلام and there are other such commonly ligated forms.

Context-based positioning

Depending on the font, Arabic letters often have special rules for joining between certain characters, and diacritic marks generally vary in height and horizontal position depending on the size of the base character.

Another example of contextual positioning rules is the placement of ِ [U+0650 ARABIC KASRA] (zir) in vowelled text when it appears on the same letter as ّ [U+0651 ARABIC SHADDA] (tašdid). Usually, zir appears below the base letter, and this is how it can be distinguished from َ [U+064E ARABIC FATHA] (zebar). However, when combined, zir may be placed relative to the shadda diacritic, rather than relative to the base character, as seen in fig_kasra_placement.

مَمِمّمَّمِّ

The word تپّه with vowel diacritics, showing the zir below the tašdid, rather than below the base letter.

Positioning of cursive joining forms is particularly complicated in the nastaliq style. See more details in the Urdu page.

Baselines, line height, etc.

The alphabetic baseline is a strong feature of Arabic script on the whole, since characters tend to join there. The nastaliq style of the script, on the other hand, uses arrangements of joined glyphs that cascade downwards from right to left, and ressemble a strongly sloping baseline.

Font styles

tbd

Punctuation & inline features

Grapheme boundaries

tbd

Word boundaries

Words are separated by spaces.

Phrase & section boundaries

،␣؛␣:␣.␣؟␣!
phrase

، [U+060C ARABIC COMMA]

؛ [U+061B ARABIC SEMICOLON]

: [U+003A COLON]

sentence

. [U+002E FULL STOP]

؟ [U+061F ARABIC QUESTION MARK] 

! [U+0021 EXCLAMATION MARK]

Persian uses a mixture of western and arabic punctuation.

Parentheses & brackets

(␣)
  start end
standard

( [U+0028 LEFT PARENTHESIS]

) [U+0029 RIGHT PARENTHESIS]

Mirrored characters

The words 'left' and 'right' in the Unicode names for parentheses, brackets, and other paired characters should be ignored. LEFT should be read as if it said START, and RIGHT as END. The direction in which the glyphs point will be automatically determined according to the base direction of the text.

a > b > c
ا > ب > ج
Both of these lines use > [U+003E GREATER-THAN SIGN], but the direction it faces depends on the base direction at the point of display.

The number of characters that are mirrored in this way is around 550, most of which are mathematical symbols. Some are single characters, rather than pairs. The following are some more common ones.

(␣)␣<␣>␣[␣]␣{␣}␣«␣»␣‹␣›

Quotations

«␣»
  start end
initial

« [U+00AB LEFT-POINTING DOUBLE ANGLE QUOTATION MARK]

» [U+00BB RIGHT-POINTING DOUBLE ANGLE QUOTATION MARK]

The quotation marks typically have rounded glyph shapes, rather than sharp angles.

Emphasis

tbd

Abbreviation, ellipsis & repetition

tbd

Inline notes & annotations

tbd

Other inline ranges

tbd

Other punctuation

tbd

Line & paragraph layout

Line breaking & hyphenation

Basic line-break opportunities occur between the space-separated words.

They are not broken at the small gaps that appear where a character doesn't join on the left.

Breaking between Latin words

When a line break occurs in the middle of an embedded left-to-right sequence, the items in that sequence are rearranged visually so that the reading direction remains top-to-bottom. latin_line_breaks shows how two Latin words are apparently reordered in the flow of text to accommodate this rule.

کنسرسیوم یونیکد برای اولین بار Unicode Standard را در سال 1991 منتشر کرد (نسخه 1.0)

کنسرسیوم یونیکد برای اولین بار Unicode Standard را در سال 1991 منتشر کرد (نسخه 1.0)

Persian with embedded Latin text. The lower of these two images shows the result of decreasing the line width, so that text wraps between a sequence of Latin words.

In digital text the rearrangement is automatic. Only the positions of the font glyphs are changed: nothing affects the order of the characters in memory.

Show (default) line-breaking properties for characters in the Persian orthography described here.

Text alignment & justification

ـ

See the section on justification for the Arabic language.

Letter spacing

See the section on letter spacing for the Arabic language.

Counters, lists, etc.

You can experiment with counter styles using the Counter styles converter. Patterns for using these styles in CSS can be found in Ready-made Counter Styles, and we use the names of those patterns here to refer to the various styles.

The Persian language uses 1 numeric and 2 fixed styles.

Numeric

The persian numeric style is decimal-based and uses these digits.rmcs,#arabic-styles

۰␣۱␣۲␣۳␣۴␣۵␣۶␣۷␣۸␣۹

Examples:

۱␣۲␣۳␣۴␣۱۱␣۲۲␣۳۳␣۴۴␣۱۱۱␣۲۲۲␣۳۳۳␣۴۴۴

Fixed

The arabic-abjad fixed style uses these letters. It is only able to count to 28.rmcs,#arabic-styles

ا␣ب␣ج␣د␣ه‍␣و␣ز␣ح␣ط␣ی␣ک␣ل␣م␣ن␣س␣ع␣ف␣ص␣ق␣ر␣ش␣ت␣ث␣خ␣ذ␣ض␣ظ␣غ

Note that the 5th counter includes a zero-width joiner formatting character. This makes the shape distinguishable from ٥ [U+0665 ARABIC-INDIC DIGIT FIVE].

The persian-alphabetic fixed style uses these letters. It is able to count to 32. The letters are arranged by shape.rmcs,#arabic-styles

ا␣ب␣پ␣ت␣ث␣ج␣چ␣ح␣خ␣د␣ذ␣ر␣ز␣ژ␣س␣ش␣ص␣ض␣ط␣ظ␣ع␣غ␣ف␣ق␣ک␣گ␣ل␣م␣ن␣و␣ه‍␣ی

The 31st counter also includes a zero-width joiner formatting character.

Prefixes and suffixes

Persian lists generally use a full stop suffix as a separator.

Styling initials

tbd

Page & book layout

This section is for any features that are specific to Persian and that relate to the following topics: general page layout & progression; grids & tables; notes, footnotes, etc; forms & user interaction; page numbering, running headers, etc.

General page layout & progression

Persian books, magazines, etc., are bound on the right-hand side, and pages progress from right to left.

عنوان كتاب

Binding configuration for Persian books, magazines, etc.

Columns are vertical but run right-to-left across the page.

Grids & tables

Tables, grids, and other 2-dimensional arrangements progress from right to left across a page.

Forms & user interaction

Form controls should display Persian text from right to left, starting at the right side of the input field. Form controls should also usually be arranged from right to left.

fig_form shows some form fields from an Arabic language web page. Note the position of the labels relative to the input fields and the checkbox, mirror-imaging a similar page in English. Note also that the input text in the first field appears to the right of the box.

A set of form fields on an Arabic web page

The position of a scrollbar should depend on the user's environment, not on the content of a page. A non-Persian user viewing a web page in Persian shouldn't have to look for the scroll bar on the left side of the window. In a system that is set up for an Persian user, however, the scrollbar may appear on the left.

Languages using the Arabic script

According to ScriptSource, the Arabic script is used for the following languages:

References