Transformation of text

Transformations of text are strategies to perform geometric transformations on text (reversals, rotations, etc.), particularly in systems that do not natively support transformation, such as HTML, seven-segment displays and plain text.


Many systems, such as HTML, seven-segment displays and plain text, do not support transformation of text. In the case of HTML, this limitation in display may eventually be addressed through standard cascading style sheets (CSS), since proposed specifications for CSS3 include rotation for block elements.[1] In the meantime, several ways of producing the visual effects of text transformations have come into use.

The most common of these transformations are rotation and reflection.

Unicode supports a variety of characters that resemble transformed characters, primarily for various forms of phonetic transcription. Each of these character names indicates what kind of transformation the characters have undergone:

Upside-down text

Strategies can be used to render words upside down in languages such as HTML that do not permit rotation of text; using Unicode characters (especially those in the IPA), a very close approximation of upside-down text (also called flip text) can be achieved. The letters s, x, z and o are rotationally symmetrical, while pairs such as b/q, d/p and n/u are rotations of each other. The rest of the letters have been encoded into the Unicode IPA section, generating a complete set of upside-down lowercase letters. With the addition of the Fraser alphabet to the Unicode standard in version 5.2, full (or at least near-full) support for upside-down capital letters is now available. Number support is incomplete; four numbers are universally strobogrammatic (0, 8, and 6/9), and the upside-down versions of numbers 2 and 3 have been provisionally assigned Unicode points for use in dozenal notation; however, other numbers still are not supported. Punctuation (by use of such characters as the interpunct and the inverted question mark and exclamation point) is mostly covered. Several Internet utilities exist for the transformation of regular text to (and sometimes from) upside-down text; each has its own slightly different algorithm for letters not precisely or well covered. A list of converters and algorithms can be found at the list below.

A similar process is USD encoding, which uses characters entirely within the ASCII character set. Because it is almost entirely alphanumeric, it is far more compatible with other programs that do not support Unicode, and more readily typed by hand. However, the text created by using USD encoding is far less legible, and in fact more closely resembles Leet. Another problem is that because not all letters fit well, the USD algorithms cannot be a complete involution (i.e., completely convertible back and forth) and contain a complete set of letters at the same time. For instance, the Albartus USD algorithm example seen in the "Examples" section below has k, T, t, and R still in their upright positions. Another issue with USD encoding is the use of italic type. The letter "a" will, in most typefaces using italic fonts, render it as a "one-story" Latin alpha, thus causing problems with any word using that letter as a lowercase "e." Oblique type does not have this problem.

Below is a conversion table that can be used to transform lowercase, uppercase numeric and punctuation output. These characters require unicode version 8.0 minimum (in particular the ᘔ and Ɛ from the duodecimal block).

Lowercase Letters z ʎ x ʍ ʌ n ʇ s ɹ b d o u ɯ l ʞ ɾ ɥ ɓ ɟ ǝ p ɔ q ɐ
007A 028E 0078 028D 028C 006E 0287 0073 0279 0062 0064 006F 0075 026F 006C 029E 027E 0131 0265 0253 025F 01DD 0070 0254 0071 0250
Capital Letters Z X M Λ S Ό Ԁ O N W ſ I H Ǝ Ɔ
005A 2144 0058 004D 039B 2229 22A5 0053 1D1A 038C 0500 004F 004E 0057 2142 22CA 017F 0049 0048 2141 2132 018E 15E1 0186 10412 2200
Numbers 0 6 8 9 5 ߈ Ɛ
0030 0036 0038 3125 0039 03DA 07C8 218B 218A 21C2
Punctuation ¿ ¡ , ˙ ' ؛
214B 203E 00BF 00A1 201E 002C 02D9 0027 061B

Sideways text

Sideways text presents a unique problem. Though it is likely the most practical (as opposed to artistic) form of text transformation, it is the least supported and is the most difficult to implement. Unlike rotating text 180 degrees, the number of sideways characters falls far short of what would be needed for most purposes, and because text is rendered horizontally, it would be very difficult to render beyond one line of vertical text in a well-aligned manner without columns, especially in proportional fonts (furthermore, each character would require a line break after it). The process of using alternate characters for sideways text is further complicated by the fact that most fonts space letters further apart vertically (to accommodate underlining and overlining) than horizontally, and that most fonts are taller than they are wider, making simulated sideways text look significantly more awkward.

Internet Explorer has a CSS property that will rotate normally entered text 90 degrees clockwise:

<div style="writing-mode:tb-rl;">

However, no other major browsers (Mozilla Firefox, Opera nor WebKit based browsers) support this writing-mode property. However, all major standards-compliant browsers now support CSS3 rotation for block elements, which makes the visual rotation of HTML text available.[2]

The most common way around these problems is to use images of text, which can then be rotated and transformed in an image editor at will, and to represent the text in those images with the alt attribute so that search engines and text-only browsers can read it properly. The use of ANSI art and box-drawing characters to manually draw sideways text has the advantage of being copiable and pastable (whereas images are not in most plain text situations), but generally creates large characters.

Reversed text

Main article: Mirror writing

Though less widespread, text can also be reversed to be a mirror image of itself. Letters A, H, I, M, O/o, T, U, V/v, W/w, X/x, Y, and in some fonts i and l are symmetrical in the y-axis; the pairs of b/d and p/q transform to each other. The letters И, Я, and г from Cyrillic, among other sources, are among the numerous characters that can be used to further generate this effect. Reversed text can use capital letters mixed with lowercase, as opposed to the strict lowercase used by upside-down transformation (upside-down lowercase and capital letters do not generally align as they would upright, though reversed letters do).

X-axis symmetry is visible in the letters B, C, D, E, H, I, K, O, X, and in some fonts a and l, as well as in the pairs of a/g, b/p, d/q, e/G, and f/t. Expanding to Cyrillic and Greek produces more symmetries, such as Λ/V and Γ/L.

The Fixedsys Excelsior typeface includes a complete set of reversed characters like this in its Private Use Area. However, online utilities to create mirrored text are not readily available, and most sites that claim to "mirror text" or "reverse text" in fact only change the order of the letters and do not actually flip the letters themselves.

Dilated text

Through the use of Unicode's small capitals and subscript and superscript phonetic modifiers, text can be created that is smaller than the inline text. This is generally only necessary for applications that only support one-size plain text, since HTML and CSS support different text sizes.


Question: How can you tell an introvert from an extrovert?
Answer: ˙sǝoɥs s,ʎnƃ ɹǝɥʇo ǝɥʇ ʇɐ sʞool ʇɹǝʌoɹʇxǝ ǝɥʇ 'sɹoʇɐʌǝlǝ ǝɥʇ uı (Using the Revfad algorithm)
Or: 'saoys s.hn6 R3HTO ayt te skool tJa^oJtxa ayt `sJote^ala ayt uI (using the Albartus USD algorithm)

Example of reversed text reflected along a y-axis:

Example:...иiɒəɒ иɘqo x иoiƨиɘмib oɟ lɒɟɿoq ɘнɟ ɟʇɘl γbodɘмoƧ (Somebody left the portal to Dimension X open again...)

Poet Darius Bacon has written two examples of palindromic poetry that reads the same upside-down as it does upside right.[3]

Comparison of algorithms

Converter Lowercase Uppercase Numbers Reconvert
Reflection Character set Provides HTML
Text Filter
Yes Yes Yes Yes Yes All of Unicode No
Flip-O-Matic Yes Incomplete Errors Yes No All of Unicode No, but highly recommended
CedarCreek Labs Yes Yes Yes No No Latin characters in Unicode No
Yes No No Yes No Latin characters in Unicode No
Yes Yes Yes No No All of Unicode
Uses combining characters for some
No Yes No No No No Latin, with one exception No Yes No Yes Yes No Latin for letters
all of Unicode for numbers
No Yes Yes No Yes No Latin, with one exception No
Upside Down Letters
Yes No No Yes No Latin, with one exception No
USD encoding
Incomplete Incomplete Yes Yes No ASCII N/A
Yes Yes No, errors Almost No Windows-1252 No Yes Incomplete Yes No X-axis, Incomplete All of Unicode Yes Yes Incomplete Yes Yes Errors Latin characters with Unicode Yes Yes No No Yes No Latin for letters
all of Unicode for numbers
Calculator spelling Incomplete Incomplete Incomplete No No, but conceivable Numerals 0-9 No
Write upside down
Yes Yes Yes Yes No All of Unicode No
Lunicode Yes Yes Yes Yes Y-axis All of Unicode Yes


  1. Bert Bos, ed. (August 9, 2007). "CSS basic box model". W3C. Retrieved 2012-11-19.
  2. Can I use... CSS3 transforms
  3. Bacon, Darius. a poem and deus am. The Palindromist #4.
This article is issued from Wikipedia - version of the 9/3/2016. The text is available under the Creative Commons Attribution/Share Alike but additional terms may apply for the media files.