My understanding of .chars is that it returns "the number of characters in the string in graphemes". My understanding of .ords is that it returns "a list of codepoint numbers, one for the base character of each grapheme in the string". That is, .chars returns the number of graphemes and .ords returns one codepoint (the base) per grapheme. However, the behavior I am seeing in Rakudo 2016.07.1 on MoarVM 2016.07 doesn't seem to match that:
> "\x[2764]\x[fe0e]".chars
1
> "\x[2764]\x[fe0e]".ords.fmt("U+%04x")
U+2764 U+fe0e
> "e\x[301]".ords.fmt("U+%04x")
U+00e9
> "0\x[301]".ords.fmt("U+%04x")
U+0030
The .chars method returns the expect 1 for the HEAVY BLACK HEART and VARIATION SELECTOR-15 (text representation ❤︎ rather than emoji ❤️, U+2764 U+fe0f), but then .ords returns both codepoints rather than just the base (I expected just U+2764). Even more confusing, if you call .ords on LATIN SMALL LETTER E and COMBINING ACUTE ACCENT, you get back U+00e9 (LATIN SMALL LETTER E WITH ACUTE). I was expecting U+0065 as LATIN SMALL LETTER E is the base codepoint. I do get back the expected result when there isn't an NFC version of the string (eg U+0030 for 0́).
Is my understanding of .chars and .ords just flawed, or is this a bug?
Documentation bug regarding the .ords method. One of the core developers has just updated the docs with this commit:
https://github.com/perl6/doc/commit/12ec5fc35e
Which should appear on the site shortly.
Related
What are the rules for reading a UTF-16 byte stream, to determine how many bytes a character takes up? I've read the standards, but based on empirical observations of real-world UTF-16 encoded streams, it looks like there are certain where the standards don't hold true (or there's an aspect of the standard that I'm missing).
From the reading the UTF-16 standard https://www.rfc-editor.org/rfc/rfc2781:
Value of leading 2 bytes
Resulting character length (bytes)
0x0000-0xC7FF
2
0xD800-0xDBFF
4
0xDC00-0xDFFF
Invalid sequence (RFC2781 2.2.2)
0xDFFF-0xFFFF
4
In practice, this appears to hold true, for some cases at least. Using an ad-hoc SQL script (SQL Server 2019; UTF-16 collation), but also verified with an online decoder:
Character
Unicode Name
ISO 10646
UTF-16 Encoding (hexadecimal, big endian)
Size (bytes)
A
LATIN CAPITAL LETTER A
U+0041
00 41
2
Б
CYRILLIC CAPITAL LETTER BE
U+0411
04 11
2
ァ
KATAKANA LETTER SMALL A
U+30A1
30 A1
2
🐰
RABBIT FACE
U+1F430
D8 3D DC 30
4
However when encoding the following ISO 10646 character into UTF-16, it appears to be 4 bytes, but reading the leading 2 bytes appears to give no indication that it will be this long:
Character
Unicode Name
UTF-16 Encoding (hexadecimal, big endian)
Size (bytes)
⚕️
STAFF OF AESCULAPIUS
26 95 FE 0F
4
Whilst I'd rather keep my question software-agnostic; the following SQL will reproduce this behaviour on Microsoft SQL Server 2019, with default collation and default language. (Note that SQL Server is little endian).
select cast(N'⚕️' as varbinary);
----------
0x95260FFE
Quite simply, how/why do you read 0x2695 and think "I'll need to read in the next word for this character."? Why doesn't this appear to align with the published UTF-16 standard?
The formal definition of all of this is called an "extended grapheme cluster," and it's defined in the Unicode Text Segmentation report. As Joachim Sauer notes, it's wise to be careful with the term "character" in Unicode.
Code points are what "U+...." syntax is referring to, and is attempting to capture a "unit" of written language, for example "an acute accent." But what a reader would think of a character (for example "an e with an acute accent") is a "grapheme cluster" and is made up of one or more code points. What is ultimately rendered to the screen is a "glyph" which is both context- and font-dependent.
Grapheme clusters in Unicode are actually more subtle than this. Unicode attempts to define them in a "neutral" way. (There's really no such thing as "neutral" when thinking about languages, but Unicode does try.) For example, in Slovak, ch, dz, and dž are each one letter, but are considered two grapheme clusters in Unicode. (Try to count the "letters" in a Slovak word. There are words that contain the letter dz and other words that have the letter d followed by the letter z. Oh human writing systems. I love you so much.)
The mapping of grapheme clusters to glyphs is also complex. For example, in Arabic, the single glyph لا is actually two grapheme clusters, ل (ARABIC LETTER LAM) followed by ا (ARABIC LETTER ALEF). If you use your mouse to select the glyph, you'll see there are two selectable pieces, and if you copy and paste them to another window you'll see them transform into their component parts. (Just to make thing even more complicated, Unicode also defines a single code point for ligature, ARABIC LIGATURE LAM WITH ALEF ISOLATED FORM: ﻻ. If you try to select part of that one, you'll find you can't. It's one "character.")
Your specific case is a bit more special. The Variation Selector predates Unicode, and is mostly designed to handle different variations of Han (Chinese) characters. However, as with every Unicode feature, it eventually has come to be used primarily for emoji. VS-16 is the "emoji" presentation form. The most famous example is the red heart, which is HEAVY BLACK HEART ❤, followed by VS-16: ❤️.
Similarly, your character U+2695 STAFF OF AESCULAPIUS is a single code point, and it looks like this by default (text style): ⚕. When you add VS-16, it is rendered in "emoji style": ⚕️. In some ways it's the same "character." Or is it? Depends on what you're using it for.
Emoji style is typically a bit larger and centered in its block, sometimes adding color. Notice where the period after the staff is drawn in each case (there are no extra spaces in the second example; the glyph is just much wider).
There are other combining systems as well:
U+0031: 1
U+0031 U+20e3: 1⃣ (+ COMBINING ENCLOSING KEYCAP, default text style)
U+0031 U+20e3 U+fe0f: 1⃣️ (+ VARIATION SELECTOR-16, emoji style)
All of these predate Unicode. Modern emoji is dramatically more complicated, and includes several combining systems of its own (including two that are currently just used for flags).
But luckily, to your actual question, your wife is correct, and you can generally just consume all trailing code points that are marked "combining" to form an extended grapheme cluster, and that is kind of a "character" for some broad enough definition of "character."
All of your assertions are completely correct; your interpretation of the UTF-16 standards is correct and complete.
In your empirical observations however, you've assumed that you only have one character. In actuality, you've ran into a nuance of the Unicode implementation. Your "character" is actually two (albeit technically, not visually): U+2695 "STAFF OF AESCULAPIUS" followed by U+FE0F "VARIATION SELECTOR-16". The second character is a non-spacing mark which combines with the base character for the purpose of rendering a character variant.
This results in the byte sequence 26 95 FE 0F, however as you note neither of the words fall within the UTF-16 reserved extension character range. But this is because neither of them require the UTF-16 4 byte extension. They're simply classified as two discrete Unicode characters.
From 7.9 Combining Marks in ISO 10646: Universal Coded Character Set (UCS):,
Combining marks are a special class of characters in the Unicode Standard that are
intended to combine with a preceding character, called their base.
Combining marks usually have a visible glyphic form... a combining mark may interact graphically with neighbouring characters in various ways.
http://unicode.org/L2/L2010/10038-fcd10646-main.pdf
To explain why I'm answering my own question; I had my SO question all ready to fire off. My wife came into my office; after looking over my shoulder she whispered into my ear, "You know combination characters are a thing, right?". I've however still asked the question and answered it myself, in case my wife's sweet nothings help another member of the community.
The following two code practically did the same thing
for character in "Dog!🐶".characters {
print(character)
}
for character in "Dog!🐶".unicodeScalars {
print(character)
}
However, when I check for more detail information behind the sense, I found the difference. The characters property is the type of CharacterView while unicodeScalars is the type of UnicodeScalarView.
Question:
What's the difference between them?
Which property is preferred for what situation? (would be nice to have an example)
Many Thanks
This comes down to the difference between a Character and UnicodeScalar.
Unicode Scalars
Behind the scenes, Swift’s native String type is built from Unicode scalar values. A Unicode scalar is a unique 21-bit number for a character or modifier, such as U+0061 for LATIN SMALL LETTER A ("a"), or U+1F425 for FRONT-FACING BABY CHICK ("🐥").
...
Extended Grapheme Clusters
Every instance of Swift’s Character type represents a single extended grapheme cluster. An extended grapheme cluster is a sequence of one or more Unicode scalars that (when combined) produce a single human-readable character.
Here’s an example. The letter é can be represented as the single Unicode scalar é (LATIN SMALL LETTER E WITH ACUTE, or U+00E9). However, the same letter can also be represented as a pair of scalars—a standard letter e (LATIN SMALL LETTER E, or U+0065), followed by the COMBINING ACUTE ACCENT scalar (U+0301). The COMBINING ACUTE ACCENT scalar is graphically applied to the scalar that precedes it, turning an e into an é when it is rendered by a Unicode-aware text-rendering system.
From the Strings and Characters section of the Swift Programming Language Guide.
In most cases that I can think of you'll want to be dealing with Character instances, as they as the smallest unit of Human language. I can't imagine a situation where you'd want to operate on a modifier without considering the full extended grapheme cluster.
What is the likelihood that I'll run into COMBINING LATIN SMALL LETTER C (U+0368) in "real life" (besides clever Scottish folk)?
I'm asking since it's in both the Unicode Block Combining Diacritical Marks and the Category Mark, Nonspacing [Mn].
As a result, it seems to gets treated the same as characters such as COMBINING GRAVE ACCENT (U+0300) by Utilities such as the ICU Transliterator (using either the suggested "NFD; [:Nonspacing Mark:] Remove; NFC" or a straight "Latin-ASCII" transliteration).
The likelihood is very close to zero, but not exactly zero. You cannot prevent anyone from using a Unicode character as he likes. There is no specific information about U+0368 in the Unicode Standard, but it has definitely been defined as a combining character that will cause a symbol (c) to be displayed above the preceding character. I would expect to find it mostly in digitized forms of medieval manuscripts, or something like that.
Using it after a space character, as in the “clever” page mentioned, is not the intended use, but not invalid either. Unicode lets you use any combining mark after any character, whether it makes sense or not.
It has no canonical or compatibility decomposition, so there is no clear-cut way to deal with in a context where you cannot, or do not want to, retain the character.
The likelihood is utterly indeterminate except to say that if you expect it not to occur, then it will occur.
Quoted from here:
Security may also be impacted by a characteristic of several character
encodings, including UTF-8: the "same thing" (as far as a user can
tell) can be represented by several distinct character sequences. For
instance, an e with acute accent can be represented by the precomposed
U+00E9 E ACUTE character or by the canonically equivalent sequence
U+0065 U+0301 (E + COMBINING ACUTE). Even though UTF-8 provides a
single byte sequence for each character sequence, the existence of
multiple character sequences for "the same thing" may have security
consequences whenever string matching, indexing,
Is this a hidden feature of UTF-8 that I've never tackled before?
This issue is not actually specific to UTF-8 at all. It happens with all encodings that can represent all (or at least most) Unicode codepoints.
The general idea of Unicode is to not provide so-called pre-composed characters (e.g. U+00E9 E ACUTE), instead they usually like to provide the base character (e.g. U+0065 LATIN SMALL LETTER E) and the combining character (e.g. U+0301 COMBINING ACUTE ACCENT). This has the advantage of not having to provide every possible combination as its own character.
Note: the U+xxxx notation is used to refer to unicode codepoints. It's the encoding-independent way to refer to Unicode characters.
However when Unicode was first designed an important goal was to have round-trip compatibility for existing, widely-used encodings, so some pre-composed characters were included (in fact most of the diacritic characters from the latin and related alphabets are included).
So yes (and tl;dr): in a correctly working Unicode-capable application U+00E9 should render the same way and be treated the same way as U+0065 followed by U+0301.
There's a non-trivial process called normalization that helps work with these differences by reducing a given string to one of four normal forms.
For example passing both strings (U+00E9 and U+0065 U+0301) will result in U+00E9 when using NFC and will result in U+0065 U+0301 when using NFD.
Very short and visualized example: the character "é" can either be represented using the Unicode code point U+00E9 (LATIN SMALL LETTER E WITH ACUTE, é), or the sequence U+0065 (LATIN SMALL LETTER E, e) followed by U+0301 (COMBINING ACUTE ACCENT, ´), which together look like this: é.
In UTF-8, é has the byte sequence xC3 xA9, while é has the byte sequence x65 xCC x81.
Note: Due to technical limitations this post does not contain the actual combination characters.
Actually I don't understand what it means by :
"Even though UTF-8 provides a single byte sequence for each character
sequence[...]"
What the quote wants to say is:
"Any given sequence of Unicode code points is mapped to one (and precisely one) sequence of bytes by the UTF-8 encoding." That is, UTF-8 is a bijection between sequences of (abstract) Unicode code points and bytes.
The problem, which the text wants to illustrate, is that there is no bijection between "letters of a text" (as commonly understood) and Unicode code points, because the same text can be represented by different sequences of Unicode code points (as explained in the example).
Actually, this has nothing to do with UTF-8 specifically; it is a fundamental property of Unicode: Many texts have more than one representations as Unicode code points. This is important to keep in mind when comparing texts expressed in Unicode (no matter in what encoding).
One (partial) solution to this is normalization. It defines various Normal forms for Unicode text, which are unique representations of a text.
Can anybody please tell me what is the range of Unicode printable characters? [e.g. Ascii printable character range is \u0020 - \u007f]
See, http://en.wikipedia.org/wiki/Unicode_control_characters
You might want to look especially at C0 and C1 control character http://en.wikipedia.org/wiki/C0_and_C1_control_codes
The wiki says, the C0 control character is in the range U+0000—U+001F and U+007F (which is the same range as ASCII) and C1 control character is in the range U+0080—U+009F
other than C-control character, Unicode also has hundreds of formatting control characters, e.g. zero-width non-joiner, which makes character spacing closer, or bidirectional text control. This formatting control characters are rather scattered.
More importantly, what are you doing that requires you to know Unicode's non-printable characters? More likely than not, whatever you're trying to do is the wrong approach to solve your problem.
This is an old question, but it is still valid and I think there is more to usefully, but briefly, say on the subject than is covered by existing answers.
Unicode
Unicode defines properties for characters.
One of these properties is "General Category" which has Major classes and subclasses. The Major classes are Letter, Mark, Punctuation, Symbol, Separator, and Other.
By knowing the properties of your characters, you can decide whether you consider them printable in your particular context.
You must always remember that terms like "character" and "printable" are often difficult and have interesting edge-cases.
Programming Language support
Some programming languages assist with this problem.
For example, the Go language has a "unicode" package which provides many useful Unicode-related functions including these two:
func IsGraphic(r rune) bool
IsGraphic reports whether the rune is defined as a Graphic by Unicode. Such
characters include letters, marks, numbers, punctuation, symbols, and spaces,
from categories L, M, N, P, S, Zs.
func IsPrint(r rune) bool
IsPrint reports whether the rune is defined as printable by Go. Such
characters include letters, marks, numbers, punctuation, symbols, and
the ASCII space character, from categories L, M, N, P, S and the ASCII
space character. This categorization is the same as IsGraphic except
that the only spacing character is ASCII space, U+0020.
Notice that it says "defined as printable by Go" not by "defined as printable by Unicode". It is almost as if there are some depths the wizards at Unicode dare not plumb.
Printable
The more you learn about Unicode, the more you realise how unexpectedly diverse and unfathomably weird human writing systems are.
In particular whether a particular "character" is printable is not always obvious.
Is a zero-width space printable? When is a hyphenation point printable? Are there characters whose printability depends on their position in a word or on what characters are adjacent to them? Is a combining-character always printable?
Footnotes
ASCII printable character range is \u0020 - \u007f
No it isn't. \u007f is DEL which is not normally considered a printable character. It is, for example, associated with the keyboard key labelled "DEL" whose earliest purpose was to command the deletion of a character from some medium (display, file etc).
In fact many 8-bit character sets have many non-consecutive ranges which are non-printable. See for example C0 and C1 controls.
First, you should remove the word 'UTF8' in your question, it's not pertinent (UTF8 is just one of the encodings of Unicode, it's something orthogonal to your question).
Second: the meaning of "printable/non printable" is less clear in Unicode. Perhaps you mean a "graphical character" ; and one can even dispute if a space is printable/graphical. The non-graphical characters would consist, basically, of control characters: the range 0x00-0x0f plus some others that are scattered.
Anyway, the vast majority of Unicode characters (more than 200.000) are "graphical". But this certainly does not imply that they are printable in your environment.
It seems to me a bad idea, if you intend to generate a "random printable" unicode string, to try to include all "printable" characters.
What you should do is pick a font, and then generate a list of which Unicode characters have glyphs defined for your font. You can use a font library like freetype to test glyphs (test for FT_Get_Char_Index(...) != 0).
Taking the opposite approach to #HoldOffHunger, it might be easier to list the ranges of non-printable characters, and use not to test if a character is printable.
In the style of Regex (so if you wanted printable characters, place a ^):
[\u0000-\u0008\u000B-\u001F\u007F-\u009F\u2000-\u200F\u2028-\u202F\u205F-\u206F\u3000\uFEFF]
Which accounts for things like separator spaces and joiners
Note that unlike their answer which is a whitelist that ignores all non-latin languages, this blacklist wont permit non-printable characters just because they're in blocks with printable characters (their answer wholly includes Non-Latin, Language Supplement blocks as 'printable', even though it contains things like 'zero-width non-joiner'..).
Be aware though, that if using this or any other solution, for sanitation for example, you may want to do something more nuanced than a blanket replace.
Arguably in that case, non-breaking spaces should change to space, not be removed, and invisible separator should be replaced with comma conditionally.
Then there's invalid character ranges, either [yet] unused or reserved for encoding purposes, and language-specific variation selectors..
NB when using regular expressions, that you enable unicode awareness if it isn't that way by default (for javascript it's via /.../u).
You can tell if you have it correct by attempting to create the regular expression with some multi-byte character ranges.
For example, the above, plus the invalid character range \u{E0100}-\u{E01EF} in javascript:
/[\u0000-\u0008\u000B-\u001F\u007F-\u009F\u2000-\u200F\u2028-\u202F\u205F-\u206F\u3000\uFEFF\u{E0100}-\u{E01EF}]/u
Without u \u{E0100}-\u{E01EF} equates to \uDB40(\uDD00-\uDB40)\uDDEF, not (\uDB40\uDD00)-(\uDB40\uDDEF), and if replacing you should always enable u even when not including multbyte unicode in the regex itself as you might break surrogate pairs that exist in the text.
What characters are valid?
At present, Unicode is defined as starting from U+0000 and ending at U+10FFFF. The first block, Basic Latin, spans U+0000 to U+007F and the last block, Supplementary Private Use Area-B, spans U+100000 to 10FFFF. If you want to see all of these blocks, see here: Wikipedia.org: Unicode Block; List of Blocks.
Let's break down what's valid/invalid in the Latin Block1.
The Latin Block: TLDR
If you're interested in filtering out either invisible characters, you'll want to filter out:
U+0000 to U+0008: Control
U+000E to U+001F: Device (i.e., Control)
U+007F: Delete (Control)
U+008D to U+009F: Device (i.e., Control)
The Latin Block: Full Ranges
Here's the Latin block, broken up into smaller sections...
U+0000 to U+0008: Control
U+0009 to U+000C: Space
U+000E to U+001F: Device (i.e., Control)
U+0020: Space
U+0021 to U+002F: Symbols
U+0030 to U+0039: Numbers
U+003A to U+0040: Symbols
U+0041 to U+005A: Uppercase Letters
U+005B to U+0060: Symbols
U+0061 to U+007A: Lowercase Letters
U+007B to U+007E: Symbols
U+007F: Delete (Control)
U+0080 to U+008C: Latin1-Supplement symbols.
U+008D to U+009F: Device (i.e., Control)
U+00A0: Non-breaking space. (i.e., )
U+00A1 to U+00BF: Symbols.
U+00C0 to U+00FF: Accented characters.
The Other Blocks
Unicode is famous for supporting non-Latin character sets, so what are these other blocks? This is just a broad overview, see the wikipedia.org page for the full, complete list.
Latin1 & Latin1-Related Blocks
U+0000 to U+007F : Basic Latin
U+0080 to U+00FF : Latin-1 Supplement
U+0100 to U+017F : Latin Extended-A
U+0180 to U+024F : Latin Extended-B
Combinable blocks
U+0250 to U+036F: 3 Blocks.
Non-Latin, Language blocks
U+0370 to U+1C7F: 55 Blocks.
Non-Latin, Language Supplement blocks
U+1C80 to U+209F: 11 Blocks.
Symbol blocks
U+20A0 to U+2BFF: 22 Blocks.
Ancient Language blocks
U+2C00 to U+2C5F: 1 Block (Glagolitic).
Language Extensions blocks
U+2C60 to U+FFEF: 66 Blocks.
Special blocks
U+FFF0 to U+FFFF: 1 Block (Specials).
One approach is to render each character to a texture and manually check if it is visible. This solution excludes spaces.
I've written such a program and used it to determine there are roughly 467241 printable characters within the first 471859 code points. I've selected this number because it covers all of the first 4 Planes of Unicode, which seem to contain all printable characters. See https://en.wikipedia.org/wiki/Plane_(Unicode)
I would much like to refine my program to produce the list of ranges, but for now here's what I am working with for anyone who needs immediate answers:
https://editor.p5js.org/SamyBencherif/sketches/_OE8Y3kS9
I am posting this tool because I think this question attracts a lot of people who are looking for slightly different applications of knowing printable ranges. Hopefully this is useful, even though it does not fully answer the question.
The printable Unicode character range, excluding the hex, is 32 to 126 in the int datatype.
Unicode, stict term, has no range. Numbers can go infinite.
What you gave is not UTF8 which has 1 byte for ASCII characters.
As for the range, I believe there is no range of printable characters. It always evolves. Check the page I gave above.