Define \pN One-letter name Unicode character class with reference to unicode spec.

[nathan-at-least]

What version of regex are you using?

1.7.3

Describe the bug at a high level.

The crate docs under Matching one character for syntax defines:

\pN           One-letter name Unicode character class

However, I cannot find a definition for "name Unicode character class" anywhere in the documentation, or through web searches.

Lower in Unicode features, certain features link to Unicode specifications, which is helpful, but from these I still cannot determine what "name character class" is. For example the unicode-gencat feature enables category-specification in patterns which correspond to this link Unicode general categories which does not have a "name" category.

What are the steps to reproduce the behavior?

Read the documentation and try to define what "name Unicode character class" is.

What is the actual behavior?

The specific definition is not found.

What is the expected behavior?

A definition would be provided, such as:

The name Unicode character class is specified in <link>

-or if it's an equivalent to another pattern, that could also be provided. For example, my (unverified) guess is that the name category might be equivalent to the Unicode Letter category:

The \Pn pattern is equivalent to \P{Letter}

[BurntSushi]

The crate docs link to UNICODE.md which I believe answers your question here.

Or are you specifically asking for a link to all the possible single level property values? Because if so, I don't think such a link exists. That's contained inside the Unicode Character Database. With that said, at this specific moment in time, you can see the list of categories and their abbreviations. Abbreviations with a single letter can be written in \pX syntax:

regex/regex-syntax/src/unicode_tables/property_values.rs

Lines 71 to 154 in 9582040

	"General_Category",
	&[
	("c", "Other"),
	("casedletter", "Cased_Letter"),
	("cc", "Control"),
	("cf", "Format"),
	("closepunctuation", "Close_Punctuation"),
	("cn", "Unassigned"),
	("cntrl", "Control"),
	("co", "Private_Use"),
	("combiningmark", "Mark"),
	("connectorpunctuation", "Connector_Punctuation"),
	("control", "Control"),
	("cs", "Surrogate"),
	("currencysymbol", "Currency_Symbol"),
	("dashpunctuation", "Dash_Punctuation"),
	("decimalnumber", "Decimal_Number"),
	("digit", "Decimal_Number"),
	("enclosingmark", "Enclosing_Mark"),
	("finalpunctuation", "Final_Punctuation"),
	("format", "Format"),
	("initialpunctuation", "Initial_Punctuation"),
	("l", "Letter"),
	("lc", "Cased_Letter"),
	("letter", "Letter"),
	("letternumber", "Letter_Number"),
	("lineseparator", "Line_Separator"),
	("ll", "Lowercase_Letter"),
	("lm", "Modifier_Letter"),
	("lo", "Other_Letter"),
	("lowercaseletter", "Lowercase_Letter"),
	("lt", "Titlecase_Letter"),
	("lu", "Uppercase_Letter"),
	("m", "Mark"),
	("mark", "Mark"),
	("mathsymbol", "Math_Symbol"),
	("mc", "Spacing_Mark"),
	("me", "Enclosing_Mark"),
	("mn", "Nonspacing_Mark"),
	("modifierletter", "Modifier_Letter"),
	("modifiersymbol", "Modifier_Symbol"),
	("n", "Number"),
	("nd", "Decimal_Number"),
	("nl", "Letter_Number"),
	("no", "Other_Number"),
	("nonspacingmark", "Nonspacing_Mark"),
	("number", "Number"),
	("openpunctuation", "Open_Punctuation"),
	("other", "Other"),
	("otherletter", "Other_Letter"),
	("othernumber", "Other_Number"),
	("otherpunctuation", "Other_Punctuation"),
	("othersymbol", "Other_Symbol"),
	("p", "Punctuation"),
	("paragraphseparator", "Paragraph_Separator"),
	("pc", "Connector_Punctuation"),
	("pd", "Dash_Punctuation"),
	("pe", "Close_Punctuation"),
	("pf", "Final_Punctuation"),
	("pi", "Initial_Punctuation"),
	("po", "Other_Punctuation"),
	("privateuse", "Private_Use"),
	("ps", "Open_Punctuation"),
	("punct", "Punctuation"),
	("punctuation", "Punctuation"),
	("s", "Symbol"),
	("sc", "Currency_Symbol"),
	("separator", "Separator"),
	("sk", "Modifier_Symbol"),
	("sm", "Math_Symbol"),
	("so", "Other_Symbol"),
	("spaceseparator", "Space_Separator"),
	("spacingmark", "Spacing_Mark"),
	("surrogate", "Surrogate"),
	("symbol", "Symbol"),
	("titlecaseletter", "Titlecase_Letter"),
	("unassigned", "Unassigned"),
	("uppercaseletter", "Uppercase_Letter"),
	("z", "Separator"),
	("zl", "Line_Separator"),
	("zp", "Paragraph_Separator"),
	("zs", "Space_Separator"),
	],
	),

[BurntSushi]

which correspond to this link Unicode general categories which does not have a "name" category.

This actually looks like a good link to me. It lists out all the general categories and even clearly shows which ones have one-letter abbreviations. I'm not understanding what has confused you here?

Although, there may be scripts that have one-letter abbreviations too, in which case, those work in \pX syntax too.

[nathan-at-least]

I'm still confused what this means:

\pN One-letter name Unicode character class

Does this mean a pattern can literally include \pN to match a "name Unicode character class"?

Based on your comments, I'm wondering if instead it means patterns can contain specifiers like \pc' to mean the Otherclass, or\pnto mean theNumberclass, etc… Is that the case? That makes sense, but I didn't realizeN` was a syntactic variable! I'm not sure how to clarify in that case.

[BurntSushi]

I still don't understand your confusion. \pL is precisely equivalent to \p{Letter}. It is a short-hand. There's nothing else to it. The set of \pN things that exist correspond to the things that have a single letter abbreviation, as defined by Unicode.

[Erutuon]

It sounds like there's a problem with parsing "One-letter name Unicode character class", a long series of noun phrases where the relationships aren't made clear by function words. It should be parsed as [one-letter name] [Unicode character class], but it sounds like @nathan-at-least is parsing it as [one-letter] [name [Unicode character class]]. It might be easier to understand if it were rephrased as "Unicode character class identified by a one-letter name" or something.

[BurntSushi]

Ah I see okay, I have a fix incoming for this using your wording @Erutuon. Thanks!

[nathan-at-least]

Thanks! The clarification looks like it would be less confusing to past-me. Kind of embarrassing after I understood the semantics, but I guess that's the nature of docs.

[BurntSushi]

@nathan-at-least No worries! It's great feedback. The docs are better for it. :-)