regex: implement the short \N escape sequence from Perl/PCRE

With `.`, Kakoune matches newlines by default. Therefore
`\N` comes in very handy, as it is quicker to type than
the commonly used patterns `(?S).` and `[^\n]`.

Author: jo-he

doc/pages/changelog.asciidoc

@@ -16,6 +16,8 @@ released versions.
 
 * `number-lines -full-relative` switch to keep a smaller line number gutter.
 
+* Support the `\N` escape sequence in regex (like in PCRE, matches `[^\n]`).
+
 == Kakoune 2025.06.03
 
 * Expose env vars that are mentioned in the arguments passed to shell expansions

doc/pages/faq.asciidoc

@@ -292,8 +292,8 @@ In most distributions, `/bin/sh` will end up being used.
 
 Data in buffers is a stream of characters, and newlines do not receive special
 treatment compared to other characters, with regards to regex matching. In
-order to select data in a line without any trailing newline characters, one could
-use the `[^\n]+` pattern, which is arguably a good compromise when
+order to select data in a line without any trailing newline characters, one
+could use the `\N+` pattern, which is arguably a good compromise when
 balanced against the ability to select data over several lines.
 
 You can instruct the regex engine to stop matching newline characters with

doc/pages/highlighters.asciidoc

@@ -131,9 +131,9 @@ highlighter is replaced with the new one.
     This highlights C++ style comments in cyan, with an eventual 'TODO:'
     in yellow on red background:
 
---------------------------------------------------------------------
-add-highlighter window/ regex //\h*(TODO:)[^\n]* 0:cyan 1:yellow,red
---------------------------------------------------------------------
+-----------------------------------------------------------------
+add-highlighter window/ regex //\h*(TODO:)\N* 0:cyan 1:yellow,red
+-----------------------------------------------------------------
 
     capture_id can be either the capture number, or its name if a
     named capture is used in the regex (See

doc/pages/regex.asciidoc

@@ -74,6 +74,8 @@ is equivalent to `[\d]`.
 `.` matches any character, including newlines, by default.
 (see <<regex#modifiers,:doc regex modifiers>> on how to change it)
 
+`\N` matches any character but newlines (not affected by modifiers).
+
 == Groups
 
 Regex atoms can be grouped using `(` and `)` or `(?:` and `)`. If `(` is
@@ -184,7 +186,7 @@ by <https://www.ecma-international.org/ecma-262/8.0/>; some divergence
 exists for ease of use, or performance reasons:
 
 * Lookarounds are not arbitrary, but lookbehind is supported.
-* `\K`, `\Q..\E`, `\A`, `\h` and `\z` are added.
+* `\K`, `\N`, `\Q..\E`, `\A`, `\h` and `\z` are added.
 * Stricter handling of escaping, as we introduce additional escapes;
   identity escapes like `\X` with `X` being a non-special character
   are not accepted, to avoid confusions between `\h` meaning literal

src/main.cc

@@ -55,6 +55,7 @@ struct {
         "» {+b}%val\\{buffile}{} is now empty for scratch buffers\n"
         "» {+b}FocusIn{}/{+b}FocusOut{} events on suspend\n"
         "» {+u}number-lines -full-relative{} switch to keep a smaller line number gutter\n"
+        "» support the {+b}\\N{} escape sequence in regex (matches {+b}[^\\n]{})\n"
     }, {
         20250603,
         "» kak_* appearing in shell arguments will be added to the environment\n"

src/regex_vm.cc

@@ -357,6 +357,9 @@ struct RegexParser
     {
         const Codepoint cp = *m_pos++;
 
+        if (cp == 'N')
+            return add_node(ParsedRegex::AnyCharExceptNewLine);
+
         if (cp == 'Q')
         {
             auto escaped_sequence = add_node(ParsedRegex::Sequence);
@@ -1562,6 +1565,12 @@ auto test_regex = UnitTest{[]{
         kak_assert(*vm.captures()[0] == 'f');
     }
 
+    {
+        TestVM<> vm{R"(a(?<=\N)\N+(?=.\N)\s(?S)d.+(?!.)\s(?<!\N)g)"};
+        kak_assert(vm.exec("abc\ndef\ng"));
+        kak_assert(not vm.exec("abc\ndef g"));
+    }
+
     {
         TestVM<> vm{R"([d-ea-dcf-k]+)"};
         kak_assert(vm.exec("abcde"));