Unicode display width (#5)

README.md

@@ -39,6 +39,7 @@ assert "The quIck? (“bRown”) fox cAn’T jUMp 32.3 feet, rIGHt?".toLower ==
 assert cmpCaseless("AbCσ", "aBcΣ")
 assert toValidUtf8("a\xffb", replacement = "") == "ab"
 assert verifyUtf8("a\xffb") == 1
+assert width("이건 테스트야", cjk=true) == 13
 ```
 
 [docs](https://nitely.github.io/nim-unicodeplus/)

src/unicodeplus.nim

@@ -1,14 +1,18 @@
 ## This module provides common unicode operations
 
 from std/unicode import
-  Rune, runes, `==`, fastRuneAt,
-  fastToUtf8Copy
+  Rune, runes, `==`, fastRuneAt, fastToUtf8Copy
 
 import pkg/unicodedb/properties
 import pkg/unicodedb/types
 import pkg/unicodedb/casing
+import pkg/unicodedb/widths
 import pkg/segmentation
 
+const enableWidth = (NimMajor, NimMinor) >= (2, 0)
+when enableWidth:
+  import pkg/graphemes
+
 export Rune
 
 type
@@ -460,3 +464,35 @@ func toValidUtf8*(s: string, replacement = "\uFFFD"): string =
       oldLen = result.len
     i += j+1
   result.add2 toOpenArray(s, i, s.len-1)
+
+when enableWidth:
+  func width(c: Rune, cjk: bool): int =
+    if c.int == 0xFE0F:  # emoji style
+      return 2
+    result = case c.unicodeWidth()
+      of uwdtFull, uwdtWide: 2
+      of uwdtAmbiguous:
+        if cjk: 1 else: 2
+      else: 1
+
+  func width*(s: string, cjk = false): int =
+    ## Return the display width of `s`.
+    ## This is usually correct for monospaced fonts,
+    ## but it may not be accurate in some cases.
+    ## Requires Nim >= 2.
+    result = 0
+    var i = 0
+    var i2 = 0
+    var c = Rune(0)
+    var w = 0
+    for bounds in s.graphemeBounds:
+      i = 0
+      i2 = -1
+      c = Rune(0)
+      w = 0
+      while i < bounds.b-bounds.a+1:
+        doAssert i > i2, "invalid utf8"
+        i2 = i
+        fastRuneAt(toOpenArray(s, bounds.a, bounds.b), i, c)
+        w = max(w, c.width(cjk))
+      result += w

tests/tests.nim

@@ -1,10 +1,10 @@
-import unittest
-from unicode import Rune, toUTF8
-from strutils import contains
+import std/unittest
+from std/unicode import Rune, toUTF8
+from std/strutils import contains
 
-from unicodedb/casing import caseFold
+from pkg/unicodedb/casing import caseFold
 
-import unicodeplus
+import ../src/unicodeplus
 
 when (NimMajor, NimMinor) >= (2, 0):
   type MyAssertionDefect = ref AssertionDefect
@@ -582,3 +582,15 @@ when true:
     check toValidUtf8("\xC2abc") == "\uFFFDabc"
     check toValidUtf8("\xF0\x80\x80") == "\uFFFD"
     check toValidUtf8("\x80\x80") == "\uFFFD"
+
+when (NimMajor, NimMinor) >= (2, 0):
+  test "width":
+    doAssert width("이건 테스트야", cjk=true) == 13
+    doAssert width("🐤node お名前=☜(ﾟヮﾟ☜)") == 22
+    doAssert width("👨‍👩‍👧‍👦") == 2
+    doAssert width("👨‍👩‍👧‍👦👨‍👩‍👧‍👦") == 4
+    doAssert width("👨‍👩‍👧‍👦🥱🧛🏻‍♂️") == 6
+    doAssert width("u̲n̲d̲e̲r̲l̲i̲n̲e̲d̲") == 20
+    doAssert width("a") == 1
+    doAssert width("ab") == 2
+    doAssert width("abc") == 3

unicodeplus.nimble

@@ -11,6 +11,7 @@ srcDir = "src"
 requires "nim >= 1.0"
 requires "unicodedb >= 0.8"
 requires "segmentation >= 0.1"
+requires "graphemes >= 0.12"
 
 task test, "Test":
   exec "nim c -r tests/tests"