wc: Do a chunked read with proper UTF-8 handling

This brings the results mostly in line with GNU wc and solves nasty
behavior with long lines.

tests/by-util/test_wc.rs

@@ -53,11 +53,16 @@ fn test_utf8() {
         .args(&["-lwmcL"])
         .pipe_in_fixture("UTF_8_test.txt")
         .run()
-        .stdout_is("    300    4969   22781   22213      79\n");
-    // GNU returns "    300    2086   22219   22781      79"
-    //
-    // TODO: we should fix the word, character, and byte count to
-    // match the behavior of GNU wc
+        .stdout_is("    303    2119   23025   22457      79\n");
+}
+
+#[test]
+fn test_utf8_extra() {
+    new_ucmd!()
+        .arg("-lwmcL")
+        .pipe_in_fixture("UTF_8_weirdchars.txt")
+        .run()
+        .stdout_is("     25      87     513     442      48\n");
 }
 
 #[test]

tests/fixtures/wc/UTF_8_weirdchars.txt

@@ -0,0 +1,25 @@
+zero-width space inbetween these: x​x
+and inbetween two spaces: [ ​ ]
+and at the end of the line: ​
+
+non-breaking space: x x [   ]  
+
+simple unicode: xµx [ µ ] µ
+
+wide: xｗx [ ｗ ] ｗ
+
+simple emoji: x👩x [ 👩 ] 👩
+
+complex emoji: x👩‍🔬x [ 👩‍🔬 ] 👩‍🔬
+
+Ｈｅｌｌｏ, ｗｏｒｌｄ!
+
+line feed: xx [  ] 
+
+vertical tab: xx [  ] 
+
+horizontal tab: x	x [ 	 ]
+this should be the longest line:
+1234567	12345678	123456781234567812345678
+
+Control character: xx [  ]