From cb936db78ecadbe4208f593921aaa21d9675a368 Mon Sep 17 00:00:00 2001
From: Vladimir Sedach <vas@oneofus.la>
Date: Sun, 21 Oct 2018 02:06:28 -0700
Subject: [PATCH] Better output of Unicode strings

---
 src/printer.lisp        | 38 +++++++++++++++++++-------------------
 tests/output-tests.lisp | 14 +++++++++++---
 2 files changed, 30 insertions(+), 22 deletions(-)

diff --git a/src/printer.lisp b/src/printer.lisp
index c53c0eb..3637d6f 100644
--- a/src/printer.lisp
+++ b/src/printer.lisp
@@ -158,30 +158,30 @@ vice-versa.")
       (newline-and-indent))))
 
 (defparameter *js-lisp-escaped-chars*
-  '((#\' . #\')
-    (#\" . #\")
-    (#\\ . #\\)
-    (#\b . #\Backspace)
-    (#\f . #.(code-char 12))
-    (#\n . #\Newline)
-    (#\r . #\Return)
-    (#\t . #\Tab)))
+  (list #\'            #\'
+        #\"            #\"
+        #\\            #\\
+        #\Backspace    #\b
+        (code-char 12) #\f
+        #\Newline      #\n
+        #\Return       #\r
+        #\Tab          #\t))
 
 (defmethod ps-print ((char character))
   (ps-print (string char)))
 
 (defmethod ps-print ((string string))
-  (flet ((lisp-special-char-to-js (lisp-char)
-           (car (rassoc lisp-char *js-lisp-escaped-chars*))))
-    (psw *js-string-delimiter*)
-    (loop for char across string
-          for code = (char-code char)
-          for special = (lisp-special-char-to-js char)
-          do (cond (special (psw #\\) (psw special))
-                   ((or (<= code #x1f) (>= code #x80))
-                    (format *psw-stream* "\\u~:@(~4,'0x~)" code))
-                   (t (psw char))))
-    (psw *js-string-delimiter*)))
+  (psw *js-string-delimiter*)
+  (loop for char across string do
+       (acond ((getf *js-lisp-escaped-chars* char)
+               (psw #\\ it))
+              ((or (<= (char-code char) #x1F)
+                   (<= #x80 (char-code char) #x9F)
+                   (member (char-code char) '(#xA0 #xAD #x200B #x200C)))
+               (format *psw-stream* "\\u~:@(~4,'0x~)" (char-code char)))
+              (t
+               (psw char))))
+  (psw *js-string-delimiter*))
 
 (defmethod ps-print ((number number))
   (format *psw-stream* (if (integerp number) "~D" "~F") number))
diff --git a/tests/output-tests.lisp b/tests/output-tests.lisp
index 57c7cab..93129f9 100644
--- a/tests/output-tests.lisp
+++ b/tests/output-tests.lisp
@@ -1019,15 +1019,19 @@ return element.innerHTML = ['<textarea', disabled || !authorized ? [' disabled=\
          `((#\\     . #\\)
            (#\b     . #\Backspace)
            (#\f     . ,(code-char 12))
-           ("u000B" . ,(code-char #x000b))    ; vertical tab
+           ("u000B" . ,(code-char #xB))       ; vertical tab
            (#\n     . #\Newline)
            (#\r     . #\Return)
            (#\'     . #\')
            (#\"     . #\")
            (#\t     . #\Tab)
-           ("u001F" . ,(code-char #x001f))    ; character below 32
+           ("u001F" . ,(code-char #x1F))      ; character below 32
            ("u0080" . ,(code-char 128))       ; character over 127
-           ("uABCD" . ,(code-char #xabcd))))) ; well above ASCII
+           ("u00A0" . ,(code-char 160))       ; non-breaking space
+           ("u00AD" . ,(code-char 173))       ; soft hyphen
+           ("u200B" . ,(code-char #x200B))    ; zero-width space
+           ("u200C" . ,(code-char #x200C))    ; zero-width non-joiner
+           )))
     (loop for (js-escape . lisp-char) in escapes
           for generated = (ps-doc* (format nil "hello~ahi" lisp-char))
           for wanted = (format nil "'hello\\~ahi';" js-escape)
@@ -4121,6 +4125,10 @@ function foo(x) {
     return { 'abc' : (x = obj['blah'], x ? 123 : 456) };
 };")
 
+(test-ps-js unicode-strings
+  "фоо бар"
+  "'фоо бар';")
+
 ;;; broken
 
 ;; (test-ps-js let-defun-toplevel
-- 
2.11.4.GIT