bug#31006: [PATCH] Handle quotation marks and apostrophes in ‘sgml-quote’

[Michal Nazarewicz]

To be able to use text in an HTML argument, quotation marks need
to be replaced with an appropriate character reference.  Make
‘sgml-quote’ do that.

While at it, fix entiteis not being unquoted if they lack closing
semicolon (e.g. ‘&amp’) occuring at the very end of a region.
Even though unlikely, make ‘sgml-quote’ handle this scenario.

* lisp/textmodes/sgml-mode.el (sgml-quote): Handle quotation marks and
apostrophes.  Match entities lacking semicolon at the end of regions.
* test/lisp/textmodes/sgml-mode-tests.el (sgml-quote-works): New test
case for ‘sgml-quote’ function.
---
 etc/NEWS                               |  6 ++++++
 lisp/textmodes/sgml-mode.el            | 26 +++++++++++++++++++-------
 test/lisp/textmodes/sgml-mode-tests.el | 30 ++++++++++++++++++++++++++++++
 3 files changed, 55 insertions(+), 7 deletions(-)

If everyone is fine with this I’m gonna push it in a few days.

diff --git a/etc/NEWS b/etc/NEWS
index 07f6d04a74..2da16e53fe 100644
--- a/etc/NEWS
+++ b/etc/NEWS
@@ -168,6 +168,12 @@ Can be controlled via the new variable 
'footnote-align-to-fn-text'.
 formats (e.g. "black" => "#000000" => "rgb(0, 0, 0)") has been added,
 bound to 'C-c C-f'.
 
+** SGML mode
+
+*** 'sgml-quote' now handles double quotes and apostrophes
+when escaping text and in addition all numeric entities when
+unescaping text.
+
 ** Dired
 
 +++
diff --git a/lisp/textmodes/sgml-mode.el b/lisp/textmodes/sgml-mode.el
index f6bdfc6384..52d14bd800 100644
--- a/lisp/textmodes/sgml-mode.el
+++ b/lisp/textmodes/sgml-mode.el
@@ -1241,8 +1241,11 @@ sgml-value
 
 (defun sgml-quote (start end &optional unquotep)
   "Quote SGML text in region START ... END.
-Only &, < and > are quoted, the rest is left untouched.
-With prefix argument UNQUOTEP, unquote the region."
+Only &, <, >, ' and \" characters are quoted, the rest is left
+untouched.  This is sufficient to use quoted text as SGML argument.
+
+With prefix argument UNQUOTEP, unquote the region.  All numeric entities,
+\"amp\", \"lt\", \"gt\" and \"quot\" named entities are unquoted."
   (interactive "r\nP")
   (save-restriction
     (narrow-to-region start end)
@@ -1250,14 +1253,23 @@ sgml-quote
     (if unquotep
        ;; FIXME: We should unquote other named character references as well.
        (while (re-search-forward
-               "\\(&\\(amp\\|\\(l\\|\\(g\\)\\)t\\)\\)[][<>&;\n\t \"%!'(),/=?]"
+               
"\\(&\\(amp\\|quot\\|lt\\|gt\\|#\\([0-9]+\\|[xX][0-9a-fA-F]+\\)\\)\\)\\([][<>&;\n\t
 \"%!'(),/=?]\\|$\\)"
                nil t)
-         (replace-match (if (match-end 4) ">" (if (match-end 3) "<" "&")) t t
-                        nil (if (eq (char-before (match-end 0)) ?\;) 0 1)))
-      (while (re-search-forward "[&<>]" nil t)
+          (replace-match
+           (string
+            (or (cdr (assq (char-after (match-beginning 2))
+                           '((?a . ?&) (?q . ?\") (?l . ?<) (?g . ?>))))
+                (let ((num (match-string 3)))
+                  (if (or (eq ?x (aref num 0)) (eq ?X (aref num 0)))
+                      (string-to-number (substring num 1) 16)
+                    (string-to-number num 10)))))
+           t t nil (if (eq (char-before (match-end 0)) ?\;) 0 1)))
+      (while (re-search-forward "[&<>\"']" nil t)
        (replace-match (cdr (assq (char-before) '((?& . "&amp;")
                                                  (?< . "&lt;")
-                                                 (?> . "&gt;"))))
+                                                 (?> . "&gt;")
+                                                  (?\" . "&#34;")
+                                                  (?' . "&#39;"))))
                       t t)))))
 
 (defun sgml-pretty-print (beg end)
diff --git a/test/lisp/textmodes/sgml-mode-tests.el 
b/test/lisp/textmodes/sgml-mode-tests.el
index 7ca6e676c6..6c0070ccb1 100644
--- a/test/lisp/textmodes/sgml-mode-tests.el
+++ b/test/lisp/textmodes/sgml-mode-tests.el
@@ -131,5 +131,35 @@ sgml-with-content
    (sgml-delete-tag 1)
    (should (string= "Winter is comin'" (buffer-string)))))
 
+(ert-deftest sgml-quote-works ()
+  (let ((text "Foo<Bar> \"Baz\" 'Qux'\n"))
+    (with-temp-buffer
+      ;; Back and forth transformation.
+      (insert text)
+      (sgml-quote (point-min) (point-max))
+      (should (string= "Foo&lt;Bar&gt; &#34;Baz&#34; &#39;Qux&#39;\n"
+                       (buffer-string)))
+      (sgml-quote (point-min) (point-max) t)
+      (should (string= text (buffer-string)))
+
+      ;; The same text escaped differently.
+      (erase-buffer)
+      (insert "Foo&lt;Bar&gt; &#34;Baz&quot; &#x27;Qux&#X27;\n")
+      (sgml-quote (point-min) (point-max) t)
+      (should (string= text (buffer-string)))
+
+      ;; Lack of semicolon.
+      (erase-buffer)
+      (insert "&amp&amp")
+      (sgml-quote (point-min) (point-max) t)
+      (should (string= "&&" (buffer-string)))
+
+      ;; Double quoting
+      (sgml-quote (point-min) (point-max))
+      (sgml-quote (point-min) (point-max))
+      (sgml-quote (point-min) (point-max) t)
+      (sgml-quote (point-min) (point-max) t)
+      (should (string= "&&" (buffer-string))))))
+
 (provide 'sgml-mode-tests)
 ;;; sgml-mode-tests.el ends here
-- 
2.16.2