Add textsec-domain-suspicious-p

* .gitignore: Ignore idna-mapping.el.

* admin/notes/unicode: Note idna-mapping file.
* admin/unidata/IdnaMappingTable.txt: New file.

* admin/unidata/Makefile.in (all): Generate idna-mapping.el.

* admin/unidata/unidata-gen.el (unidata-gen-idna-mapping): Generate.

* lisp/international/textsec.el (textsec-domain-suspicious-p): New
function.

.gitignore

@@ -221,6 +221,7 @@ lisp/international/emoji-zwj.el
 lisp/international/emoji-labels.el
 lisp/international/eucjp-ms.el
 lisp/international/uni-*.el
+lisp/international/idna-mapping.el
 lisp/language/pinyin.el
 
 # Documentation.

admin/notes/unicode

@@ -26,13 +26,15 @@ Emacs uses the following files from the Unicode Character Database
   . BidiCharacterTest.txt
 
 Emacs also uses the file emoji-test.txt which should be imported from
-the Unicode's Public/emoji/ directory.
+the Unicode's Public/emoji/ directory, and IdnaMappingTable.txt from
+the Public/idna/ directory.
 
-First, the first 14 files and emoji-test.txt need to be copied into
+First, the first 14 files, emoji-test.txt and IdnaMappingTable.txt
-admin/unidata/, and the file https://www.unicode.org/copyright.html
+need to be copied into admin/unidata/, and the file
-should be copied over copyright.html in admin/unidata (some of them
+https://www.unicode.org/copyright.html should be copied over
-might need trailing whitespace removed before they can be committed to
+copyright.html in admin/unidata (some of them might need trailing
-the Emacs repository).
+whitespace removed before they can be committed to the Emacs
+repository).
 
 Then Emacs should be rebuilt for them to take effect.  Rebuilding
 Emacs updates several derived files elsewhere in the Emacs source

admin/unidata/Makefile.in

@@ -45,7 +45,8 @@ unifiles = $(addprefix ${unidir}/,$(sort $(shell sed -n 's/^[ \t][ \t]*${lparen}
 
 all: ${top_srcdir}/src/macuvs.h ${unifiles} ${unidir}/charscript.el \
   ${unidir}/charprop.el ${unidir}/emoji-zwj.el ${unidir}/emoji-labels.el \
-  ${unidir}/uni-scripts.el ${unidir}/uni-confusable.el
+  ${unidir}/uni-scripts.el ${unidir}/uni-confusable.el \
+  ${unidir}/idna-mapping.el
 
 ## Specify .elc as an order-only prereq so as to not needlessly rebuild
 ## target just because the .elc is missing.
@@ -95,6 +96,11 @@ ${unidir}/uni-confusable.el: ${srcdir}/unidata-gen.el \
 	$(AM_V_GEN)${emacs} -L ${srcdir} \
 		-l unidata-gen.el -f unidata-gen-confusable $@
 
+${unidir}/idna-mapping.el: ${srcdir}/unidata-gen.el \
+	  ${srcdir}/IdnaMappingTable.txt
+	$(AM_V_GEN)${emacs} -L ${srcdir} \
+		-l unidata-gen.el -f unidata-gen-idna-mapping $@
+
 .PHONY: charscript.el
 charscript.el: ${unidir}/charscript.el
 

admin/unidata/README

@@ -61,6 +61,6 @@ PropertyValueAliases.txt
 https://www.unicode.org/Public/UCD/latest/ucd/PropertyValueAliases.txt
 2022-01-17
 
-confusables.txt
+IdnaMappingTable.txt
-https://www.unicode.org/Public/security/latest/confusables.txt
+https://www.unicode.org/Public/idna/latest/IdnaMappingTable.txt
 2022-01-18

admin/unidata/unidata-gen.el

@@ -1602,6 +1602,46 @@ Property value is a symbol `o' (Open), `c' (Close), or `n' (None)."
       (insert ")")
       (unidata-gen-charprop file (buffer-string)))))
 
+(defun unidata-gen-idna-mapping (&optional file)
+  ;; Running from Makefile.
+  (unless file
+    (setq file (pop command-line-args-left)))
+  (let ((map (make-char-table nil)))
+    (with-temp-buffer
+      (unidata-gen--insert-file "IdnaMappingTable.txt")
+      (while (re-search-forward "^\\([0-9A-F]+\\)\\(?:\\.\\.\\([0-9A-F]+\\)\\)? +; +\\([^ ]+\\) +\\(?:; +\\([ 0-9A-F]+\\)\\)?"
+                                nil t)
+        (let ((start (match-string 1))
+              (end (match-string 2))
+              (status (match-string 3))
+              (mapped (match-string 4)))
+          ;; Make reading the file slightly faster by using `t'
+          ;; instead of `disallowed' all over the place.
+          (when (string-match-p "\\`disallowed" status)
+            (setq status "t"))
+          (unless (or (equal status "valid")
+                      (equal status "deviation"))
+            (set-char-table-range
+             map
+             (if end
+                 (cons (string-to-number start 16)
+                       (string-to-number end 16))
+               (string-to-number start 16))
+             (cond
+              ((equal status "mapped")
+               (apply #'string
+                      (mapcar (lambda (char)
+                                (string-to-number char 16))
+                              (split-string (string-trim mapped)))))
+              (t
+               (intern status))))))))
+    (with-temp-buffer
+      (insert "(defconst idna-mapping-table\n")
+      (let ((print-length nil))
+        (prin1 map (current-buffer)))
+      (insert ")")
+      (unidata-gen-charprop file (buffer-string)))))
+
 
 
 ;;; unidata-gen.el ends here

lisp/international/textsec.el

@@ -26,6 +26,8 @@
 (require 'cl-lib)
 (require 'uni-confusable)
 (require 'ucs-normalize)
+(require 'idna-mapping)
+(require 'puny)
 
 (defvar textsec--char-scripts nil)
 
@@ -222,6 +224,18 @@ STRING isn't a single script string."
        (textsec-single-script-p string1)
        (textsec-single-script-p string2)))
 
+(defun textsec-domain-suspicious-p (domain)
+  (catch 'found
+    (seq-do
+     (lambda (char)
+       (when (eq (elt idna-mapping-table char) t)
+         (throw 'found (format "Disallowed character: `%s' (#x%x)"
+                               (string char) char))))
+     domain)
+    (unless (puny-highly-restrictive-domain-p domain)
+      (throw 'found "%s is not highly restrictive"))
+    nil))
+
 (provide 'textsec)
 
 ;;; textsec.el ends here

test/lisp/international/textsec-tests.el

@@ -110,4 +110,9 @@
   (should-not (textsec-whole-script-confusable-p "paypal" "pаypаl"))
   (should (textsec-whole-script-confusable-p "scope""ѕсоре")))
 
+(ert-deftest test-suspiction-domain ()
+  (should (textsec-domain-suspicious-p "foo/bar.org"))
+  (should-not (textsec-domain-suspicious-p "foo.org"))
+  (should (textsec-domain-suspicious-p "f\N{LEFT-TO-RIGHT ISOLATE}oo.org")))
+
 ;;; textsec-tests.el ends here

test/lisp/net/puny-tests.el

@@ -68,4 +68,10 @@
                  "xn--b.com-gra"))
   (should (equal (puny-encode-string "Bä.com") "xn--b.com-gra")))
 
+;;; TODO!
+;; puny-resources/IdnaTestV2.txt has a bunch of tests, and they should
+;; be implemented.  However, the puny encoding does not fully
+;; implement https://www.unicode.org/reports/tr46/#Conformance yet, so
+;; it'll fail.
+
 ;;; puny-tests.el ends here