GitLab

at 2024-05-13T14:23:47-07:00

Ignore some categories when finding the upper or lower case character

For `unicode-upper`, we ignore characters with the category Lt, Mn,
Nl, and So and just return the character unchanged.  Likewise, for
`unicode-lower`, we ignore Lt, Nl, and So.

@@ -41,7 +41,7 @@ else
  fi
  cd ../ansi-test
 -git checkout issue-288-new-failures
 +git checkout issue-316-support-roundtrip-char-casing
  make LISP="$LISP batch -noinit -nositeinit"
  # There should be no unexpected successes or failures; check these separately

@@ -883,7 +883,9 @@
    (unless (unidata-scase *unicode-data*) (load-scase))
    (let* ((scase (unidata-scase *unicode-data*))
  	 (n (logand (qref32 scase code) #xFF)))
 -    (if (zerop n)
 +    (if (or (zerop n)
 +            ;; Ignore category Lt, Mn, Nl, So
 +            (member (unicode-category code) '(92 32 75 109)))
  	code
  	(let* ((m (aref (scase-svec scase) (logand n #x7F))))
  	  (if (logbitp 7 n) (+ code m) (- code m))))))
@@ -894,7 +896,9 @@
    (unless (unidata-scase *unicode-data*) (load-scase))
    (let* ((scase (unidata-scase *unicode-data*))
  	 (n (logand (ash (qref32 scase code) -8) #xFF)))
 -    (if (zerop n)
 +    (if (or (zerop n)
 +            ;; Ignore category Lt, Nl, So
 +            (member (unicode-category code) '(92 75 109)))
  	code
  	(let ((m (aref (scase-svec scase) (logand n #x7F))))
  	  (if (logbitp 7 n) (+ code m) (- code m))))))

Raymond Toy pushed to branch issue-316-support-roundtrip-char-casing at cmucl / cmucl

Commits:

2 changed files:

Changes: