[nongnu] elpa/adoc-mode 63abada690 074/199: test: adoctest-trans learned setting mark using <>

[ELPA Syncer]

branch: elpa/adoc-mode
commit 63abada690f68448de71a9078097f9895e63da0d
Author: Florian Kaufmann <sensorflo@gmail.com>
Commit: Florian Kaufmann <sensorflo@gmail.com>

    test: adoctest-trans learned setting mark using <>
---
 adoc-mode-test.el | 48 ++++++++++++++++++++++++++++++++++++------------
 1 file changed, 36 insertions(+), 12 deletions(-)

diff --git a/adoc-mode-test.el b/adoc-mode-test.el
index 91a31d17d8..b4259323db 100644
--- a/adoc-mode-test.el
+++ b/adoc-mode-test.el
@@ -58,24 +58,44 @@
       (kill-buffer buf-name))))
 
 (defun adoctest-trans (original-text expected-text transform)
-  (if (string-match "!" original-text)
+  "Calling TRANSFORM on EXPECTED-TEXT, ORIGINAL-TEXT `should' result.
+ORIGINAL-TEXT is put in an temporary buffer and TRANSFORM is
+evaluated using `eval'. The resulting buffer content is compared
+to EXPECTED-TEXT.
+
+ORIGINAL-TEXT optionaly may contain the following special
+charachters. Escaping them is not (yet) supported. They are
+removed before TRANSFORM is evaluated.
+
+!  Position of point before TRANSFORM is evaluated
+
+<> Position of mark (<) and point (>) before TRANSFORM is
+   evaluatred"
+  (if (string-match "[!<>]" original-text)
       ;; original-text has ! markers
-      (let ((pos 0)
-           (pos-old 0)
-           (pos-list)
-           (new-original-text ""))
+      (let ((pos 0)      ; pos in original-text
+           (pos-old 0)  ; pos of the last iteration
+           (pos-in-new-region-start 0)
+           (pos-new-list) ; list of positions in new-original-text
+           (new-original-text "")) ; as original-text, but with < > ! stripped
        ;; original-text -> new-original-text by removing ! and remembering 
their positions
        (while (and (< pos (length original-text))
-                   (setq pos (string-match "!" original-text pos)))
+                   (setq pos (string-match "[!<>]" original-text pos)))
          (setq new-original-text (concat new-original-text (substring 
original-text pos-old pos)))
-         (setq pos-list (cons (length new-original-text) pos-list))
+         (cond
+          ((eq (aref original-text pos) ?<)
+           (setq pos-in-new-region-start (length new-original-text)))
+          ((eq (aref original-text pos) ?>)
+           (setq pos-new-list (cons (cons pos-in-new-region-start (length 
new-original-text)) pos-new-list)))
+          (t
+           (setq pos-new-list (cons (length new-original-text) pos-new-list))))
          (setq pos (1+ pos))
          (setq pos-old pos))
        (setq new-original-text (concat new-original-text (substring 
original-text pos-old pos)))
        ;; run adoctest-trans-inner for each remembered pos 
-       (while pos-list
-         (adoctest-trans-inner new-original-text expected-text transform (car 
pos-list))
-         (setq pos-list (cdr pos-list))))
+       (while pos-new-list
+         (adoctest-trans-inner new-original-text expected-text transform (car 
pos-new-list))
+         (setq pos-new-list (cdr pos-new-list))))
     ;; original-text has no ! markers
     (adoctest-trans-inner original-text expected-text transform)))
 
@@ -89,8 +109,12 @@
          (delete-region (point-min) (point-max))
          (adoc-mode)
          (insert original-text)
-         (when pos
-           (goto-char (1+ pos)))       ; buffer pos starts at 1, but string 
pos at 0
+         (cond ; 1+: buffer pos starts at 1, but string pos at 0
+          ((consp pos)
+           (goto-char (1+ (car pos)))
+           (set-mark (1+ (cdr pos))))
+          (pos
+           (goto-char (1+ pos)))) 
          ;; exercise
          (eval transform)
          ;; verify