doublep / iter2 Goto Github PK

Since your work is a faster, fully compatible version of generator.el, wouldn't it be a good idea to just replace generator by iter2? This way everyone benefits from it, and we don't have to duplicate packages which already rely on generator.el.

(I understand if you just say "It's too much work to interact with the Emacs mailing list", seems reasonable!)

To reproduce, on master, eval:

(iter2-defun f ()
  (dolist (e '(1 2 3))
    (iter-yield e)))

which gives error wrong-type-argument listp $value. Fair enough.

Mysteriously, however, macroexpanding the above is fine:

(macroexpand '(iter2-defun f ()
  (dolist (e '(1 2 3))
    (iter-yield e))))

gives no error.

I don't know how to start fixing this, since macroexpand works just fine.

When running code that uses save-restriction under ert, in certain circumstances an error is encountered. I don't have it at hand right now, but essentially, save-restriction attempts to restore point-max in the *ert* buffer, not the buffer I created for the test and in which I originally ran save-restriction, and that fails because my buffer is larger than the *ert* buffer.

In general, it seems like internally save-restriction, save-excursion, save-current-buffer are implemented essentially using unwind-protect (example). Basically, it seems like save-current-buffer just saves the current buffer, save-excursion saves instead a marker to the current point in the current buffer (as well as the current window), and save-restriction saves markers to the min and max points in the current buffer. iter2's save-restriction, however, does not restore the current buffer from the markers before calling narrow-to-region, which I believe causes the bug.

In general, it may be a better idea to manually convert these 3 macros to equivalent unwind-protect forms, following the Emacs source code, and then convert those.

As part of kidd/org-gcal.el#160, I'm working on some code that uses iter2 (via a library https://github.com/skeeto/emacs-aio that I ported to use iter2 instead of iter). I'm having some issues with my code, though. Namely, the loop that I've linked seems to somehow yield early, so that only the first entry in org-gcal-fetch-file-alist is actually processed. What would be the best way to debug the code to see where exactly I'm running into an issue? I know about iter2-generate-tracing-functions, but I'm unsure how exactly to interpret the output of this.