# in .emacs_workgroup causes parsing error

[chipschap]

This is a re-opening of a bug from back in 2014. I saved an org-agenda file which resulted in this string in the workgroup file:

#<buffer etc.

This gave a syntax read error whenever trying to create or open a new workgroup. It turns out the '#' needs to be escaped and then all is well.

This hackish rewrite of wg-write-sexp-to-file appears to fix the issue. It could surely be done better but it works for me.

(defun wg-write-sexp-to-file (sexp file)
"Write a printable (and human-readable) representation of SEXP to FILE."
(with-temp-buffer
(wg-insert-and-indent sexp)
;;; rjn
(goto-char (point-min))
(while (search-forward " #<" nil t)
(replace-match " \#<" nil t))
;;; rjn
(write-file file)))

The space in " #<" is necessary to avoid re-doing previous edits on future invocations.

I hope this will be of some use.

[redguardtoo]

thanks , could you send me a pull request?

[chipschap]

I can go and send pull request but not sure what to do after that? Sorry I am not so familiar with how GitHub works. Bob Newell

…

On Sat, Mar 5, 2022 at 6:44 PM Chen Bin ***@***.***> wrote: thanks , could you send me a pull request? — Reply to this email directly, view it on GitHub <#113 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAJELKTW5MPR562QNFQSEQTU6QZZPANCNFSM5P4Q4I7A> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

-- Bob Newell Honolulu, Hawai`i Via Linux/Emacs/Gnus/BBDB.

[redguardtoo]

I might need a bit more understanding about below code,

(while (search-forward " #<" nil t)
  (replace-match " \#<" nil t))

Why this code work?

A minimum case to reproduce the problem is also very helpful.

[chipschap]

Chen Bin ***@***.***> writes:

I need a bit more understanding about below code, (while (search-forward " #<" nil t) (replace-match " \#<" nil t)) Why this code works?

The problem I ran into was this. Let's say I have an Org Agenda buffer showing and I do wg-create-workgroup (and I don't yet have an .emacs_workgroup file). The .emacs_workgroup is created but it contains a string that looks partially like #<buffer followed by other stuff. This happens with buffers that are not visiting a file. Now, later when I try to wg-open-workgroup or wg-create-workgroup or anything that tries o read .emacs_workgroup then I will get a syntax read error because it doesn't like the '#' character in .emacs_workgroup. So I found by experimenting that if I escape the "#" by making it into "\#" then the file will be read correctly. So the code I added searches for #< and replaces it with \#<. It has to search for " #<" to find instances of " #<buffer" because there can be # present elsewhere which don't need to have escape like many "#s" occurrences. Of course search and replace is a very crude solution. Really the right way would have been to write out the \# in the first place rather than fixing it later on. But I didn't figure out how to do it the right way :( Thank you for your time and attention and I hope this explains a little better.

…

-- Bob Newell Honolulu, Hawai`i - Via GNU/Linux/Emacs/Gnus/BBDB

[redguardtoo]

It's because there is no wg-deserialize-org-agenda-mode-buffer,

[chipschap]

Ah, thank you, I see, and this is symptomatic of a bigger problem with buffers not visiting a file. However, a user (like me!) who may create a workgroup with Org Agenda or similar will then have the .emacs_workgroups file made unreadable because of the bare "#" character in front of "<buffer Org Agenda>". This seems like harsh punishment to lose all the saved workspaces!

[redguardtoo]

342fc30 store buffer name instead of buffer object (Chen Bin)