@rbarzic commented on Mon Nov 26 2018

Hi,

I've installed realgud and tried it with ipdb and get this strange behaviour where the source code position indicated by ipdb is not updated in the source window (basically the green cursor in the source window is "late" compared to the current source line seen by pdb)

I have the following message at the startup of ipdb 👍
WARNING: your terminal doesn't support cursor position requests (CPR).

I don't have the problem when running with pdb

I've using realgud-20180925.10 from elpa on emacs 26.1

@rocky commented on Mon Nov 26 2018

See if commit b8b587b fixes this.

It looks like a new ANSI control sequence was added before the prompt (> ...) and that messed (messes?) up the location tracking.

@rocky commented on Mon Nov 26 2018

One other thing - if the synchonization occasionally gets off, you can sometimes for a resync by issuing up 0 in (i)pdb.

@rbarzic commented on Mon Nov 26 2018

Thanks for the tip - Unfortunatly, I'm struggling a bit to install the realgud stuff from git when not having sudo access and a local installation of emacs under /opt and be sure that the right version is used by emacs. I think I've install the latest version properly but the bug seems to be still here. Is there a way to get the version of realgud from emacs ?

@rocky commented on Mon Nov 26 2018

This version of realgud is already on melpa.

To see whether you are getting the new code, you should not be seeing ASCII escape characters like
�[J> before where it shows where you are stopped or �[6C after listing the source code.

Also, you need to make sure you are running in new comint shell, so that the changed functions are picked up.: shell hook, which is one of the places where a change was made, stores a copy of the function and doesn't automatcally get updated when the function changes.

The simplest thing is to do here then is run a new emacs where you load everything from scratch.

Since Emacs is dynamic, all you need to do is load-file any of the files that have changed, before running realgud for the first time. To get a list of files that have changed, look at the commit.

And lastly, I note that for a guy that seems to request this much assistance you are relucant to star any project

@rbarzic commented on Mon Nov 26 2018

Thanks - I was confused about the version to use because the commit you pointed to appears to be done 9 hours ago while the latest melpa version is 20180925.10 - which I read as "25 September 2018"- so I was not expected to get the fix through MELPA - thus I had to be sure to remove any realgud versions already installed and do a reinstallation using git. I think I got it right now.
And I was asking for a possible way to get the realgud version as some other modes( like Verilog-mode ) sometimes provide a way to get it easily from a menu or a command.

@rocky commented on Mon Nov 26 2018

Acually you are correct melpa: is out of date. I read the date wrong. But I guess it will catch up eventually.

If you look at the README.md there is a link to instructions for setting up. You read that, right?

@rbarzic commented on Mon Nov 26 2018

Yes I did - And I don't have any ASCII escape sequence on screen. Only the message "WARNING: your terminal doesn't support cursor position requests (CPR)" at the beginning of the session.

@rocky commented on Mon Nov 26 2018

Ok. Well, if this doesn't do it, then I guess I would like to leave it here. Things work for me well enough with this. (Actually, I don't use ipdb at all.)

As this is open source feel free to dig in and investigate and/or fix. Or we can wait until someone else is interested in this.

@rbarzic commented on Mon Nov 26 2018

ok I understand - I'm not particularly familiar with emacs interaction with other processes so I may not be able to help much on this unfortunatly

@rocky commented on Mon Nov 26 2018

No one is born understanding emacs interaction with other processes.

That's all I get for ipdb related functions:

• realgud:ipdb-completion-at-point
• realgud:ipdb-minibuffer-history

I'm unable to autocomplete anything.

@stnbu commented on Thu Apr 20 2017

I'm quite new to emacs so please forgive any stupid...

I have the following in my ~/.emacs

(add-hook 'shell-mode-hook 'ansi-color-for-comint-mode-on)
(add-to-list 'comint-output-filter-functions 'ansi-color-process-output)

This does address a similar problem (re ANSI color sequences) with elpy+ipython.

But when I

M-x load-library [RET] realgud [RET]

M-x realgud:ipdb [RET]

from within a python buffer, I get a new ipdb interpreter in another window (that appears to be working as expected). But I get the above mentioned ANSI color escape sequence problem:

Current directory: /foo/
Command: ipdb /foo/__init__.py
> /foo/__init__.py(1)<module>()
----> 1 import paramiko
      2 
      3 

�[6n�[Jipdb> �[6D�[6C
�[6D�[Jipdb> �[6D
�[J�[6n�[Jipdb> �[6D�[6C�[6D�[J�[6n�[Jipdb> �[6D�[6C

From my package listing buffer:

 realgud       20170218.740  installed

@eflanigan00 commented on Thu Jul 13 2017

Anyone know anything about this? I'm having the same problem.

@xkal36 commented on Sun Aug 13 2017

I came across this problem also. I detailed my fix here: https://emacs.stackexchange.com/questions/34854/ipdb-displaying-ansi-color-escape-sequences

@rocky commented on Sun Aug 13 2017

@xkal36 I'd be grateful if you documented the workaround in off of https://github.com/realgud/realgud/wiki

Although beggars can't be choosers, my suggestion would be to document the problem in general: ANSI escape codes in Emacs shells, and then under ipdb notes add a mention to that along with your sample shell script that disables ANSI control codes. Thanks in advance.

I've added entries to the wiki. I may also try more aggressive ANSI Escape code filtering in the future.

@rocky commented on Tue Aug 15 2017

@stnbu @eflanigan00 @xkal36 I have just created a branch that may address this by filtering out all ANSI escape codes. Try this patch and let me know if this helps:

diff --git a/realgud/common/track.el b/realgud/common/track.el
index f26a0ed..3097de0 100644
--- a/realgud/common/track.el
+++ b/realgud/common/track.el
@@ -41,6 +41,11 @@
   :type 'symbolp
   :group 'realgud)
 
+(defcustom realgud-strip-ansi? nil
+"If non-nil, set strip ANSI control characters from output text we see"
+  :type 'symbolp
+  :group 'realgud)
+
 (declare-function fn-p-to-fn?-alias                   'realgud-helper)
 (declare-function realgud-bp-add-info                 'realgud-bp)
 (declare-function realgud-bp-del-info                 'realgud-bp)
@@ -182,7 +187,8 @@ evaluating (realgud-cmdbuf-info-loc-regexp realgud-cmdbuf-info)"
 
   (interactive "r")
   (if (> from to) (psetq to from from to))
-  (let* ((text (buffer-substring-no-properties from to))
+  (let* ((raw-text (buffer-substring-no-properties from to))
+	 (text (if realgud-strip-ansi? (ansi-color-filter-apply raw-text) raw-text))
 	 (loc (realgud-track-loc text cmd-mark))
 	 ;; If we see a selected frame number, it is stored
 	 ;; in frame-num. Otherwise, nil.

It is also in the strip-ansi-var branch on github.

Hi,

What do I do first to start installing this?

This package looks awesome, I installed it with elpa (I'm using spacemacs).
realgud:pdb works just fine, ipdb by itself works fine as well, but I have a repeating error/traceback when I start realgud:ipdb

Exception in thread Thread-1:
Traceback (most recent call last):
  File "/home/pierre/Apps/anaconda3/lib/python3.7/threading.py", line 926, in _bootstrap_inner
    self.run()
  File "/home/pierre/Apps/anaconda3/lib/python3.7/threading.py", line 870, in run
    self._target(*self._args, **self._kwargs)
  File "/home/pierre/Apps/anaconda3/lib/python3.7/site-packages/IPython/terminal/debugger.py", line 122, in in_thread
    line = self.pt_app.prompt()
  File "/home/pierre/Apps/anaconda3/lib/python3.7/site-packages/prompt_toolkit/shortcuts/prompt.py", line 1011, in prompt
    return get_event_loop().run_until_complete(self._dumb_prompt(self.message))
  File "/home/pierre/Apps/anaconda3/lib/python3.7/asyncio/events.py", line 644, in get_event_loop
    % threading.current_thread().name)
RuntimeError: There is no current event loop in thread 'Thread-1'.

This error repeats indefinitely, incrementing the thread number at each iteration. It fills up the screen so fast that you get to millions of characters in the buffer in a matter of seconds.

To reproduce this:

• install realgud and realgud-ipdb (in spacemacs these packages are inserted in dotspacemacs-additional-packages)
• open a python script
• realgud is already loaded, but I need to load realgud-ipdb: M-x load-library then realgud-ipdb
• M-x realgud:ipdb
• I use the default command: ipdb script.py
• I get the error above

I checked, all the package dependencies specified in the realgud tutorial are installed.
Thanks for your help and that cool package!