I think you sent me a message to my personal email rather than the mailing
list, but I would like to give you some friendly advice.
I think you should reconsider your position when you see me or someone else
disagreeing with your points. I have worked on SymPy longer, and if I
wanted to push every complaint I have, it would be 100 times longer, but I
don't do it for good reason. The expectation that I should agree with your
personal complaints doesn't sound fair in this case.
I don't think that 'list_symbols' is useful for everyone. At least I
haven't found such an inconvenience personally, and I haven't needed it.
That's not because I have better memory or such, but because I already know
how to navigate such situations.
The problem that things get confusing after global variables are
overwritten does not only exist with respect to SymPy, but for all Python
variables. And that is especially problematic if the definition is
recursive, such as:
x = x + 1
or
api_key = requests.get(... + api_key)
because running the same script can break things
If we implement something that is too specific to SymPy, like SymPy
symbols, I would also find that too restrictive to be useful anyway.
The reason that I (and some others) don't complain about such issues is
that we understand these are user workflow issues, not library issues.
I generally suggest that:
- You should not keep the terminal open too long before you forget too
many stateful changes. You should exit and restart the terminal often when
you complete a task.
- You should be careful about copy-pasting any code to the terminal and
check the inputs, because running the same script twice or running it in
the wrong position can break things.
- You should take notes of the code in case you need to rerun the
terminal. For example, Jupyter notebook is good for this purpose, or any
Python file or function can work.
Anyway, the point is that if you are wise, you would realize it's
definitely better to improve your own skills and terminal usage practices
than attempting to fix the library.
On Saturday, June 28, 2025 at 1:57:28 PM UTC+2 peter.st...@gmail.com wrote:
> Dear Aasim,
>
>
>
> thanks!
>
> Just learned something new.
>
>
>
> Peter
>
>
>
> *From:* sy...@googlegroups.com <sy...@googlegroups.com> *On Behalf Of *
> Aasim
> *Sent:* Saturday, June 28, 2025 12:12 PM
> *To:* sy...@googlegroups.com
> *Subject:* Re: [sympy] Listing all symbols in sympy
>
>
>
> I agree with Sangyub here, it is problematic when we work with globals()
> and locals() (have suffered before).
>
> And to Peter, yes, Krishnav's one liner code works, at least it lists all
> the Symbol objects created
>
> globally, not locally though(like inside a function or inside a class
> defined in the terminal application) .
>
> We don't really need such typical methods for a very trivial use case in
> the library(any contributor can
>
> stumble upon a maintainer mentioning this exact same thing very often).
>
> Using globals() and locals(), again, is concerning. The concern that
> stands out the most for me here is
>
> thread safety. For example, let us assume that for once we indeed start
> working on adding the list_symbols()
>
> method that Stephen mentioned, then I will use Krishnav's one liner,
> something like:
>
> In [1]: from sympy import Symbol
>
>
>
> In [2]: def generate_symbols():
>
> ...: for i in range(10):
>
> ...: globals()[f"s{i}"] = Symbol(f"s{i}")
>
> ...:
>
>
>
> In [3]: def list_symbols():
>
> ...: return [name for name, obj in globals().items() if isinstance(obj,
> Symbol)] # Krishnav's one liner
>
> ...:
>
>
>
> In [4]: import threading
>
>
>
> In [5]: thread1 = threading.Thread(target=generate_symbols)
>
> ...: thread2 = threading.Thread(target=list_symbols)
>
>
>
> In [6]: thread1.start()
>
> ...: thread2.start()
>
>
>
> thread1 might still be mutating the global namespace while thread2 tries
> accessing it,
>
> which is unsafe (a very simple example of a race condition here).
>
> Similarly, when you use globals() or locals(), you're creating references
> to objects that
>
> might otherwise be eligible for garbage collection. Because the globals()
> dictionary often
>
> keeps references alive for longer than we might intend it to.
>
>
>
> Another example could be, in pretty simple web applications the dev might
> add a secret
>
> API key to the application's code before hosting it and if in the python
> backend they are
>
> using SymPy which in turn uses something like list_symbols() proposed
> here, can introduce
>
> a vulnerability that publicly exposes the secret API key (hypothesizing
> here but very much possible).
>
> I would suggest the same thing as Peter in conclusion:
> Stephen, you can just create a simple startup script for yourself that
> defines such trivial functionality(ies).
>
> I use IPython myself and I have a startup script written for it, if I had
> a similar use case as yours I would just
>
> add something like this to my startup_ipython.py:
>
> def list_symbols():
>
> return [name for name, obj in globals().items() if isinstance(obj,
> Symbol)] # Krishnav's one liner
>
>
> and then use the shell by calling `ipython -i startup_ipython.py`.
>
>
>
> On Sat, 28 Jun 2025 at 15:29, <peter.st...@gmail.com> wrote:
>
> So why don’t you define you own function:
>
>
>
> def list_symbols():
>
> return Krishnaw’s list
>
>
>
> Would that not work?
>
>
>
> *From:* sy...@googlegroups.com <sy...@googlegroups.com> *On Behalf Of *
> stephen.j...@gmail.com
> *Sent:* Saturday, June 28, 2025 11:31 AM
> *To:* sy...@googlegroups.com
> *Subject:* Re: [sympy] Listing all symbols in sympy
>
>
>
> Yes, that single line of code works but it’s simpler and faster to be able
> to just type a short command like “list_symbols()” to list all the symbols
> created in terminal.
>
>
>
>
>
> Sent from my iPhone.
>
>
>
> On 28 Jun 2025, at 10:14, peter.st...@gmail.com wrote:
>
>
>
> I have been following this discussion as I use a library of sympy
> (sympy.physics.mechanics) quite a bit.
>
> Would Krishnav’s one line code not do what Stephen wants, or am I missing
> something?
>
>
>
> Thanks!
>
>
>
>
>
>
>
> *From:* sy...@googlegroups.com <sy...@googlegroups.com> *On Behalf Of
> *Sangyub
> Lee
> *Sent:* Saturday, June 28, 2025 10:59 AM
> *To:* sympy <sy...@googlegroups.com>
> *Subject:* Re: [sympy] Listing all symbols in sympy
>
>
>
> I don't understand `locals()` or `globals()` very well, but it's always
> good that we don't implement something directly based on functionality that
> we don't feel safe using. Or will you put in the effort to address those
> concerns anyway
>
> Regardless, it is primarily your responsibility to prove a strong reason
> why we need such functionality in SymPy. We don't have a responsibility to
> implement something general to solve the trivial terminal problems that you
> face.
>
>
>
> On Saturday, June 28, 2025 at 10:41:48 AM UTC+2 stephen.j...@gmail.com
> wrote:
>
> What does “…locals() or globals() introspection…” mean?
>
>
>
> Please prove to me that “… it lead to issues with garbage collection,
> introduce security vulnerabilities and create thread safety concerns”.
>
>
>
> I won’t be moving beyond Python.
>
>
>
> What do you mean by “…domain-specific…”!
>
>
>
>
>
> Sent from my iPhone.
>
>
>
> On 28 Jun 2025, at 07:26, Sangyub Lee <syle...@gmail.com> wrote:
>
> I don't think we should add such functionality to the library. I would
> not use locals() or globals() introspection in library because it lead to
> issues with garbage collection, introduce security vulnerabilities, and
> create thread safety concerns.
>
>
>
> I think you would need to move beyond Python and use a different
> domain-specific language or user interface to implement such functionality
> in a secure way. I suggest consulting with experienced software engineers
> or IT agencies to implement that approach.
>
> On Wednesday, June 25, 2025 at 8:17:49 PM UTC+2 stephen.j...@gmail.com
> wrote:
>
> My “use case” is that suppose you are executing Python code in the Mac
> terminal application to test out some code and you lose track of what
> variables you have created and you want to delete some or all of them to
> start over.
>
> A function which lists all the created variables would be useful so you
> can delete them all and start over. A very basic function.
>
> Sent from my iPhone.
>
> > On 25 Jun 2025, at 17:46, Aaron Meurer <asme...@gmail.com> wrote:
> >
> > What is your use-case for such a function? In most cases if you are
> > writing a script you know what symbols are defined because you have
> > defined them all statically somewhere, so there's no need to
> > programmatically get a list of them.
> >
> > Aaron Meurer
> >
> >> On Mon, Jun 23, 2025 at 9:14 AM Stephen Learmonth
> >> <stephen.j...@gmail.com> wrote:
> >>
> >> Thanks Krishnav,
> >>
> >> It would have been really useful if Sympy just had a function like
> list_symbols() oir something from the start.
> >>
> >> It''s such a basic thing to want to be able to do!
> >>
> >> Stephen
> >>
> >>> On Monday, 23 June 2025 at 16:00:38 UTC+1 bajoria...@gmail.com wrote:
> >>>
> >>> Hi Stephen,
> >>>
> >>> If you want to list all sympy symbols that you have created in a
> Python script, one straightforward way is to use Python’s built-in
> globals() to scan the current namespace and check for instances of
> sympy.Symbol.
> >>>
> >>> Here's a simple example to illustrate:
> >>>
> >>>>>> from sympy import Symbol
> >>>>>> from sympy.abc import a, b, c
> >>>>>> x = Symbol('x')
> >>>>>> y = Symbol('y')
> >>>>>> z = 3 # this is not a SymPy symbol
> >>>>>> symbols_used = [name for name, obj in globals().items() if
> isinstance(obj, Symbol)]
> >>>>>> print("Symbols used :", symbols_used)
> >>>
> >>> This would output:
> >>>
> >>> Symbols used : ['a', 'b', 'c', 'x', 'y']
> >>>
> >>> But do let me know if you are looking for something more specific.
> Happy to help further!
> >>>
> >>> Best regards,
> >>>
> >>> Krishnav Bajoria.
> >>>
> >>>
> >>> On Mon, Jun 23, 2025 at 8:07 PM Stephen Learmonth <
> stephen.j...@gmail.com> wrote:
> >>>>
> >>>> How do I list ALL sympy symbols created in a Python script?
> >>>>
> >>>> --
> >>>> You received this message because you are subscribed to the Google
> Groups "sympy" group.
> >>>> To unsubscribe from this group and stop receiving emails from it,
> send an email to sympy+un...@googlegroups.com.
> >>>> To view this discussion visit
> https://groups.google.com/d/msgid/sympy/2f682bea-f462-4188-9873-82c9ebafe889n%40googlegroups.com.
>
>
> >>
> >> --
> >> You received this message because you are subscribed to the Google
> Groups "sympy" group.
> >> To unsubscribe from this group and stop receiving emails from it, send
> an email to sympy+un...@googlegroups.com.
> >> To view this discussion visit
> https://groups.google.com/d/msgid/sympy/c41df7a9-b27c-4e81-9fcd-85a52789b4e6n%40googlegroups.com.
>
>
> >
> > --
> > You received this message because you are subscribed to the Google
> Groups "sympy" group.
> > To unsubscribe from this group and stop receiving emails from it, send
> an email to sympy+un...@googlegroups.com.
> > To view this discussion visit
> https://groups.google.com/d/msgid/sympy/CAKgW%3D6%2B%2B_feNU%3D1psGQ1H7PapFHO67DtaqHKW8PiR59iZmrRDA%40mail.gmail.com.
>
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "sympy" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sympy+un...@googlegroups.com.
>
> To view this discussion visit
> https://groups.google.com/d/msgid/sympy/f0234176-abe1-4451-a6bd-9c58bb35bb06n%40googlegroups.com
>
> <https://groups.google.com/d/msgid/sympy/f0234176-abe1-4451-a6bd-9c58bb35bb06n%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>
> --
> You received this message because you are subscribed to the Google Groups
> "sympy" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sympy+un...@googlegroups.com.
> To view this discussion visit
> https://groups.google.com/d/msgid/sympy/9f2e3377-26ce-4894-8f14-2b15c4869812n%40googlegroups.com
>
> <https://groups.google.com/d/msgid/sympy/9f2e3377-26ce-4894-8f14-2b15c4869812n%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>
> --
> You received this message because you are subscribed to the Google Groups
> "sympy" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sympy+un...@googlegroups.com.
> To view this discussion visit
> https://groups.google.com/d/msgid/sympy/0c0901dbe80d%2414fff820%243effe860%24%40gmail.com
>
> <https://groups.google.com/d/msgid/sympy/0c0901dbe80d%2414fff820%243effe860%24%40gmail.com?utm_medium=email&utm_source=footer>
> .
>
> --
> You received this message because you are subscribed to the Google Groups
> "sympy" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sympy+un...@googlegroups.com.
> To view this discussion visit
> https://groups.google.com/d/msgid/sympy/126633FB-0907-496F-BE65-EADA8192C5FF%40gmail.com
>
> <https://groups.google.com/d/msgid/sympy/126633FB-0907-496F-BE65-EADA8192C5FF%40gmail.com?utm_medium=email&utm_source=footer>
> .
>
> --
> You received this message because you are subscribed to the Google Groups
> "sympy" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sympy+un...@googlegroups.com.
> To view this discussion visit
> https://groups.google.com/d/msgid/sympy/0c3801dbe813%2450a8bd20%24f1fa3760%24%40gmail.com
>
> <https://groups.google.com/d/msgid/sympy/0c3801dbe813%2450a8bd20%24f1fa3760%24%40gmail.com?utm_medium=email&utm_source=footer>
> .
>
> --
> You received this message because you are subscribed to the Google Groups
> "sympy" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sympy+un...@googlegroups.com.
>
> To view this discussion visit
> https://groups.google.com/d/msgid/sympy/CAO0%2BxCUyzFy-x1i5P2_u%2BjA%3DBLHnH35LE98p3kDCQBwYKSkt4A%40mail.gmail.com
>
> <https://groups.google.com/d/msgid/sympy/CAO0%2BxCUyzFy-x1i5P2_u%2BjA%3DBLHnH35LE98p3kDCQBwYKSkt4A%40mail.gmail.com?utm_medium=email&utm_source=footer>
> .
>
--
You received this message because you are subscribed to the Google Groups
"sympy" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to sympy+unsubscr...@googlegroups.com.
To view this discussion visit
https://groups.google.com/d/msgid/sympy/412c1e09-4273-406f-8ab6-ed0580139da0n%40googlegroups.com.
