get_compounds fails when given multiple smiles

[simanjo]

When trying to get compounds by smiles, the server responds with an error, when giving the identifiers as list:

\>\>\> pubchempy.get_compounds(['C', 'O'], 'smiles')
Traceback (most recent call last):
[...]
urllib.error.HTTPError: HTTP Error 400: PUGREST.BadRequest

The above exception was the direct cause of the following exception:
[...]
pubchempy.BadRequestError: PubChem HTTP Error 400 PUGREST.BadRequest: Unable to standardize the given structure - perhaps some special 
characters need to be escaped or data packed in a MIME form? (error: , status: 400, output: Caught ncbi::CException: Standardization 
failed, Output Log:, Record 1: Warning: Cactvs Ensemble cannot be created from input string, 
Record 1: Error: Unable to convert input into a compound object, , )`

I omitted the trace, bc I believe this is not supported by pubchem to start with according to the documentation:

<identifiers> = comma-separated list of positive integers (e.g. cid, sid, aid) or identifier strings (source, inchikey, formula); in some cases only a single identifier string (name, smiles, xref; inchi, sdf by POST only)

This similarly happens with just a single smiles in a list.
When trying with names in a list, the result is empty, so I figure pubchem interprets the comma-seperated list as single name.

I'm happy to provide more details on setup and the like, if necessary.

[nbehrnd]

@simanjo In an instance of Linux Debian 13/trixie (stable), a pristine virtual environment for Python, and a pip install from PyPI (and thus sourced from the blessed repository), I'm able to search by SMILES.

My log on the CLI:

no@deb:~/Desktop$ mkdir test
no@deb:~/Desktop$ cd test/
no@deb:~/Desktop/test$ python -m venv venv
no@deb:~/Desktop/test$ source ./venv/bin/activate
(venv) no@deb:~/Desktop/test$ pip install pubchempy\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00
Collecting pubchempy
  Downloading pubchempy-1.0.5-py3-none-any.whl.metadata (4.3 kB)
Downloading pubchempy-1.0.5-py3-none-any.whl (21 kB)
Installing collected packages: pubchempy
Successfully installed pubchempy-1.0.5
(venv) no@deb:~/Desktop/test$ 
(venv) no@deb:~/Desktop/test$ pip list
Package   Version
--------- -------
pip       25.1.1
PubChemPy 1.0.5
(venv) no@deb:~/Desktop/test$ 
(venv) no@deb:~/Desktop/test$ python
Python 3.13.5 (main, Jun 25 2025, 18:55:22) [GCC 14.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import pubchempy as pcp
>>> from pubchempy import get_compounds, Compound
>>> 
>>> compound = pcp.get_compounds("CC", "smiles")
>>> print(type(compound))
<class 'list'>
>>> print(compound[0].inchi)
InChI=1S/C2H6/c1-2/h1-2H3
>>> 
>>> exit()
(venv) no@deb:~/Desktop/test$ lsb_release -a
No LSB modules are available.
Distributor ID:	Debian
Description:	Debian GNU/Linux 13 (trixie)
Release:	13
Codename:	trixie
(venv) no@deb:~/Desktop/test$ 

The repeated \00 only is how the shell (here BASH) retains the incremental progress bar.

Prior to fall 2025, there were some issues to reliably access pubchem via pubchempy 1.0.4 in Windows. Matt however fixed these (and many more) prior to the update 1.0.5 now available via PyPI.

[simanjo]

@nbehrnd
Thanks for the fast reply!

I am able to search by smiles, but I can not use the (undocumented?) and only implied batch download feature.
At least, that's how I interpret the type signature of get_compounds:

PubChemPy/pubchempy.py

Line 472 in 9935a14

identifier: str | int | list[str | int],

This will (most likely) work when using multiple cid's but fails for multiple smiles (and most likely for names as well).

This also fails when giving a single smile in a list

pcp.get_compounds("[CC]", "smiles")

which could probably be fixed by more appropriate handling in the requests function

PubChemPy/pubchempy.py

Lines 350 to 353 in 9935a14

	if isinstance(identifier, int):
	identifier = str(identifier)
	if not isinstance(identifier, str):
	identifier = ",".join(str(x) for x in identifier)

I am actually unsure about a useful handling of this issue, and mostly wanted to highlight the (at least to me) misleading type annotation, suggesting a feature which is supposedly not supported by pubchem.

As for my setup, and demonstrating the implied batch download feature working for compound ids, and failing (by server response) for smiles:

user@ws:~/test$ cat /proc/version
Linux version 6.8.0-106-generic (buildd@lcy02-amd64-074) (x86_64-linux-gnu-gcc-13 (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0, GNU ld (GNU Binutils for Ubuntu) 2.42) #106-Ubuntu SMP PREEMPT_DYNAMIC Fri Mar  6 07:58:08 UTC 2026
user@ws:~/test$ python3 --version
Python 3.12.3
user@ws:~/test$ python3 -m venv env
user@ws:~/test$ source env/bin/activate
(env) user@ws:~/test$ pip install pubchempy
Collecting pubchempy
  Using cached pubchempy-1.0.5-py3-none-any.whl.metadata (4.3 kB)
Using cached pubchempy-1.0.5-py3-none-any.whl (21 kB)
Installing collected packages: pubchempy
Successfully installed pubchempy-1.0.5
(env) user@ws:~/test$ python
Python 3.12.3 (main, Mar  3 2026, 12:15:18) [GCC 13.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> from pubchempy import get_compounds
>>> id1 = get_compounds("CC", 'smiles')[0].cid
>>> id2 = get_compounds("O", 'smiles')[0].cid
>>> get_compounds(["CC", "O"], 'smiles')
Traceback (most recent call last):
  File "/home/user/test/env/lib/python3.12/site-packages/pubchempy.py", line 380, in request
    response = urlopen(apiurl, postdata, context=context)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib/python3.12/urllib/request.py", line 215, in urlopen
    return opener.open(url, data, timeout)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib/python3.12/urllib/request.py", line 521, in open
    response = meth(req, response)
               ^^^^^^^^^^^^^^^^^^^
  File "/usr/lib/python3.12/urllib/request.py", line 630, in http_response
    response = self.parent.error(
               ^^^^^^^^^^^^^^^^^^
  File "/usr/lib/python3.12/urllib/request.py", line 559, in error
    return self._call_chain(*args)
           ^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib/python3.12/urllib/request.py", line 492, in _call_chain
    result = func(*args)
             ^^^^^^^^^^^
  File "/usr/lib/python3.12/urllib/request.py", line 639, in http_error_default
    raise HTTPError(req.full_url, code, msg, hdrs, fp)
urllib.error.HTTPError: HTTP Error 400: PUGREST.BadRequest

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/user/test/env/lib/python3.12/site-packages/pubchempy.py", line 494, in get_compounds
    results = get_json(identifier, namespace, searchtype=searchtype, **kwargs)
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/user/test/env/lib/python3.12/site-packages/pubchempy.py", line 441, in get_json
    get(
  File "/home/user/test/env/lib/python3.12/site-packages/pubchempy.py", line 421, in get
    response = request(
               ^^^^^^^^
  File "/home/user/test/env/lib/python3.12/site-packages/pubchempy.py", line 383, in request
    raise create_http_error(e) from e
pubchempy.BadRequestError: PubChem HTTP Error 400 PUGREST.BadRequest: Unable to standardize the given structure - perhaps some special characters need to be escaped or data packed in a MIME form? (error: , status: 400, output: Caught ncbi::CException: Standardization failed, Output Log:, Record 1: Warning: Cactvs Ensemble cannot be created from input string, Record 1: Error: Unable to convert input into a compound object, , )
>>> comps = get_compounds([id1, id2], 'cid')
>>> comps
[Compound(6324), Compound(962)]
>>> exit()