read_csv from filehandle, when turned into a view, stops working with `No files found that match the pattern "DUCKDB_INTERNAL_OBJECTSTORE://...`

Open
#477 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Assessment

Difficulty
3/5
Estimated time
1-2 days
Newbie friendliness
55/100
Issue type
Bug
Clarity
Mostly clear
Activity status
Quiet
Tech stack
python
Domain
database

Research direction

Start with the Python API path exercised by cursor.read_csv(StringIO) and DuckDBPyRelation.to_view(), using the reproduction in the issue. Check the behavior when the relation is temporary or goes out of scope, then verify that a view created from a filehandle remains queryable afterward, including the direct chained call case.

Written by the indexing model from the issue text.

Description

needs triage
What happens?

cursor.read_csv(filehandle) returns a DuckDBPyRelation object on which you can call .to_view(viewname) but the view isn't usable later, once the returned DuckDBPyRelation object has gone out of scope. If you try to do something like cursor.read_csv(filehandle).to_view('viewname') then it doesn't work at all.

This doesn't seem to be a problem for opening a csv by filename, or for relations made into tables, just for csvs opened from filehandles and made into views. I think I can understand why it's happening, but it is

(In case you're wondering, I'm opening files from filehandles as a workaround for https://github.com/duckdb/duckdb/discussions/12232 ... so more typically with bzip2.open(filename) as fh: cursor.read_csv(fh).to_view(viewname) or similar but using a StringIO makes for a simpler demo to reproduce.)

To Reproduce
import duckdb
from io import StringIO

cursor = duckdb.connect()

csv_file = StringIO("foo,bar\nhello,world")
rel = cursor.read_csv(csv_file)
rel.to_view("view1")
print(rel.alias)
print(cursor.sql("select * from view1"))

csv_file = StringIO("foo,bar\nhello,world")
cursor.read_csv(csv_file).to_view("view2")
print(cursor.sql("select * from view2"))

The first way works, the second way doesn't:

$ python x.py 
DUCKDB_INTERNAL_OBJECTSTORE://b38cc260dcc16094
┌─────────┬─────────┐
│   foo   │   bar   │
│ varchar │ varchar │
├─────────┼─────────┤
│ hello   │ world   │
└─────────┴─────────┘

Traceback (most recent call last):
  File "/home/nick/Work/wehi/countess/x.py", line 14, in <module>
    print(cursor.sql("select * from view2"))
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
duckdb.duckdb.IOException: IO Error: No files found that match the pattern "DUCKDB_INTERNAL_OBJECTSTORE://ce4251be44deb137"

It also fails with the same error if rel is deleted or goes out of scope before the SQL query of the view. Note also that read_csv(filename).to_view(viewname) works fine.

OS:

Linux 6.8.0 x86_64

DuckDB Package Version:

1.5.3 from pypi

Also source build duckdb-python 1.6.0-dev45 @ ab63b5f2d2713d8f57486209e3c78e0ba246c6c6
w/ duckdb v1.5.2-4685-g01eda16d6e

Python Version:

3.12.3

Full Name:

Nick Moore

Affiliation:

Mnemote Pty Ltd

What is the latest build you tested with? If possible, we recommend testing with the latest nightly build.

I have tested with a stable release 1.5.3 also 1.3.1
I have tested with 1.6.0.dev45 @ ab63b5f2d2713d8f57486209e3c78e0ba246c6c6

Did you include all relevant data sets for reproducing the issue?

Yes

Did you include all code required to reproduce the issue?
  • Yes, I have
Did you include all relevant configuration to reproduce the issue?
  • Yes, I have
Dominant language
Python
Stars
186
Forks
113
Avg merge
20h 58m
Merged PRs (30d)
11

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from duckdb/duckdb-python

All issues in duckdb/duckdb-python

Similar issues

More Python issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.