Usage

Every example below assumes the import:

from pyfilechoose import file_choose, files_choose

Pick a single file

path = file_choose()
print(path)   # e.g. C:\Users\you\Documents\data.csv

file_choose() returns a single absolute path as a str. If the user closes the dialog without choosing anything, it raises FileNotFoundError.

Read a file directly with pandas

Because the return value is a path string, it drops into any function that accepts a file path:

import pandas as pd

# Equivalent to df <- read.csv(file.choose()) in R
df = pd.read_csv(file_choose())

You can combine it with the usual reader options:

df = pd.read_csv(file_choose(filetypes=[("CSV files", "*.csv")]), sep=";")

Set the dialog title

path = file_choose(title="Select your input dataset")

Filter by file type

filetypes takes a list of (label, pattern) tuples. The label is shown in the dialog's type dropdown; the pattern controls which files are visible.

path = file_choose(
    filetypes=[
        ("CSV files", "*.csv"),
        ("Excel files", "*.xlsx *.xls"),
        ("All files", "*.*"),
    ]
)

Patterns follow the platform's glob rules. Several extensions can share one label by separating them with spaces, as in "*.xlsx *.xls".

Open in a specific directory

path = file_choose(initialdir="~/Documents")

If initialdir is omitted, the dialog opens wherever the platform last left it.

Pick several files at once

paths = files_choose()
for p in paths:
    print(p)

files_choose() returns a list[str] of absolute paths in selection order. It raises FileNotFoundError if nothing is selected.

images = files_choose(
    title="Select images to process",
    filetypes=[("Images", "*.png *.jpg *.jpeg")],
)

Handle cancellation

Both functions raise FileNotFoundError on cancel, so wrap the call when a user might back out:

try:
    path = file_choose()
except FileNotFoundError:
    print("No file selected, aborting.")
    raise SystemExit