Skip to content

⚡️ Lazy-load rich_utils to reduce startup time #1128

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 3 commits into
base: master
Choose a base branch
from
+51 −10
Open

⚡️ Lazy-load rich_utils to reduce startup time #1128

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
a75a6d1
Lazy-load rich_utils to reduce startup time
oefe Jan 19, 2025
3f4cf32
Add unit test
oefe Mar 29, 2025
f2c8976
Merge branch 'master' into master
oefe Mar 29, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 15 additions & 0 deletions tests/assets/print_modules.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,15 @@
import sys

import typer

app = typer.Typer()


@app.command()
def main():
for m in sys.modules:
print(m)


if __name__ == "__main__":
app()
21 changes: 21 additions & 0 deletions tests/test_rich_import.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,21 @@
import subprocess
import sys
from pathlib import Path

ACCEPTED_MODULES = {"rich._extension", "rich"}


def test_rich_not_imported_unnecessary():
file_path = Path(__file__).parent / "assets/print_modules.py"
result = subprocess.run(
[sys.executable, "-m", "coverage", "run", str(file_path)],
capture_output=True,
encoding="utf-8",
)
modules = result.stdout.splitlines()
modules = [
module
for module in modules
if module not in ACCEPTED_MODULES and module.startswith("rich")
]
assert not modules
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is a matter of personal preference, but for clarity of code I would prefer

Suggested change
assert not modules
assert len(modules) == 0

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is arguably the most idiomatic style, see also PEP8:

For sequences, (strings, lists, tuples), use the fact that empty sequences are false:

# Correct:
if not seq:
if seq:
# Wrong:
if len(seq):
if not len(seq):

But I can change it if you prefer.

Copy link
Member

@svlandeg svlandeg Apr 18, 2025 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'll leave the final decision with Tiangolo 🙏

UPDATE: Note that I'm not suggesting to check truthness of len(seq) as in your code above, I'm suggesting to check len(seq) == 0.

Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

There are even pylint / Ruff rules that enforce this behavior: https://docs.astral.sh/ruff/rules/len-test/

assert not modules is the right thing.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm not reading this with the same interpretation. I'm not suggesting to check truthness of len(x), I'm suggesting to explicitely check the length being zero, as suggested on the page you cited:

You can either remove the call to len or compare the length against a scalar.

Comment on lines +16 to +21
Copy link

@basnijholt basnijholt Jun 26, 2025 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Fair point!

Suggested change
modules = [
module
for module in modules
if module not in ACCEPTED_MODULES and module.startswith("rich")
]
assert not modules
assert not any(
module in ACCEPTED_MODULES and module.startswith("rich")
for module in modules
)

is even better I guess.

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

These are all good suggestions, however in addition to the readability of the code, we should also consider the readability of the error messages.
This is important for example if someone introduces a regression by inadvertently importing rich again.
They should see right away which module caused the regression.

If an assertion fails, pytest does a good job printing meaningful error messages based on the condition. But this works only with relatively simple conditions.

To demonstrate this, I changed ACCEPTED_MODULES to an empty set to cause a test failure. With assert not modules, I get:

E       AssertionError: assert not ['rich._extension', 'rich']

This tells me immediately which modules were unexpectedly imported, and thus helps finding the cause.

With assert len(modules) == 0, I get:

E       AssertionError: assert 2 == 0
E        +  where 2 = len(['rich._extension', 'rich'])

This is a bit more lengthy and indirect, but I can still see which modules were unexpectedly imported.

With assert not any (...), I get:

E       assert not True
E        +  where True = any(<generator object test_rich_not_imported_unnecessary_any.<locals>.<genexpr> at 0x101803510>)

This is rather cryptic and unhelpful.

I therefore prefer the first variant. I find the second variant (with len) acceptable too, but it wouldn't be my choice. I don't like the third variant because of the cryptic error message.

However, in the end, this is @tiangolo's project, so I will use whatever he prefers.

3 changes: 2 additions & 1 deletion typer/cli.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,6 @@
import rich

has_rich = True
from . import rich_utils

except ImportError: # pragma: no cover
has_rich = False
Expand Down Expand Up @@ -275,6 +274,8 @@ def get_docs_for_click(
def _parse_html(input_text: str) -> str:
if not has_rich: # pragma: no cover
return input_text
from . import rich_utils

return rich_utils.rich_to_html(input_text)


Expand Down
10 changes: 8 additions & 2 deletions typer/core.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -33,8 +33,6 @@
try:
import rich

from . import rich_utils

DEFAULT_MARKUP_MODE: MarkupMode = "rich"

except ImportError: # pragma: no cover
Expand Down Expand Up @@ -213,6 +211,8 @@ def _main(
raise
# Typer override
if rich and rich_markup_mode is not None:
from . import rich_utils

rich_utils.rich_format_error(e)
else:
e.show()
Expand Down Expand Up @@ -243,6 +243,8 @@ def _main(
raise
# Typer override
if rich and rich_markup_mode is not None:
from . import rich_utils

rich_utils.rich_abort_error()
else:
click.echo(_("Aborted!"), file=sys.stderr)
Expand Down Expand Up @@ -688,6 +690,8 @@ def main(
def format_help(self, ctx: click.Context, formatter: click.HelpFormatter) -> None:
if not rich or self.rich_markup_mode is None:
return super().format_help(ctx, formatter)
from . import rich_utils

return rich_utils.rich_format_help(
obj=self,
ctx=ctx,
Expand Down Expand Up @@ -751,6 +755,8 @@ def main(
def format_help(self, ctx: click.Context, formatter: click.HelpFormatter) -> None:
if not rich or self.rich_markup_mode is None:
return super().format_help(ctx, formatter)
from . import rich_utils

return rich_utils.rich_format_help(
obj=self,
ctx=ctx,
Expand Down
12 changes: 5 additions & 7 deletions typer/main.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -50,11 +50,6 @@

try:
import rich
from rich.traceback import Traceback

from . import rich_utils

console_stderr = rich_utils._get_rich_console(stderr=True)

except ImportError: # pragma: no cover
rich = None # type: ignore
Expand Down Expand Up @@ -82,16 +77,19 @@ def except_hook(
supress_internal_dir_names = [typer_path, click_path]
exc = exc_value
if rich:
from .rich_utils import MAX_WIDTH
from rich.traceback import Traceback

from . import rich_utils

rich_tb = Traceback.from_exception(
type(exc),
exc,
exc.__traceback__,
show_locals=exception_config.pretty_exceptions_show_locals,
suppress=supress_internal_dir_names,
width=MAX_WIDTH,
width=rich_utils.MAX_WIDTH,
)
console_stderr = rich_utils._get_rich_console(stderr=True)
console_stderr.print(rich_tb)
return
tb_exc = traceback.TracebackException.from_exception(exc)
Expand Down
Loading