Skip to content

fix: trim repeat-dict metadata before validation #166

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

Merged
njzjz merged 2 commits into from
Aug 31, 2026
Merged

fix: trim repeat-dict metadata before validation #166

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
f973619
fix: trim repeat-dict metadata before validation
njzjz-bot Aug 30, 2026
037aa3c
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Aug 31, 2026
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
57 changes: 49 additions & 8 deletions dargs/dargs.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -350,6 +350,7 @@ def traverse(
path: list[str] | None = None,
allow_ref: bool = False,
_ref_base_dir: str | None = None,
_trim_pattern: str | None = None,
) -> None:
# first, do something with the key
# then, take out the vaule and do something with it
Expand All @@ -370,6 +371,7 @@ def traverse(
newpath,
allow_ref,
_ref_base_dir,
_trim_pattern,
)

def traverse_value(
Expand All @@ -382,6 +384,7 @@ def traverse_value(
path: list[str] | None = None,
allow_ref: bool = False,
_ref_base_dir: str | None = None,
_trim_pattern: str | None = None,
) -> None:
# this is not private, and can be called directly
# in the condition where there is no leading key
Expand All @@ -397,6 +400,7 @@ def traverse_value(
path,
allow_ref,
_ref_base_dir,
_trim_pattern,
)
elif self.repeat and isinstance(value, list):
for idx, item in enumerate(value):
Expand All @@ -409,8 +413,14 @@ def traverse_value(
[*path, str(idx)],
allow_ref,
_ref_base_dir,
_trim_pattern,
)
elif self.repeat and isinstance(value, dict):
# Repeat dictionaries use their keys as item names. Trim comment or
# metadata entries before visiting items, since those entries may
# not contain dictionaries and must not be type-checked as items.
if _trim_pattern is not None:
trim_by_pattern(value, _trim_pattern)
for kk, item in value.items():
self._traverse_sub(
item,
Expand All @@ -421,6 +431,7 @@ def traverse_value(
[*path, kk],
allow_ref,
_ref_base_dir,
_trim_pattern,
)

def _traverse_sub(
Expand All @@ -433,6 +444,7 @@ def _traverse_sub(
path: list[str] | None = None,
allow_ref: bool = False,
_ref_base_dir: str | None = None,
_trim_pattern: str | None = None,
) -> None:
if path is None:
path = [self.name]
Expand All @@ -458,6 +470,7 @@ def _traverse_sub(
path,
allow_ref,
ref_base_dir,
_trim_pattern,
)

# above are general traverse part
Expand Down Expand Up @@ -633,13 +646,20 @@ def normalize(
key_hook=Argument._convert_alias,
variant_hook=Variant._convert_choice_alias,
allow_ref=allow_ref,
_trim_pattern=trim_pattern,
)
if do_default:
self.traverse(
argdict, key_hook=Argument._assign_default, allow_ref=allow_ref
argdict,
key_hook=Argument._assign_default,
allow_ref=allow_ref,
_trim_pattern=trim_pattern,
)
self.traverse(
argdict, key_hook=Argument._handle_empty_dict, allow_ref=allow_ref
argdict,
key_hook=Argument._handle_empty_dict,
allow_ref=allow_ref,
_trim_pattern=trim_pattern,
)
if trim_pattern is not None:
trim_by_pattern(argdict, trim_pattern, reserved=[self.name])
Expand All @@ -649,6 +669,7 @@ def normalize(
d, trim_pattern, a.flatten_sub(d, p).keys()
),
allow_ref=allow_ref,
_trim_pattern=trim_pattern,
)
return argdict

Expand Down Expand Up @@ -693,13 +714,20 @@ def normalize_value(
key_hook=Argument._convert_alias,
variant_hook=Variant._convert_choice_alias,
allow_ref=allow_ref,
_trim_pattern=trim_pattern,
)
if do_default:
self.traverse_value(
value, key_hook=Argument._assign_default, allow_ref=allow_ref
value,
key_hook=Argument._assign_default,
allow_ref=allow_ref,
_trim_pattern=trim_pattern,
)
self.traverse_value(
value, key_hook=Argument._handle_empty_dict, allow_ref=allow_ref
value,
key_hook=Argument._handle_empty_dict,
allow_ref=allow_ref,
_trim_pattern=trim_pattern,
)
if trim_pattern is not None:
self.traverse_value(
Expand All @@ -708,6 +736,7 @@ def normalize_value(
d, trim_pattern, a.flatten_sub(d, p).keys()
),
allow_ref=allow_ref,
_trim_pattern=trim_pattern,
)
return value

Expand Down Expand Up @@ -1186,15 +1215,27 @@ def trim_by_pattern(
rep = fnmatch.translate(pattern) if not use_regex else pattern
rem = re.compile(rep)
if reserved:
# Use lambda instead of rem.match for ty type checker compatibility
conflict = list(filter(lambda x: rem.match(x) is not None, reserved))
# Use lambda instead of rem.match for ty type checker compatibility.
# Keys from Python dictionaries are not required to be strings; only
# string keys can match a glob/regex pattern.
conflict = list(
filter(
lambda x: isinstance(x, str) and rem.match(x) is not None,
reserved,
)
)
if conflict:
raise ValueError(
f"pattern `{pattern}` conflicts with the "
f"following reserved names: {', '.join(conflict)}"
)
# Use lambda instead of rem.match for ty type checker compatibility
unrequired = list(filter(lambda x: rem.match(x) is not None, argdict.keys()))
# Skip non-string keys instead of passing them to the regular expression.
unrequired = list(
filter(
lambda x: isinstance(x, str) and rem.match(x) is not None,
argdict.keys(),
)
)
for key in unrequired:
argdict.pop(key)

Expand Down
26 changes: 26 additions & 0 deletions tests/test_normalizer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -80,6 +80,32 @@ def test_trim(self) -> None:
self.assertDictEqual(end1, ref)
self.assertTrue(end1 is beg)

def test_trim_repeat_dict_container(self) -> None:
"""Trim metadata entries before treating repeat-dict keys as items."""
ca = Argument("base", dict, [Argument("value", int)], repeat=True)
beg = {
"base": {
"_container_comment": "ignored",
"item": {"value": 1, "_item_comment": "ignored"},
}
}
ref = {"base": {"item": {"value": 1}}}

self.assertDictEqual(ca.normalize(beg, trim_pattern="_*"), ref)
self.assertDictEqual(
beg["base"]["item"], {"value": 1, "_item_comment": "ignored"}
)
self.assertDictEqual(
ca.normalize_value(beg["base"], trim_pattern="_*"), ref["base"]
)

def test_trim_repeat_dict_non_string_key(self) -> None:
"""Do not pass non-string Python mapping keys to the trim regex."""
ca = Argument("base", dict, [Argument("value", int)], repeat=True)
value = {1: {"value": 1}}

self.assertDictEqual(ca.normalize_value(value, trim_pattern="_*"), value)

def test_combined(self) -> None:
ca = Argument(
"base",
Expand Down