feat: Create validation context classes from schema.meta.context

.github/workflows/build-test-deploy.yml

@@ -36,6 +36,7 @@ jobs:
       - uses: actions/checkout@v4
         with:
           fetch-depth: 0
+          submodules: true
       - uses: hynek/build-and-inspect-python-package@v2
         with:
           # Use attestation only if the action is triggered inside the repo
@@ -65,6 +66,9 @@ jobs:
 
     steps:
       - uses: actions/checkout@v4
+        with:
+          submodules: true
+
       - name: Set git name/email
         run: |
           git config --global user.email "[email protected]"

.gitmodules

@@ -0,0 +1,3 @@
+[submodule "tests/data/bids-examples"]
+	path = tests/data/bids-examples
+	url = https://github.com/bids-standard/bids-examples

pyproject.toml

@@ -23,6 +23,9 @@ classifiers = [
 requires-python = ">=3.9"
 dependencies = [
     "bidsschematools >=1.0",
+    "typing_extensions",
+    "attrs",
+    "httpx",
 ]
 
 [project.optional-dependencies]
@@ -55,12 +58,14 @@ tag_prefix = ""
 parentdir_prefix = ""
 
 [tool.pytest.ini_options]
+norecursedirs = ["data"]
 doctest_optionflags = ["NORMALIZE_WHITESPACE", "IGNORE_EXCEPTION_DETAIL"]
 
+
 [tool.coverage.run]
 branch = true
 parallel = true
-source = ["bids_validator"]
+source = ["bids_validator", "tests"]
 omit = [
   "setup.py",
   "*/_version.py",
@@ -77,6 +82,7 @@ exclude_lines = [
   "no cov",
   "if __name__ == .__main__.:",
   "if TYPE_CHECKING:",
+  "pytest.skip",
 ]
 
 # Disable black

src/bids_validator/__main__.py

@@ -0,0 +1,56 @@
+import typer
+
+from bids_validator import BIDSValidator
+from bids_validator.types.files import FileTree
+
+validator = BIDSValidator()
+
+app = typer.Typer()
+
+
+def check_children(tree: FileTree):
+    """Iterate over children of a FileTree and check if they are a directory or file.
+
+    If it's a directory then run again recursively, if it's a file file check the file name is
+    BIDS compliant.
+
+    Parameters
+    ----------
+    tree : FileTree
+        FileTree object to iterate over
+
+    """
+    for child in tree.children.values():
+        if child.is_dir:
+            check_children(child)
+        else:
+            check_bids(child.relative_path)
+
+
+def check_bids(path: str):
+    """Check if the file path is BIDS compliant.
+
+    Parameters
+    ----------
+    path : str
+        Path to check of compliance
+
+    """
+    # The output of the FileTree.relative_path method always drops the initial for the path which
+    # makes it fail the validator.is_bids check. Not sure if it's a Windows specific thing.
+    # This line adds it back.
+    path = f'/{path}'
+
+    if not validator.is_bids(path):
+        print(f'{path} is not a valid bids filename')
+
+
+@app.command()
+def main(bids_path: str):
+    root_path = FileTree.read_from_filesystem(bids_path)
+
+    check_children(root_path)
+
+
+if __name__ == '__main__':
+    app()

src/bids_validator/bidsignore.py

@@ -0,0 +1,129 @@
+"""Utilities for working with .bidsignore files."""
+
+import os
+import re
+from functools import lru_cache
+from typing import Protocol, Union
+
+import attrs
+
+from .types.files import FileTree
+
+
+@lru_cache
+def compile_pat(pattern: str) -> Union[re.Pattern, None]:
+    """Compile .gitignore-style ignore lines to regular expressions."""
+    orig = pattern
+    # A line starting with # serves as a comment.
+    if pattern.startswith('#'):
+        return None
+
+    # An optional prefix "!" which negates the pattern;
+    invert = pattern.startswith('!')
+
+    # Put a backslash ("\") in front of the first hash for patterns that begin with a hash.
+    # Put a backslash ("\") in front of the first "!" for patterns that begin with a literal "!"
+    if pattern.startswith((r'\#', r'\!')):
+        pattern = pattern[1:]  # Unescape
+
+    # Trailing spaces are ignored unless they are quoted with backslash ("\").
+    pattern = re.sub(r'(?<!\\) +$', '', pattern)
+
+    # A blank line matches no files, so it can serve as a separator for readability.
+    if pattern == '':
+        return None
+
+    # If there is a separator at the beginning or middle (or both) of the pattern,
+    # then the pattern is relative to the [root]
+    relative_match = pattern == '/' or '/' in pattern[:-1]
+    # If there is a separator at the end of the pattern then the pattern will only match
+    # directories, otherwise the pattern can match both files and directories.
+    directory_match = pattern.endswith('/')
+
+    # This does not handle character ranges correctly except when they are also valid regex
+    parts = [
+        '.*'
+        if part == '**'
+        else part.replace('*', '[^/]*').replace('?', '[^/]').replace('.', r'\.')
+        for part in pattern.strip('/').split('/')
+    ]
+
+    prefix = '^' if relative_match else '^(.*/|)'
+    postfix = r'/\Z' if directory_match else r'/?\Z'
+
+    # "**/" matches zero or more directories, so the separating slash needs to be optional
+    out_pattern = '/'.join(parts).replace('.*/', '.*/?')
+    out_pattern = f'{prefix}{out_pattern}{postfix}'
+
+    if invert:
+        raise ValueError(f'Inverted patterns not supported: {orig}')
+        # out_pattern = f'(?!{out_pattern})'
+
+    return re.compile(out_pattern)
+
+
+class HasMatch(Protocol):  # noqa: D101
+    def match(self, relpath: str) -> bool: ...  # noqa: D102
+
+
+@attrs.define
+class Ignore:
+    """Collection of .gitignore-style patterns.
+
+    Tracks successfully matched files for reporting.
+    """
+
+    patterns: list[str] = attrs.field(factory=list)
+    history: list[str] = attrs.field(factory=list, init=False)
+
+    @classmethod
+    def from_file(cls, pathlike: os.PathLike):
+        """Load Ignore contents from file."""
+        with open(pathlike) as fobj:
+            return cls([line.rstrip('\n') for line in fobj])
+
+    def match(self, relpath: str) -> bool:
+        """Match a relative path against a collection of ignore patterns."""
+        if any(compile_pat(pattern).match(relpath) for pattern in self.patterns):
+            self.history.append(relpath)
+            return True
+        return False
+
+
+@attrs.define
+class IgnoreMany:
+    """Match against several ignore filters."""
+
+    ignores: list[Ignore] = attrs.field()
+
+    def match(self, relpath: str) -> bool:
+        """Return true if any filters match the given file.
+
+        Will short-circuit, so ordering is significant for side-effects,
+        such as recording files ignored by a particular filter.
+        """
+        return any(ignore.match(relpath) for ignore in self.ignores)
+
+
+def filter_file_tree(filetree: FileTree) -> FileTree:
+    """Read .bidsignore and filter file tree."""
+    bidsignore = filetree.children.get('.bidsignore')
+    if not bidsignore:
+        return filetree
+    ignore = IgnoreMany([Ignore.from_file(bidsignore), Ignore(['/.bidsignore'])])
+    return _filter(filetree, ignore)
+
+
+def _filter(filetree: FileTree, ignore: HasMatch) -> FileTree:
+    items = filetree.children.items()
+    children = {
+        name: _filter(child, ignore)
+        for name, child in items
+        if not ignore.match(child.relative_path)
+    }
+
+    # XXX This check may not be worth the time. Profile this.
+    if any(children.get(name) is not child for name, child in items):
+        filetree = attrs.evolve(filetree, children=children)
+
+    return filetree

src/bids_validator/context.py

@@ -0,0 +1,38 @@
+"""Validation context for schema-based BIDS validation."""
+
+from .context_generator import get_schema, load_schema_into_namespace
+
+schema = get_schema()
+load_schema_into_namespace(schema['meta']['context'], globals(), 'Context')
+
+
+__all__ = [  # noqa: F822
+    'Context',
+    'Schema',
+    'Dataset',
+    'DatasetDescription',
+    'Tree',
+    'Subjects',
+    'Subject',
+    'Sessions',
+    'Entities',
+    'Sidecar',
+    'Associations',
+    'Events',
+    'Aslcontext',
+    'M0scan',
+    'Magnitude',
+    'Magnitude1',
+    'Bval',
+    'Bvec',
+    'Channels',
+    'Coordsystem',
+    'Columns',
+    'Json',
+    'Gzip',
+    'NiftiHeader',
+    'DimInfo',
+    'XyztUnits',
+    'Ome',
+    'Tiff',
+]