def _collect_stages(self):
from dvc.dvcfile import Dvcfile, is_valid_filename
stages = []
outs = set()
for root, dirs, files in self.tree.walk(self.root_dir):
for file_name in filter(is_valid_filename, files):
path = os.path.join(root, file_name)
stage_loader = Dvcfile(self, path).stages
stages.extend(stage_loader.values())
outs.update(out.fspath for stage in stages
for out in (out for out in stage.outs
if out.scheme == "local"))
dirs[:] = [d for d in dirs if os.path.join(root, d) not in outs]
return stages
def collect_granular(self, target, *args, **kwargs):
from ..dvcfile import Dvcfile, is_valid_filename
if not target:
return [(stage, None) for stage in self.stages]
file, name = parse_target(target)
if is_valid_filename(file) and not kwargs.get("with_deps"):
# Optimization: do not collect the graph for a specific .dvc target
stages = Dvcfile(self, file).stages.filter(name)
return [(stage, None) for stage in stages.values()]
try:
(out, ) = self.find_outs_by_path(file, strict=False)
filter_info = PathInfo(os.path.abspath(file))
return [(out.stage, filter_info)]
except OutputNotFoundError:
stages = self.collect(target, *args, **kwargs)
return [(stage, None) for stage in stages]
