Source code for

"""Utilities for dealing with continous integration systems."""

import copy
import math
import os

import yaml

from planemo import (
from planemo.shed import REPO_METADATA_FILES

[docs]def filter_paths(ctx, raw_paths, path_type="repo", **kwds): """Filter ``paths``. ``path_type`` is ``repo`` or ``file``. """ cwd = os.getcwd() filter_kwds = copy.deepcopy(kwds) changed_in_commit_range = kwds.get("changed_in_commit_range", None) diff_paths = None if changed_in_commit_range is not None: diff_files = git.diff(ctx, cwd, changed_in_commit_range) if path_type == "repo": diff_dirs = {os.path.dirname(p) for p in diff_files} diff_paths = set() for diff_dir in diff_dirs: diff_path = metadata_file_in_path(diff_dir) if diff_path: diff_paths.add(diff_path) else: diff_paths = diff_files unique_paths = {os.path.relpath(p, cwd) for p in raw_paths} if diff_paths is not None: unique_paths = unique_paths.intersection(diff_paths) filtered_paths = sorted(io.filter_paths(unique_paths, cwd=cwd, **filter_kwds)) excluded_paths = sorted(set(unique_paths) - set(filtered_paths)) if excluded_paths: ctx.log("List of excluded paths: %s" % excluded_paths) path_count = len(filtered_paths) chunk_size = (1.0 * path_count) / kwds["chunk_count"] chunk = kwds["chunk"] chunked_paths = [] for i, path in enumerate(filtered_paths): if int(math.floor(i / chunk_size)) == chunk: chunked_paths.append(path) return chunked_paths
[docs]def metadata_file_in_path(diff_dir): while diff_dir: for metadata_file in REPO_METADATA_FILES: if os.path.isfile(os.path.join(diff_dir, metadata_file)): return diff_dir diff_dir = os.path.dirname(diff_dir)
[docs]def group_paths(paths): repos = {} for path in paths: repo = os.path.split(path)[0] if repo not in repos: repos[repo] = [] repos[repo].append(path) return [" ".join(repos[_]) for _ in repos]