Merge pull request #201 from jmchilton/idc_2

Enhancements to the IDC scripts

Author: mvdbeek

setup.py

@@ -43,7 +43,11 @@ def get_var(var_name):
         install_tool_deps=ephemeris.install_tool_deps:main
         install-tool-deps=ephemeris.install_tool_deps:main
         set-library-permissions=ephemeris.set_library_permissions:main
-"""
+        _idc-lint=ephemeris._idc_lint:main
+        _idc-split-data-manager-genomes=ephemeris._idc_split_data_manager_genomes:main
+        _idc-data-managers-to-tools=ephemeris._idc_data_managers_to_tools:main
+        """
+
 PACKAGE_DATA = {
     # Be sure to update MANIFEST.in for source dist.
 }

src/ephemeris/__init__.py

@@ -1,3 +1,5 @@
+import os
+
 import yaml
 from bioblend import galaxy
 
@@ -11,6 +13,14 @@
 RAW_CONTENT_URL = f"https://raw.github.com/{PROJECT_USERAME}/{PROJECT_NAME}/master/"
 
 
+def get_or_create_history(history_name: str, gi: galaxy.GalaxyInstance):
+    histories = gi.histories.get_histories(name=history_name)
+    if histories:
+        return histories[0]
+    else:
+        return gi.histories.create_history(name=history_name)
+
+
 def check_url(url, log=None):
     if not url.startswith("http"):
         if log:
@@ -32,7 +42,7 @@ def get_galaxy_connection(args, file=None, log=None, login_required=True):
 
     url = args.galaxy or file_content.get("galaxy_instance")
     galaxy_url = check_url(url, log)
-    api_key = args.api_key or file_content.get("api_key")
+    api_key = args.api_key or file_content.get("api_key") or os.environ.get("EPHEMERIS_API_KEY")
 
     if args.user and args.password:
         return galaxy.GalaxyInstance(url=galaxy_url, email=args.user, password=args.password)

src/ephemeris/_config_models.py

@@ -0,0 +1,85 @@
+from pathlib import Path
+from typing import (
+    Dict,
+    List,
+    Optional,
+    Union,
+)
+
+import yaml
+from pydantic import (
+    BaseModel,
+    Extra,
+)
+
+StrOrPath = Union[Path, str]
+
+
+class RepositoryInstallTarget(BaseModel):
+    name: str
+    owner: str
+    tool_shed_url: Optional[str]
+    tool_panel_section_id: Optional[str]
+    tool_panel_section_label: Optional[str]
+    revisions: Optional[List[str]]
+    install_tool_dependencies: Optional[bool]
+    install_repository_dependencies: Optional[bool]
+    install_resolver_dependencies: Optional[bool]
+
+
+class RepositoryInstallTargets(BaseModel):
+    """ """
+
+    api_key: Optional[str]
+    galaxy_instance: Optional[str]
+    tools: List[RepositoryInstallTarget]
+
+
+class DataManager(BaseModel, extra=Extra.forbid):
+    tags: List[str]
+    tool_id: str
+
+
+class DataManagers(BaseModel, extra=Extra.forbid):
+    __root__: Dict[str, DataManager]
+
+
+class Genome(BaseModel):
+    id: str  # The unique id of the data in Galaxy
+    description: str  # The description of the data, including its taxonomy, version and date
+    dbkey: Optional[str]
+    source: Optional[str]  # The source of the data. Can be: 'ucsc', an NCBI accession number or a URL to a fasta file.
+
+    # The following fields are currently purely for human consumption and unused by
+    # IDC infrastructure.
+    doi: Optional[str]  # Any DOI associated with the data
+    blob: Optional[str]  # A blob for any other pertinent information
+    checksum: Optional[str]  # A SHA256 checksum of the original
+    version: Optional[str]  # Any version information associated with the data
+
+    # Description of actions (data managers) to run on target genome.
+    indexers: Optional[
+        List[str]
+    ]  # indexers to run - keyed on repository name - see data_managers.yml for how to resolve these to tools
+    skiplist: Optional[List[str]]  # unimplemented: but if we implement classes of indexers, these will be ones to skip
+
+
+class Genomes(BaseModel):
+    genomes: List[Genome]
+
+
+def _read_yaml(path: StrOrPath):
+    with open(path) as f:
+        return yaml.safe_load(f)
+
+
+def read_data_managers(path: StrOrPath) -> DataManagers:
+    return DataManagers(__root__=_read_yaml(path))
+
+
+def read_genomes(path: StrOrPath) -> Genomes:
+    return Genomes(**_read_yaml(path))
+
+
+def read_tools(path: StrOrPath) -> RepositoryInstallTargets:
+    return RepositoryInstallTargets(**_read_yaml(path))

src/ephemeris/_idc_data_managers_to_tools.py

@@ -0,0 +1,105 @@
+#!/usr/bin/env python
+"""Helper script for IDC - not yet meant for public consumption.
+
+This script takes a data_managers.yml configuration describing the
+set of data managers the IDC configuration targets and builds a
+a tools.yml file from it for use with shed_tools.
+"""
+import argparse
+import logging
+from typing import (
+    Dict,
+    List,
+    NamedTuple,
+)
+
+import yaml
+
+from ._config_models import (
+    read_data_managers,
+    RepositoryInstallTargets,
+)
+from .common_parser import (
+    add_log_file_argument,
+    add_verbosity_argument,
+)
+from .ephemeris_log import (
+    disable_external_library_logging,
+    setup_global_logger,
+)
+
+
+class DataManager(NamedTuple):
+    tool_id: str
+    repository_name: str
+    tags: List[str]
+
+
+def read_data_managers_configuration(path: str) -> Dict[str, DataManager]:
+    raw_data_managers = read_data_managers(path)
+    data_managers: Dict[str, DataManager] = {}
+    for repository_name, data_manager_configuration in raw_data_managers.__root__.items():
+        data_manager = DataManager(
+            tool_id=data_manager_configuration.tool_id,
+            repository_name=repository_name,
+            tags=data_manager_configuration.tags or [],
+        )
+        data_managers[repository_name] = data_manager
+    return data_managers
+
+
+def build_shed_install_conf(path: str) -> dict:
+    data_managers = read_data_managers_configuration(path)
+    tools = []
+    for data_manager in data_managers.values():
+        tool_id = data_manager.tool_id
+        tool_id_parts = tool_id.split("/")
+        repo_owner = tool_id_parts[2]
+        repo_name = tool_id_parts[3]
+        entry = {
+            "name": repo_name,
+            "owner": repo_owner,
+            "tool_panel_section_label": None,
+            "tool_shed_url": "toolshed.g2.bx.psu.edu",
+        }
+        tools.append(entry)
+    tools_yaml = {"tools": tools}
+    return tools_yaml
+
+
+def write_shed_install_conf(data_manager_conf_path: str, output_path: str) -> None:
+    tools_yaml = build_shed_install_conf(data_manager_conf_path)
+
+    # validate generated dict to ensure we're writing out valid file
+    RepositoryInstallTargets(**tools_yaml)
+
+    with open(output_path, "w") as f:
+        yaml.safe_dump(tools_yaml, f)
+
+
+def _parser():
+    """returns the parser object."""
+
+    parser = argparse.ArgumentParser(add_help=False)
+    general_group = parser.add_argument_group("General options")
+    add_verbosity_argument(general_group)
+    add_log_file_argument(general_group)
+    parser.add_argument("--data-managers-conf", default="data_managers.yml")
+    parser.add_argument("--shed-install-output-conf", default="tools.yml")
+    return parser
+
+
+def main():
+    disable_external_library_logging()
+    parser = _parser()
+    args = parser.parse_args()
+    log = setup_global_logger(name=__name__, log_file=args.log_file)
+    if args.verbose:
+        log.setLevel(logging.DEBUG)
+    else:
+        log.setLevel(logging.INFO)
+    write_shed_install_conf(args.data_managers_conf, args.shed_install_output_conf)
+
+
+if __name__ == "__main__":
+    main()

src/ephemeris/_idc_lint.py

@@ -0,0 +1,44 @@
+import os
+from pathlib import Path
+
+import yaml
+
+from ._config_models import (
+    read_data_managers,
+    read_genomes,
+)
+
+
+def read_yaml(path: Path):
+    with open(path) as f:
+        return yaml.safe_load(f)
+
+
+def lint_idc_directory(directory: Path):
+    genomes_path = directory / "genomes.yml"
+    data_managers_path = directory / "data_managers.yml"
+    assert genomes_path.exists()
+    assert data_managers_path.exists()
+    data_managers = read_data_managers(data_managers_path).__root__
+    genomes = read_genomes(genomes_path)
+
+    for data_manager in data_managers.values():
+        data_manager_tool_id = data_manager.tool_id
+        if not data_manager_tool_id.startswith("toolshed.g2.bx.psu.edu/"):
+            raise Exception(
+                f"Expected a data manager repository from main Galaxy tool shed but discovered tool ID {data_manager_tool_id}"
+            )
+
+    for genome in genomes.genomes:
+        print(genome)
+        for indexer in genome.indexers or []:
+            if indexer not in data_managers:
+                raise Exception(f"Failed to find data manager {indexer} referenced for genome {genome}")
+
+
+def main():
+    lint_idc_directory(Path(os.curdir))
+
+
+if __name__ == "__main__":
+    main()