winval

winval

Workflow inputs validation python library

• Currently, supports WDL workflows.
• Constraints on inputs can be written as specially tagged (#@wv) comments in the WDL inputs section.
• The syntax of constraints is a python-similar DSL, with ANTLR defined grammar.
• The WDL + inputs.json can be validated using a script/function before submitting a workflow.

Installation:

pip install winval

Usage:

validate_wdl_constraints

From python

from winval.validate_wdl_constraints import run_winval
is_validated = run_winval(wdl_file, json_file)

From unix command-line:

python winval/validate_wdl_constraints.py --wdl <wdl_file> --json <json_file>

cloud_files_validater

Make sure google-storage permissions are equivalent to the batch workers permissions

From python

from winval.cloud_files_validator import CloudFilesValidator
is_validated = CloudFilesValidator(args.wdl, args.json).validate()

From unix command-line:

python winval/cloud_files_validator.py --wdl <wdl_file> --json <json_file>

WDL constraints example

workflow MyWorkflow {

  input {
    File file
    Int c
    File* opt_file_1
    File* opt_file_2
    Array[File] files
    Array[File] index_files
    MyStruct struct_instance

    #@wv defined(opt_file_1) <-> defined(opt_file_2)
    #@wv defined(opt_file_1) -> c > 1
    #@wv len(files) == len(index_files)
    #@wv len(files) >= 0
    #@wv len(index_files) >= 0
    #@wv c <= 1 and c >= 0
    #@wv suffix(file) == ".fasta"
    #@wv suffix(files) <= {".bam", ".cram"} 
    #@wv prefix(index_files) == files
    #@wv len(struct_instance['field_a']) > 0
  }
  ...
}
  
  struct MyStruct{
     String field_a,
     String field_b
  }

Generate parsers from grammar:

cd <project_root>/winval
antlr4 -Dlanguage=Python3 winval.g4 -visitor -o antlr

Available atomic expressions:

• int: 5
• float: 5.6
• bool: True, False
• str: "some string", 'some_string'
• workflow_variable: my_var
• evaluates to value given by json conf, or None if not defined in json
• empty_set: {}

Available python operators

• +,-,*,**,/,&,|,%
• and, or, in
• <, <=, ==, >=, >, !=
• Notice the following useful operators work for sets:
  • -: set subtraction
  • &: set intersection
  • | : is set union
  • <=: subset

Available python functions

• len(), not()
• basename(), splitext() (from os.path)

Available convenience functions and operators:

• x <-> y: if-and-only-if logical operator (if x is true if and only if y is true)
• x -> y: implies logical operator (if x is true then y should be true)
• defined(x): if x a defined variable
• prefix(x): return path prefix of String/File or list of prefixes for
• suffix(x): return path suffix of String/File or set of suffixes for Array[String]/Array[File]