tiny-config: SamplesSheet: validate group names to avoid namespace collisions in R

tests/unit_tests_configuration.py

@@ -7,6 +7,7 @@
 
 from tiny.rna.configuration import Configuration, SamplesSheet, PathsFile
 from unit_test_helpers import csv_factory, paths_template_file, make_paths_file
+from tiny.rna.util import r_reserved_keywords
 
 
 class BowtieIndexesTest(unittest.TestCase):
@@ -209,6 +210,22 @@ def test_validate_fastq_filepath(self):
                 patch('tiny.rna.configuration.os.path.isfile', return_value=True):
             SamplesSheet('mock_filename')
 
+    """Does validate_r_safe_sample_groups detect group names that will cause namespace collisions in R?"""
+
+    def test_validate_r_safe_sample_groups(self):
+        non_alphanum_chars = [bad.join(('a', 'b')) for bad in "~!@#$%^&*()+-=`<>?/,:;\"'[]{}\| \t\n\r\f\v"]
+        leading_dot_number = [".0", "X.0"]
+
+        for bad in [non_alphanum_chars, leading_dot_number]:
+            msg = " ≈ ".join(bad)
+            with self.assertRaisesRegex(AssertionError, msg):
+                SamplesSheet.validate_r_safe_sample_groups(dict.fromkeys(bad))
+
+        for kwd in r_reserved_keywords:
+            bad = (kwd, kwd + '.')
+            msg = " ≈ ".join(bad)
+            with self.assertRaisesRegex(AssertionError, msg):
+                SamplesSheet.validate_r_safe_sample_groups(dict.fromkeys(bad))
 
 class PathsFileTest(unittest.TestCase):
 

tiny/rna/configuration.py

@@ -8,12 +8,12 @@
 import re
 
 from pkg_resources import resource_filename
-from collections import Counter, OrderedDict
+from collections import Counter, OrderedDict, defaultdict
 from typing import Union, Any, Optional, List
 from glob import glob
 
 from tiny.rna.counter.validation import GFFValidator
-from tiny.rna.util import get_timestamp
+from tiny.rna.util import get_timestamp, get_r_safename
 
 
 class ConfigBase:
@@ -586,6 +586,8 @@ def validate_control_group(self, is_control: bool, group: str):
 
     @staticmethod
     def validate_deseq_compatibility(sample_groups: Counter) -> bool:
+        SamplesSheet.validate_r_safe_sample_groups(sample_groups)
+
         total_samples = sum(sample_groups.values())
         total_coefficients = len(sample_groups)
         degrees_of_freedom = total_samples - total_coefficients
@@ -598,6 +600,21 @@ def validate_deseq_compatibility(sample_groups: Counter) -> bool:
         else:
             return True
 
+    @staticmethod
+    def validate_r_safe_sample_groups(sample_groups: Counter):
+        """Determine the "syntactically valid" translation of each group name to ensure
+        that two groups won't share the same name once translated in tiny-deseq.r"""
+
+        safe_names = defaultdict(list)
+        for group in sample_groups:
+            safe_names[get_r_safename(group)].append(group)
+
+        collisions = [' ≈ '.join(cluster) for cluster in safe_names.values() if len(cluster) > 1]
+
+        assert len(collisions) == 0, \
+            "The following group names are too similar and will cause a namespace collision in R:\n" \
+            + '\n'.join(collisions)
+
     @staticmethod
     def get_sample_basename(filename):
         root, _ = os.path.splitext(filename)

tiny/rna/util.py

@@ -95,17 +95,33 @@ def make_filename(args, ext='.csv'):
     return '_'.join([str(chnk) for chnk in args if chnk is not None]) + ext
 
 
+r_reserved_keywords = [
+    "if", "else", "repeat", "while", "function",
+    "for", "in", "next", "break", "TRUE", "FALSE",
+    "NULL", "Inf", "NaN", "NA", "NA_integer_",
+    "NA_real_", "NA_complex_", "NA_character_"]
+
+
 def get_r_safename(name: str) -> str:
     """Converts a string to a syntactically valid R name
 
-    This can be used to match names along axes of DataFrames produced by R,
-    assuming that the R script takes no measures to preserve names itself.
+    This can be used as the Python equivalent of R's make.names() function.
     https://stat.ethz.ch/R-manual/R-devel/library/base/html/make.names.html
     """
 
+    # If the name starts with a non-letter character or a dot
+    # followed by a number, the character "X" is prepended
     leading_char = lambda x: re.sub(r"^(?=[^a-zA-Z.]+|\.\d)", "X", x)
-    special_char = lambda x: re.sub(r"[^a-zA-Z0-9_.]", ".", x)
-    return special_char(leading_char(name))
+
+    # If the name contains characters that aren't (locale based) letters,
+    # numbers, dot, or underscore, the characters are replaced with a dot
+    special_char = lambda x: re.sub(r"[^\w.]", ".", x)
+
+    # If the name contains R keywords, a dot is appended to the keyword
+    reserved = "|".join(r_reserved_keywords)
+    reserved_wrd = lambda x: re.sub(fr"^({reserved})$", r'\1.', x)
+
+    return reserved_wrd(special_char(leading_char(name)))
 
 
 class ReadOnlyDict(dict):