Check if phenotype exists before creating one

1 files changed, 16 insertions, 3 deletions

diff --git a/gn3/computations/gemma.py b/gn3/computations/gemma.py
index ea0f86a..42b4b36 100644
--- a/gn3/computations/gemma.py
+++ b/gn3/computations/gemma.py
@@ -1,9 +1,11 @@
 """Procedures related gemma computations"""
+import os
 import random
 import string
 
 from base64 import b64encode
 from hashlib import md5
+from typing import List
 
 def generate_random_n_string(n_length: int) -> str:
     """Generate a random string that is N chars long"""
@@ -16,11 +18,22 @@ def generate_hash_of_string(unhashed_str: str) -> str:
 the end"""
     hashed_str = md5(unhashed_str.encode("utf-8")).digest()
     return b64encode(hashed_str).decode("utf-8").replace("==", "")
+
+
 def generate_pheno_txt_file(trait_filename: str,
-                            values: str,
+                            values: List,
                             tmpdir: str = "/tmp") -> str:
-    """Given VALUES, and TMPDIR, generate a valide traits file"""
-    trait_filename += f"_{generate_random_n_string(6)}"
+    """Given VALUES, and TMPDIR, generate a valid traits file"""
+    if not os.path.isdir(f"{tmpdir}/gn2/"):
+        os.mkdir(f"{tmpdir}/gn2/")
+    ext = trait_filename.partition(".")[-1]
+    if ext:
+        trait_filename = trait_filename.replace(f".{ext}", "")
+        ext = f".{ext}"
+    trait_filename += f"_{generate_hash_of_string(''.join(values))}{ext}"
+    # Early return if this already exists!
+    if os.path.isfile(f"{tmpdir}/gn2/{trait_filename}"):
+        return f"{tmpdir}/gn2/{trait_filename}"
     with open(f"{tmpdir}/gn2/{trait_filename}", "w") as _file:
         for value in values:
             if value == "x":