Add CSV support; right now hard-coded for kaggle acher linux-kernel-size

2 files changed, 31 insertions, 3 deletions

diff --git a/bin/analyze-log.py b/bin/analyze-log.py
index d6ae27a..1deb438 100755
--- a/bin/analyze-log.py
+++ b/bin/analyze-log.py
@@ -21,15 +21,18 @@ import time
 
 
 def parse_logfile(filename):
-    lf = dfatool.utils.Logfile()
+    if ".csv" in filename:
+        loader = dfatool.utils.CSVfile()
+    else:
+        loader = dfatool.utils.Logfile()
 
     if filename.endswith("xz"):
         import lzma
 
         with lzma.open(filename, "rt") as f:
-            return lf.load(f)
+            return loader.load(f)
     with open(filename, "r") as f:
-        return lf.load(f)
+        return loader.load(f)
 
 
 def main():
diff --git a/lib/utils.py b/lib/utils.py
index af705b9..61cb6f1 100644
--- a/lib/utils.py
+++ b/lib/utils.py
@@ -33,6 +33,31 @@ class NpEncoder(json.JSONEncoder):
             return super(NpEncoder, self).default(obj)
 
 
+class CSVfile:
+    def __init__(self):
+        pass
+
+    def load(self, f):
+        observations = list()
+        for lineno, line in enumerate(f):
+            if lineno == 0:
+                param_names = line.split(",")[1:-1]
+                attr_names = line.removesuffix("\n").split(",")[-1:]
+            else:
+                param_values = list(map(soft_cast_int, line.split(",")[1:-1]))
+                attr_values = list(
+                    map(soft_cast_float, line.removesuffix("\n").split(",")[-1:])
+                )
+                observations.append(
+                    {
+                        "name": "CSVFile",
+                        "param": dict(zip(param_names, param_values)),
+                        "attribute": dict(zip(attr_names, attr_values)),
+                    }
+                )
+        return observations
+
+
 class Logfile:
     def __init__(self):
         pass