From 9c88194369311d285eca07e05da982aa1382fec9 Mon Sep 17 00:00:00 2001
From: Marty Oehme <contact@martyoeh.me>
Date: Tue, 30 Sep 2025 18:24:17 +0200
Subject: [PATCH] add generation script for unique installs

---
 code/unique.py | 56 ++++++++++++++++++++++++++++++++++++++++++++++++++
 justfile       |  5 +++++
 2 files changed, 61 insertions(+)
 create mode 100644 code/unique.py

diff --git a/code/unique.py b/code/unique.py
new file mode 100644
index 0000000..eddf419
--- /dev/null
+++ b/code/unique.py
@@ -0,0 +1,56 @@
+import csv
+import json
+from pathlib import Path
+
+
+def unique_install_csv(input_dir: Path, output_dir: Path) -> None:
+    output_file = output_dir / "unique_installs.csv"
+    with open(output_file, "w") as fw:
+        writer = csv.writer(fw)
+        writer.writerow(["date", "unique"])
+
+        for j in input_dir.glob("*.json"):
+            with open(j) as fr:
+                date = j.stem
+                data: dict[str, object] = {}
+                try:
+                    data = json.load(fr)
+                except json.JSONDecodeError:
+                    print(f"WARN: Could not decode JSON data for file {j}")
+                    continue
+
+                if "UniqueInstalls" not in data or not isinstance(
+                    data["UniqueInstalls"], int
+                ):
+                    print(
+                        f"WARN: No correct json structure containing 'UniqueInstalls' field in file {j}"
+                    )
+                    continue
+
+                p_date = date
+                p_count = data["UniqueInstalls"]
+                writer.writerow([p_date, p_count])
+
+
+def ensure_dirs(input_dir: Path, output_dir: Path):
+    if not input_dir.is_dir():
+        raise ValueError
+    output_dir.mkdir(exist_ok=True, parents=True)
+
+
+def main(input: str, output: str) -> None:
+    input_dir = Path(input)
+    output_dir = Path(output)
+    ensure_dirs(input_dir, output_dir)
+    unique_install_csv(input_dir, output_dir)
+
+
+if __name__ == "__main__":
+    import sys
+
+    if not len(sys.argv) == 3:
+        print("Please provide exactly one input directory and one output directory.")
+        sys.exit(1)
+    inp = sys.argv[1]
+    out = sys.argv[2]
+    main(inp, out)
diff --git a/justfile b/justfile
index 411b5b6..a40ee33 100644
--- a/justfile
+++ b/justfile
@@ -1,2 +1,7 @@
+default: kernels unique
+
 kernels:
     python code/kernels.py input output
+
+unique:
+    python ./code/unique.py input output