Page de statistiques

tools/stats/words_per_article.py

@@ -0,0 +1,88 @@
+#!/usr/bin/env python3
+
+import sys
+import json
+import os
+from collections import defaultdict
+
+CURRENT_DIR = os.path.dirname(os.path.abspath(__file__))
+PARENT_DIR = os.path.abspath(os.path.join(CURRENT_DIR, os.pardir))
+if CURRENT_DIR not in sys.path:
+    sys.path.append(CURRENT_DIR)
+if PARENT_DIR not in sys.path:
+    sys.path.append(PARENT_DIR)
+
+from common import load_articles, MONTH_LABELS, write_result  # noqa: E402
+
+
+def main():
+    try:
+        payload = json.load(sys.stdin)
+    except Exception as exc:  # noqa: BLE001
+        print(f"Failed to read JSON: {exc}", file=sys.stderr)
+        sys.exit(1)
+
+    content_dir = payload.get("contentDir") or "content"
+    output_path = payload.get("outputPath")
+    public_path = payload.get("publicPath")
+
+    articles = load_articles(content_dir)
+
+    buckets = defaultdict(lambda: {"words": 0, "count": 0})
+    years = set()
+    total_words = 0
+    total_articles = 0
+
+    for article in articles:
+        date = article.get("date")
+        if not date:
+            continue
+        year = date.year
+        month = date.month
+        key = (year, month)
+        years.add(year)
+        buckets[key]["words"] += article.get("wordCount") or 0
+        buckets[key]["count"] += 1
+        total_words += article.get("wordCount") or 0
+        total_articles += 1
+
+    month_numbers = list(range(1, 13))
+    labels = [MONTH_LABELS[m - 1] for m in month_numbers]
+    sorted_years = sorted(years)
+
+    series = []
+    for year in sorted_years:
+        values = []
+        for month in month_numbers:
+            entry = buckets.get((year, month))
+            if not entry or entry["count"] == 0:
+                values.append(0)
+            else:
+                values.append(round(entry["words"] / entry["count"]))
+        series.append({"label": str(year), "values": values})
+
+    average = round(total_words / total_articles) if total_articles > 0 else 0
+
+    try:
+        from render_stats_charts import render_words_per_article, setup_rcparams
+    except ImportError as exc:  # noqa: BLE001
+        print(f"Failed to import renderer: {exc}", file=sys.stderr)
+        sys.exit(1)
+
+    setup_rcparams()
+    render_words_per_article({"labels": labels, "series": series, "title": "Moyenne de mots par article (par mois)"}, output_path)
+
+    write_result(
+        {
+            "image": public_path,
+            "meta": {
+                "average": average,
+                "articles": total_articles,
+            },
+        }
+    )
+
+
+if __name__ == "__main__":
+    main()
+