Analytics page

.github/workflows/prod.yml

@@ -73,6 +73,13 @@ jobs:
           retention-days: 14
           archive: false
 
+      - name: Upload analytics page artifact
+        uses: actions/upload-artifact@v7
+        with:
+          path: dashboard.html
+          retention-days: 14
+          archive: false
+
       - name: Upload API Log artifact
         if: '!cancelled()' #This ensures this step runs even if the previous steps failed only if manually cancelled it doesnt run
         uses: actions/upload-artifact@v7

database.py

@@ -0,0 +1,29 @@
+import json
+import traceback
+import logging
+
+from pathlib import Path
+
+logger = logging.getLogger(__name__)
+
+
+def read_database(database_path):
+    path = Path(database_path)
+    if not path.exists() or path.stat().st_size == 0:
+        return {}
+
+    try:
+        with path.open() as database_file:
+            return json.load(database_file)
+    except (json.JSONDecodeError, ValueError) as exc:
+        logger.error("%s:%s", type(exc).__name__, exc)
+        traceback.print_exc()
+        return {}
+
+
+def write_database(database_path, data):
+    path = Path(database_path)
+    temporary_path = path.with_name(f"{path.name}.tmp")
+    with temporary_path.open("w") as database_file:
+        json.dump(data, database_file, indent=4)
+    temporary_path.replace(path)

main.py

@@ -13,6 +13,8 @@
 import requests
 
 from adoption_sources import SourceManual, SourceRescueGroups
+from database import read_database, write_database
+from metrics_dashboard import dashboard
 from metric_collectors.bluesky import CollectorBluesky
 from metric_collectors.instagram import CollectorInstagram
 from metric_collectors.mastodon import CollectorMastodon
@@ -103,6 +105,7 @@ def run(sources, posters, collectors=None, database_path="database.json"):
         record_publish_results(pet, published_results, database_path=database_path)
 
     collect_metrics(collectors or [], database_path=database_path)
+    dashboard(database_path=database_path)
     return results
 
 
@@ -131,7 +134,7 @@ def publish_posts(pet, posters):
 
 
 def pick_pet(pets, database_path="database.json"):
-    data = _read_database(database_path)
+    data = read_database(database_path)
     posted_pet_ids = {
         posted_pet["pet_id"] for posted_pet in data.get("posted_pets", [])
     }
@@ -149,7 +152,7 @@ def pick_pet(pets, database_path="database.json"):
 
 
 def record_publish_results(pet, results, database_path="database.json"):
-    data = _read_database(database_path)
+    data = read_database(database_path)
     posted_pets = data.setdefault("posted_pets", [])
     posts = data.setdefault("posts", [])
     posted_at = datetime.now(timezone.utc).isoformat()
@@ -182,12 +185,12 @@ def record_publish_results(pet, results, database_path="database.json"):
         for item in posts
         if datetime.fromisoformat(item["posted_at"]) >= cutoff
     ]
-    _write_database(database_path, data)
+    write_database(database_path, data)
 
 
 def collect_metrics(collectors, database_path="database.json", window_days=14):
     try:
-        data = _read_database(database_path)
+        data = read_database(database_path)
         posts = data.get("posts", [])
         if not posts:
             return
@@ -228,33 +231,11 @@ def collect_metrics(collectors, database_path="database.json", window_days=14):
                 )
 
         if updated:
-            _write_database(database_path, data)
+            write_database(database_path, data)
     except Exception as exc:
         logger.error("Metric collection failed: %s", exc)
 
 
-def _read_database(database_path):
-    path = Path(database_path)
-    if not path.exists() or path.stat().st_size == 0:
-        return {}
-
-    try:
-        with path.open() as database_file:
-            return json.load(database_file)
-    except (json.JSONDecodeError, ValueError) as exc:
-        logger.error("%s:%s", type(exc).__name__, exc)
-        traceback.print_exc()
-        return {}
-
-
-def _write_database(database_path, data):
-    path = Path(database_path)
-    temporary_path = path.with_name(f"{path.name}.tmp")
-    with temporary_path.open("w") as database_file:
-        json.dump(data, database_file, indent=4)
-    temporary_path.replace(path)
-
-
 # Slack incoming-webhook messages have a ~40k-char limit; cap the traceback
 # well below that so the post stays readable and is never rejected.
 MAX_TRACEBACK_CHARS = 2500

metrics_dashboard.py

@@ -0,0 +1,173 @@
+import pandas as pd
+import plotly.express as px
+from database import read_database
+
+
+def empty_dashboard_html(message):
+    return f"""<!DOCTYPE html>
+    <html>
+    <head>
+        <title>Cute Pets Boston Top Pets</title>
+        <style>
+            body {{ font-family: Arial, sans-serif; margin: 40px; text-align: center; background-color: #f9f9f9; }}
+            h1 {{ color: #333; }}
+        </style>
+    </head>
+    <body>
+        <h1>🐶 Cute Pets Boston Top Pets 🐶</h1>
+        <p>{message}</p>
+    </body>
+    </html>
+    """
+
+
+def dashboard(html_file_path="dashboard.html", database_path="database.json"):
+    # 1. Fetch data
+    data = read_database(database_path)
+    posts = data.get("posts", [])
+    posted_pets = data.get("posted_pets", [])
+
+    if not posts or not posted_pets:
+        with open(html_file_path, "w", encoding="utf-8") as f:
+            f.write(empty_dashboard_html("No analytics data available yet."))
+        return
+
+    df_pets = pd.DataFrame(posted_pets)
+    df_posts = pd.json_normalize(
+        posts,
+        record_path=["metrics"],  # Unpacks the nested metrics array
+        meta=["pet_id", "platform", "post_id", "post_url"],
+    )
+
+    # Merge pets metadata with detailed post metrics
+    df_metrics = pd.merge(df_pets, df_posts, on="pet_id", how="left")
+
+    # Posts with no recorded metrics yet leave the metric columns absent
+    missing = [col for col in ("likes", "reposts", "comments") if col not in df_metrics.columns]
+    df_metrics = df_metrics.reindex(
+        columns=[*df_metrics.columns, *missing]
+    )
+
+    if not df_metrics["likes"].notna().any():
+        with open(html_file_path, "w", encoding="utf-8") as f:
+            f.write(empty_dashboard_html("No metric data available yet."))
+        return
+
+    # 2. Get unique platforms including 'All Platforms'
+    platforms_html = ["All Platforms"] + sorted(
+        [
+            str(p)
+            for p in df_metrics["platform"].dropna().unique()
+            if p != "All Platforms"
+        ]
+    )
+
+    # 3. Build options HTML for the dropdown
+    dropdown_options = "".join(
+        [f'<option value="{p}">{p}</option>' for p in platforms_html]
+    )
+
+    # 4. Generate HTML blocks for each platform and metric
+    sections_html = ""
+    for p in platforms_html:
+      if p == "All Platforms":
+        filtered_platform_temp = df_metrics
+      else:
+        filtered_platform_temp = df_metrics[df_metrics["platform"] == p]
+
+      platform_charts_html = ""
+
+      for col in ["likes", "reposts", "comments"]:
+        temp = (
+            filtered_platform_temp[["post_id", "pet_id", "name", col]]
+            .groupby(["post_id", "pet_id", "name"])
+            .max()
+            .reset_index()
+        )
+
+        agg_temp = (
+            temp[["pet_id", "name", col]]
+            .groupby(["pet_id", "name"])[col]
+            .agg(["max", "sum"])
+            .reset_index()
+        )
+
+        temp1 = agg_temp.sort_values("max", ascending=False).head(10)
+        temp2 = agg_temp.sort_values("sum", ascending=False).head(10)
+
+        metric_label = col.capitalize()
+
+        fig_max = px.bar(
+            data_frame=temp1,
+            x="name",
+            y="max",
+            title=f"Top 10 Pets by Max {metric_label} [{p}]",
+            labels={"name": "Pet", "max": f"Max {metric_label}"},
+        )
+        fig_sum = px.bar(
+            data_frame=temp2,
+            x="name",
+            y="sum",
+            title=f"Top 10 Pets by Total {metric_label} [{p}]",
+            labels={"name": "Pet", "sum": f"Total {metric_label}"},
+        )
+
+        platform_charts_html += f"""
+            <div class="chart-container">{fig_max.to_html(full_html=False, include_plotlyjs='cdn' if (p == platforms_html[0] and col == 'likes') else False)}</div>
+            <div class="chart-container">{fig_sum.to_html(full_html=False, include_plotlyjs=False)}</div>
+            """
+
+      display_style = "block" if p == platforms_html[0] else "none"
+      sections_html += f"""
+        <div class="platform-section" id="section-{p}" style="display: {display_style};">
+            {platform_charts_html}
+        </div>
+        """
+
+    # 5. Assemble full HTML document with styling and switching logic
+    html_content = f"""<!DOCTYPE html>
+    <html>
+    <head>
+        <title>Cute Pets Boston Top Pets</title>
+        <style>
+            body {{ font-family: Arial, sans-serif; margin: 40px; text-align: center; background-color: #f9f9f9; }}
+            h1 {{ color: #333; }}
+            .dropdown-container {{ margin-bottom: 30px; }}
+            select {{ padding: 8px 15px; font-size: 16px; border-radius: 4px; border: 1px solid #ccc; }}
+            .chart-container {{ margin-bottom: 30px; background: white; padding: 20px; border-radius: 8px; box-shadow: 0 2px 4px rgba(0,0,0,0.1); display: inline-block; width: 80%; max-width: 800px; }}
+        </style>
+        <script>
+            function filterPlatform(selectedPlatform) {{
+                var sections = document.getElementsByClassName('platform-section');
+                for (var i = 0; i < sections.length; i++) {{
+                    sections[i].style.display = 'none';
+                }}
+                var activeSection = document.getElementById('section-' + selectedPlatform);
+                if (activeSection) {{
+                    activeSection.style.display = 'block';
+                }}
+            }}
+        </script>
+    </head>
+    <body>
+        <h1>🐶 Cute Pets Boston Top Pets 🐶</h1>
+        <div class="dropdown-container">
+            <label for="platform-select"><strong>Select Platform: </strong></label>
+            <select id="platform-select" onchange="filterPlatform(this.value)">
+                {dropdown_options}
+            </select>
+        </div>
+        {sections_html}
+    </body>
+    </html>
+    """
+
+    # 6. Write out file
+    with open(html_file_path, "w", encoding="utf-8") as f:
+      f.write(html_content)
+
+    print(
+        "Successfully generated dashboard.html with all platforms, max/sum stats,"
+        " and metrics!"
+    )
+

requirements.txt

@@ -47,4 +47,6 @@ sortedcontainers==2.4.0
 typing_extensions==4.15.0
 urllib3==2.6.3
 pytest
+pandas
+plotly
 wheel==0.47.0