Add visitor tracking code

config-sample.py

@@ -134,4 +134,5 @@ load_plugin("rophako.modules.blog")
 load_plugin("rophako.modules.photo")
 load_plugin("rophako.modules.comment")
 load_plugin("rophako.modules.emoticons")
-load_plugin("rophako.modules.contact")
+load_plugin("rophako.modules.contact")
+load_plugin("rophako.modules.tracking")

rophako/app.py

@@ -26,6 +26,7 @@ BLUEPRINT_PATHS = []
 import config
 from rophako import __version__
 from rophako.plugin import load_plugin
+import rophako.model.tracking as Tracking
 import rophako.utils
 
 app.DEBUG      = config.DEBUG
@@ -78,7 +79,8 @@ def before_request():
             "uid": 0,
             "name": "Guest",
             "role": "user",
-        }
+        },
+        "tracking": Tracking.track_visit(request, session),
     }
 
     # Default session vars.
@@ -158,4 +160,4 @@ def not_found(error):
 
 @app.errorhandler(403)
 def forbidden(error):
-    return render_template('errors/403.html', **g.info), 403
+    return render_template('errors/403.html', **g.info), 403

rophako/jsondb.py

@@ -19,7 +19,7 @@ redis_client = None
 cache_lifetime = 60*60 # 1 hour
 
 
-def get(document):
+def get(document, cache=True):
     """Get a specific document from the DB."""
     logger.debug("JsonDB: GET {}".format(document))
 
@@ -32,7 +32,7 @@ def get(document):
     stat = os.stat(path)
 
     # Do we have it cached?
-    data = get_cache(document)
+    data = get_cache(document) if cache else None
     if data:
         # Check if the cache is fresh.
         if stat.st_mtime > get_cache(document+"_mtime"):
@@ -45,12 +45,13 @@ def get(document):
     data = read_json(path)
 
     # Cache and return it.
-    set_cache(document, data, expires=cache_lifetime)
-    set_cache(document+"_mtime", stat.st_mtime, expires=cache_lifetime)
+    if cache:
+        set_cache(document, data, expires=cache_lifetime)
+        set_cache(document+"_mtime", stat.st_mtime, expires=cache_lifetime)
     return data
 
 
-def commit(document, data):
+def commit(document, data, cache=True):
     """Insert/update a document in the DB."""
 
     # Need to create the file?
@@ -69,8 +70,9 @@ def commit(document, data):
                 os.mkdir(segment, 0o755)
 
     # Update the cached document.
-    set_cache(document, data, expires=cache_lifetime)
-    set_cache(document+"_mtime", time.time(), expires=cache_lifetime)
+    if cache:
+        set_cache(document, data, expires=cache_lifetime)
+        set_cache(document+"_mtime", time.time(), expires=cache_lifetime)
 
     # Write the JSON.
     write_json(path, data)
@@ -82,6 +84,7 @@ def delete(document):
     if os.path.isfile(path):
         logger.info("Delete DB document: {}".format(path))
         os.unlink(path)
+        del_cache(document)
 
 
 def exists(document):
@@ -216,4 +219,4 @@ def del_cache(key):
     """Delete a cached item."""
     key = config.REDIS_PREFIX + key
     client = get_redis()
-    client.delete(key)
+    client.delete(key)

rophako/model/photo.py

@@ -3,7 +3,7 @@
 """Photo album models."""
 
 import os
-from flask import g, request
+from flask import g
 import time
 import requests
 from PIL import Image
@@ -12,7 +12,7 @@ import random
 
 import config
 import rophako.jsondb as JsonDB
-from rophako.utils import sanitize_name
+from rophako.utils import sanitize_name, remote_addr
 from rophako.log import logger
 
 # Maps the friendly names of photo sizes with their pixel values from config.
@@ -523,7 +523,7 @@ def process_photo(form, filename):
         }
 
     index["albums"][album][key] = dict(
-        ip=request.remote_addr,
+        ip=remote_addr(),
         author=g.info["session"]["uid"],
         uploaded=int(time.time()),
         caption=form.get("caption", ""),

rophako/model/tracking.py

@@ -0,0 +1,175 @@
+# -*- coding: utf-8 -*-
+
+"""Visitor tracking models."""
+
+import time
+import requests
+
+import rophako.jsondb as JsonDB
+from rophako.utils import remote_addr, pretty_time, server_name
+
+def track_visit(request, session):
+    """Main logic to track and log visitor details."""
+
+    # Get their tracking cookie value. The value will either be their HTTP
+    # referrer (if exists and valid) or else a "1".
+    cookie = session.get("tracking")
+    addr   = remote_addr()
+    values = dict() # Returnable traffic values
+
+    # Log hit counts. We need four kinds:
+    # - Unique today   - Unique total
+    # - Hits today     - Hits total
+    today = pretty_time("%Y-%m-%d", time.time())
+    files = {
+        "unique/{}".format(today) : "unique_today",
+        "unique/total"            : "unique_total",
+        "hits/{}".format(today)   : "hits_today",
+        "hits/total"              : "hits_total",
+    }
+
+    # Go through the hit count files. Update them only if their tracking
+    # cookie was not present.
+    for file, key in files.items():
+        dbfile = "traffic/{}".format(file)
+        if file.startswith("hits"):
+            # Hit file is just a simple counter.
+            db = dict(hits=0)
+            if JsonDB.exists(dbfile):
+                db = JsonDB.get(dbfile)
+
+            # Update it?
+            if not cookie:
+                db["hits"] += 1
+                JsonDB.commit(dbfile, db)
+
+            # Store the copy.
+            values[key] = db["hits"]
+        else:
+            # Unique file is a collection of IP addresses.
+            db = dict()
+            if JsonDB.exists(dbfile):
+                db = JsonDB.get(dbfile)
+
+            # Update with their IP?
+            if not cookie and not addr in db:
+                db[addr] = time.time()
+                JsonDB.commit(dbfile, db)
+
+            # Store the copy.
+            values[key] = len(db.keys())
+
+    # Log their HTTP referrer.
+    referrer = "1"
+    if request.referrer:
+        # Branch and check this.
+        referrer = log_referrer(request, request.referrer)
+        if not referrer:
+            # Wasn't a valid referrer.
+            referrer = "1"
+
+    # Set their tracking cookie.
+    if not cookie:
+        cookie = referrer
+        session["tracking"] = cookie
+
+    return values
+
+
+def log_referrer(request, link):
+    """Double check the referring URL."""
+
+    # Ignore if same domain.
+    if link.startswith(request.url_root):
+        print "Referrer is same host!"
+        return None
+
+    # See if the URL really links back to us.
+    hostname = server_name()
+    r = requests.get(link)
+    if hostname in r.text:
+        # Log it.
+        db = list()
+        if JsonDB.exists("traffic/referrers"):
+            # Don't cache the result -- the list can get huge!
+            db = JsonDB.get("traffic/referrers", cache=False)
+        db.append(link)
+        JsonDB.commit("traffic/referrers", db, cache=False)
+        return link
+
+    return None
+
+
+def get_visitor_details():
+    """Retrieve detailed visitor information for the frontend."""
+    result = {
+        "traffic": [],                      # Historical traffic data
+        "most_unique": [ "0000-00-00", 0 ], # Day with the most unique
+        "most_hits":   [ "0000-00-00", 0 ], # Day with the most hits
+        "oldest": None,                     # Oldest day on record.
+    }
+
+    # List all the documents.
+    hits = JsonDB.list_docs("traffic/hits")
+    for date in sorted(hits):
+        if date == "total": continue
+        if not result["oldest"]:
+            result["oldest"] = date
+
+        # Get the DBs.
+        hits_db = JsonDB.get("traffic/hits/{}".format(date), cache=False)
+        uniq_db = JsonDB.get("traffic/unique/{}".format(date), cache=False)
+
+        # Most we've seen?
+        if hits_db["hits"] > result["most_hits"][1]:
+            result["most_hits"] = [ date, hits_db["hits"] ]
+        if len(uniq_db.keys()) > result["most_unique"][1]:
+            result["most_unique"] = [ date, len(uniq_db.keys()) ]
+
+        result["traffic"].append(dict(
+            date=date,
+            hits=hits_db["hits"],
+            unique=len(uniq_db.keys()),
+        ))
+
+    return result
+
+
+def get_referrers(recent=25):
+    """Retrieve the referrer details. Returns results in this format:
+
+    ```
+    {
+        referrers: [
+            ["http://...", 20], # Pre-sorted by number of hits
+        ],
+        recent: [ recent list ]
+    }
+    ```
+    """
+    db = []
+    if JsonDB.exists("traffic/referrers"):
+        db = JsonDB.get("traffic/referrers", cache=False)
+
+    # Count the links.
+    unique = dict()
+    for link in db:
+        if not link in unique:
+            unique[link] = 1
+        else:
+            unique[link] += 1
+
+    # Sort them by popularity.
+    result = dict(
+        referrers=[],
+        recent=[],
+    )
+
+    sorted_links = sorted(unique.keys(), key=lambda x: unique[x], reverse=True)
+    for link in sorted_links:
+        result["referrers"].append([ link, unique[link] ])
+
+    recent = 0 - recent
+    result["recent"] = db[recent:]
+
+    return result

rophako/modules/admin/__init__.py

@@ -3,7 +3,6 @@
 """Endpoints for admin functions."""
 
 from flask import g, Blueprint, request, redirect, url_for, session, flash
-import re
 
 import rophako.model.user as User
 from rophako.modules.account import validate_create_form
@@ -180,4 +179,4 @@ def unimpersonate():
     )
 
     flash("No longer impersonating.")
-    return redirect(url_for("index"))
+    return redirect(url_for("index"))

rophako/modules/blog/__init__.py

@@ -2,10 +2,8 @@
 
 """Endpoints for the web blog."""
 
-from flask import Blueprint, g, request, redirect, url_for, session, flash, make_response
-import re
+from flask import Blueprint, g, request, redirect, url_for, flash, make_response
 import datetime
-import calendar
 import time
 from xml.dom.minidom import Document
 
@@ -13,7 +11,8 @@ import rophako.model.user as User
 import rophako.model.blog as Blog
 import rophako.model.comment as Comment
 import rophako.model.emoticons as Emoticons
-from rophako.utils import template, render_markdown, pretty_time, login_required
+from rophako.utils import (template, render_markdown, pretty_time,
+    login_required, remote_addr)
 from rophako.plugin import load_plugin
 from rophako.log import logger
 from config import *
@@ -264,7 +263,7 @@ def update():
                     avatar     = g.info["avatar"],
                     categories = tags,
                     privacy    = g.info["privacy"],
-                    ip         = request.remote_addr,
+                    ip         = remote_addr(),
                     emoticons  = g.info["emoticons"],
                     comments   = g.info["comments"],
                     format     = g.info["format"],

rophako/modules/comment/__init__.py

@@ -8,7 +8,8 @@ import time
 
 import rophako.model.user as User
 import rophako.model.comment as Comment
-from rophako.utils import template, pretty_time, login_required, sanitize_name
+from rophako.utils import (template, pretty_time, login_required, sanitize_name,
+    remote_addr)
 from rophako.plugin import load_plugin
 from rophako.log import logger
 from config import *
@@ -48,7 +49,7 @@ def preview():
         Comment.add_comment(
             thread=thread,
             uid=g.info["session"]["uid"],
-            ip=request.remote_addr,
+            ip=remote_addr(),
             time=int(time.time()),
             image=gravatar,
             name=form["name"],
@@ -216,4 +217,4 @@ def get_comment_form(form):
         contact = request.form.get("contact", ""),
         message = request.form.get("message", ""),
         subscribe = request.form.get("subscribe", "false"),
-    )
+    )

rophako/modules/contact/__init__.py

@@ -2,12 +2,9 @@
 
 """Endpoints for contacting the site owner."""
 
-from flask import Blueprint, g, request, redirect, url_for, session, flash
-import re
-import time
+from flask import Blueprint, request, redirect, url_for, flash
 
-from rophako.utils import template, send_email
-from rophako.log import logger
+from rophako.utils import template, send_email, remote_addr
 from config import *
 
 mod = Blueprint("contact", __name__, url_prefix="/contact")
@@ -59,7 +56,7 @@ Subject: {subject}
 
 {message}""".format(
             site_name=SITE_NAME,
-            ip=request.remote_addr,
+            ip=remote_addr(),
             ua=request.user_agent.string,
             referer=request.headers.get("Referer", ""),
             name=name,
@@ -70,4 +67,4 @@ Subject: {subject}
     )
 
     flash("Your message has been delivered.")
-    return redirect(url_for("index"))
+    return redirect(url_for("index"))

rophako/modules/tracking/__init__.py

@@ -0,0 +1,43 @@
+# -*- coding: utf-8 -*-
+
+"""Endpoints for visitor tracking functions."""
+
+from flask import Blueprint, g
+import re
+
+import rophako.model.tracking as Tracking
+from rophako.utils import template
+
+mod = Blueprint("tracking", __name__, url_prefix="/tracking")
+
+
+@mod.route("/")
+def index():
+    return template("tracking/index.html")
+
+
+@mod.route("/visitors")
+def visitors():
+    g.info["history"] = Tracking.get_visitor_details()
+    return template("tracking/visitors.html")
+
+
+@mod.route("/referrers")
+def referrers():
+    g.info["referrers"] = Tracking.get_referrers()
+
+    # Filter some of the links.
+    for i, link in enumerate(g.info["referrers"]["referrers"]):
+        # Clean up useless Google links.
+        if "google" in link[0] and re.search(r'/(?:imgres|url|search|translate\w+)?/', link[0]):
+            g.info["referrers"]["referrers"][i] = None
+
+    # Make the links word-wrap properly.
+    filtered = [
+        [ re.sub(r'(.{20})', r'\1<wbr>', x[0]), x[1] ]
+        for x in g.info["referrers"]["referrers"]
+        if x is not None
+    ]
+    g.info["referrers"]["referrers"] = filtered
+
+    return template("tracking/referrers.html")

rophako/modules/tracking/templates/tracking/index.html

@@ -0,0 +1,13 @@
+{% extends "layout.html" %}
+{% block title %}Visitor Tracking{% endblock %}
+
+{% block content %}
+
+<h1>Visitor Tracking</h1>
+
+<ul>
+    <li><a href="{{ url_for('tracking.visitors') }}">Unique Visitors &amp; Visits</a></li>
+    <li><a href="{{ url_for('tracking.referrers') }}">HTTP Referrers</a></li>
+</ul>
+
+{% endblock %}

rophako/modules/tracking/templates/tracking/referrers.html

@@ -0,0 +1,44 @@
+{% extends "layout.html" %}
+{% block title %}Referring URLs{% endblock %}
+
+{% block content %}
+
+<h1>Referring URLs</h1>
+
+This table lists the HTTP referrers to this site, in order of popularity. For
+the most recent 25 links, see <a href="#recent">the end of this page</a>.<p>
+
+<div style="height: 450px; overflow: auto">
+<table class="table" width="100%">
+    <thead>
+        <tr>
+            <th width="40">Hits</th>
+            <th>Query</th>
+        </tr>
+    </thead>
+    <tbody>
+        {% for link in referrers["referrers"] %}
+            {% if link %}
+                <tr>
+                    <td align="center" valign="top">
+                        {{ link[1] }}
+                    </td>
+                    <td align="left" valign="top" style="position: relative">
+                        {{ link[0]|safe }}
+                    </td>
+                </tr>
+            {% endif %}
+        {% endfor %}
+    </tbody>
+</table>
+</div>
+
+<h2 id="recent">25 Most Recent Links</h2>
+
+<ol>
+{% for item in referrers["recent"] %}
+    <li>{{ item }}</li>
+{% endfor %}
+</ol>
+
+{% endblock %}

rophako/modules/tracking/templates/tracking/visitors.html

@@ -0,0 +1,54 @@
+{% extends "layout.html" %}
+{% block title %}Visitor History{% endblock %}
+
+{% block content %}
+
+<h1>Visitor History</h1>
+
+Unique visitors and hit counts have been logged on this site since
+{{ history["oldest"] }}.<p>
+
+The most unique visitors on this site in one day has been
+{{ history["most_unique"][1] }} on {{ history["most_unique"][0] }}. The most
+hits total in one day has been {{ history["most_hits"][1] }} on
+{{ history["most_hits"][0] }}.<p>
+
+Here is a full list of hits over time. Percentages are relative to the current
+records.<p>
+
+<table class="table" width="100%" border="0" cellspacing="2" cellpadding="2">
+    <thead>
+        <tr>
+            <th width="20">Date</th>
+            <th>Graph</th>
+            <th width="250">Details</th>
+        </tr>
+    </thead>
+    <tbody>
+        {% for date in history["traffic"]|reverse %}
+        <tr>
+            <td align="center" valign="middle" rowspan="2">
+                {{ date["date"] }}
+            </td>
+            <td align="left" valign="middle">
+                {% set pct = (date["unique"] / history["most_unique"][1]) * 100 %}
+                <div class="visitor-graph unique" style="width: {{ pct|int }}%"></div>
+            </td>
+            <td align="left" valign="middle">
+                Unique: {{ date["unique"] }} ({{ pct|int }}%)
+            </td>
+        </tr>
+        <tr>
+            <td align="left" valign="middle">
+                {% set pct = (date["unique"] / history["most_unique"][1]) * 100 %}
+                <div class="visitor-graph hits" style="width: {{ pct|int }}%"></div>
+            </td>
+            <td align="left" valign="middle">
+                Hits: {{ date["hits"] }} ({{ pct|int }}%)
+            </td>
+        </tr>
+        {% endfor %}
+    </tbody>
+</table>
+
+{% endblock %}

rophako/utils.py

@@ -11,6 +11,7 @@ import importlib
 import smtplib
 import markdown
 import json
+import urlparse
 
 from rophako.log import logger
 from config import *
@@ -211,6 +212,20 @@ def include(endpoint, *args, **kwargs):
     return html
 
 
+def remote_addr():
+    """Retrieve the end user's remote IP address."""
+
+    # TODO: eventually support configurations with X-Forwarded-For, but for
+    # now at least we're centralizing this in one spot.
+    return request.remote_addr
+
+
+def server_name():
+    """Get the server's hostname."""
+    urlparts = list(urlparse.urlparse(request.url_root))
+    return urlparts[1]
+
+
 def pretty_time(time_format, unix):
     """Pretty-print a time stamp."""
     date = datetime.datetime.fromtimestamp(unix)