excerpt_errata.html

﻿<html>
    <head>
        <meta charset="UTF-8">
        <title>Errata for Redis in Action</title>
        <style type="text/css">
            body {background-color: #eee;}
            #body {
                margin-left: auto;
                margin-right: auto;
                width: 800px;
                border: 1px solid #444;
                background-color: #eee;
                padding: 10px;
            }
            h2 {
                margin-left: 10px;
                border-top: 1px solid #444;
            }
            h3,h4 {margin-left: 20px;}
            p {margin-left: 30px;}
            blockquote {margin-left: 60px;}
            ul.ns {list-style-type: none;}
        </style>
    </head>
    <body>
        <div id="body">
            <h1>Errata for Redis in Action</h1>
            <img id="bookcover" width="150" height="188" src=" https://manning-content.s3.amazonaws.com/book/8/00f2522-76ce-4594-8564-541254e6d8f0/carlson.png" alt="booktitle">
            <p><i>Last updated December 14, 2015</i></p>
            <p><strong>You can view the most recent version of this errata <a href="http://htmlpreview.github.io/?https://github.com/josiahcarlson/redis-in-action/blob/master/excerpt_errata.html">here</a></strong></p>
            <h3>Known errata</h3>
            <ul class="ns">
                <li><a href="#2">Chapter 2</a><br/>
                    <ul class="ns">
                        <li><a href="#2.5">Section 2.5</a>, Page 35, <a href="#l2.10">Listing 2.10</a></li>
                    </ul>
                </li>
                <li><a href="#5">Chapter 5</a><br/>
                    <ul class="ns">
                        <li><a href="#5.1.2">Section 5.1.2</a>, Page 93, <a href="#l5.2">Listing 5.2</a></li>
                    </ul>
                </li>
                <li><a href="#6">Chapter 6</a><br/>
                    <ul class="ns">
                        <li><a href="#6.2.3">Section 6.2.3</a>, Page 121, <a href="#l6.9">Listing 6.9</a></li>
                        <li><a href="#6.5.2">Section 6.5.2</a>, Page 145, <a href="#l6.28">Listing 6.28</a></li>
                    </ul>
                </li>
                <li><a href="#7">Chapter 7</a><br/>
                    <ul class="ns">
                        <li><a href="#7.3.4">Section 7.3.4</a>, Page 177, <a href="#l7.15">Listing 7.15</a></li>
                    </ul>
                </li>
                <li><a href="#8">Chapter 8</a><br/>
                    <ul class="ns">
                        <li><a href="#8.1.1">Section 8.1.1</a>, Page 187, <a href="#l8.1">Listing 8.1</a></li>
                        <li><a href="#8.3">Section 8.3</a>, Page 190-191, <a href="#l8.4">Listing 8.4</a></li>
                        <li><a href="#8.3">Section 8.3</a>, Page 191-192, <a href="#l8.5">Listing 8.5</a></li>
                        <li><a href="#8.4">Section 8.4</a>, Page 194, <a href="#l8.8">Listing 8.8</a></li>
                        <li><a href="#8.5.3">Section 8.5.3</a>, Page 201-202, <a href="#l8.14">Listing 8.14</a></li>
                        <li><a href="#8.5.3">Section 8.5.3</a>, Page 205, <a href="#l8.19">Listing 8.19</a></li>
                    </ul>
                </li>
                <li><a href="#11">Chapter 11</a><br/>
                    <ul class="ns">
                        <li><a href="#11.3.2">Section 11.3.2</a>, Page 260-261, <a href="#l11.12">Listing 11.12</a></li>
                    </ul>
                </li>
            </ul>
            <h3>Recent updates</h3>
            <ul>
                <li>2015/12/14 - <a href="#5">Chapter 6</a>, <a href="#6.5.2">Section 6.5.2</a>, Page 145, <a href="#l6.28">Listing 6.28</a></li>
                <li>2015/10/10 - <a href="#5">Chapter 5</a>, <a href="#5.1.2">Section 5.1.2</a>, Page 93, <a href="#l5.2">Listing 5.2</a></li>
                <li>2015/04/22 - <a href="#11">Chapter 11</a>, <a href="#11.3.2">Section 11.3.2</a>, Page 260-261, <a href="#l11.12">Listing 11.12</a></li>
                <li>2015/04/22 - <a href="#7">Chapter 7</a>, <a href="#7.3.4">Section 7.3.4</a>, Page 177, <a href="#l7.15">Listing 7.15</a></li>
                <li>2015/04/22 - <a href="#6">Chapter 6</a>, <a href="#6.2.3">Section 6.2.3</a>, Page 121, <a href="#l6.9">Listing 6.9</a></li>
                <li>2015/04/13 - <a href="#8">Chapter 8</a>, <a href="#8.5.3">Section 8.5.3</a>, Page 201-202, <a href="#l8.14">Listing 8.14</a></li>
                <li>2015/04/13 - <a href="#8">Chapter 8</a>, <a href="#8.4">Section 8.4</a>, Page 194, <a href="#l8.8">Listing 8.8</a></li>
                <li>2015/04/13 - <a href="#8">Chapter 8</a>, <a href="#8.3">Section 8.3</a>, Page 191-192, <a href="#l8.5">Listing 8.5</a></li>
                <li>2015/04/13 - <a href="#8">Chapter 8</a>, <a href="#8.3">Section 8.3</a>, Page 190-191, <a href="#l8.4">Listing 8.4</a></li>
                <li>2015/04/13 - <a href="#8">Chapter 8</a>, <a href="#8.1.1">Section 8.1.1</a>, Page 187, <a href="#l8.1">Listing 8.1</a></li>
                <li>2015/04/10 - <a href="#6">Chapter 6</a>, <a href="#6.5.2">Section 6.5.2</a>, Page 145, <a href="#l6.28">Listing 6.28</a></li>
                <li>2014/07/30 - <a href="#8">Chapter 8</a>, <a href="#8.5.3">Section 8.5.3</a>, Page 205, <a href="#l8.19">Listing 8.19</a></li>
                <li>2014/05/04 - <a href="#2">Chapter 2</a>, <a href="#2.5">Section 2.5</a>, Page 35, <a href="#l2.10">Listing 2.10</a></li>
                <li>2014/02/16 - <a href="#6">Chapter 6</a>, <a href="#6.2.3">Section 6.2.3</a>, Page 121, <a href="#l6.9">Listing 6.9</a></li>
                <li>2013/08/24 - <a href="#6">Chapter 6</a>, <a href="#6.2.3">Section 6.2.3</a>, Page 121, <a href="#l6.9">Listing 6.9</a></li>
            </ul>

            <h2><a name=2>Chapter 2</a></h2>
            <h3><a name="2.5">Section 2.5</a>, Page 35, <a name="l2.10">Listing 2.10</a> (2014/05/04)</h3>
            <p>There is a <strong>bug</strong> on the third line of the <tt>rescale_viewed()</tt> function definition.</p>
            <p>The full function definition reads:</p>
            <blockquote>
<pre>def rescale_viewed(conn):
    while not QUIT:
        conn.zremrangebyrank('viewed:', 20000, -1)
        conn.zinterstore('viewed:', {'viewed:': .5})
        time.sleep(300)</pre>
            </blockquote>

            <p>The third line, updated inline, should read:</p>
            <blockquote>
<pre>def rescale_viewed(conn):
    while not QUIT:
        conn.zremrangebyrank('viewed:', 0, -20001)
        conn.zinterstore('viewed:', {'viewed:': .5})
        time.sleep(300)</pre>
            </blockquote>

            <p>You can see the updated code in-context by visiting: <a href="http://goo.gl/1hlS3m">http://goo.gl/1hlS3m</a></p>

            <h2><a name=5>Chapter 5</a></h2>
            <h3><a name="5.1.2">Section 5.1.2</a>, Page 93, <a name="l5.2">Listing 5.2</a> (2015/10/10)</h3>
            <p>There is a <strong>bug</strong> caused by a missing <tt>elif</tt> clause inside the try block of the <tt>log_common()</tt> function definition.</p>
            <p>The original full function definition reads:</p>
            <blockquote>
<pre>def log_common(conn, name, message, severity=logging.INFO, timeout=5):
    severity = str(SEVERITY.get(severity, severity)).lower()
    destination = 'common:%s:%s'%(name, severity)
    start_key = destination + ':start'
    pipe = conn.pipeline()
    end = time.time() + timeout
    while time.time() < end:
        try:
            pipe.watch(start_key)
            now = datetime.utcnow().timetuple()
            hour_start = datetime(*now[:4]).isoformat()

            existing = pipe.get(start_key)
            pipe.multi()
            if existing and existing < hour_start:
                pipe.rename(destination, destination + ':last')
                pipe.rename(start_key, destination + ':pstart')
                pipe.set(start_key, hour_start)

            pipe.zincrby(destination, message)
            log_recent(pipe, name, message, severity, pipe)
            return
        except redis.exceptions.WatchError:
            continue</pre>
            </blockquote>

            <p>And with the added elif block (prefixed by a comment line below), the function should read:</p>
            <blockquote>
<pre>def log_common(conn, name, message, severity=logging.INFO, timeout=5):
    severity = str(SEVERITY.get(severity, severity)).lower()
    destination = 'common:%s:%s'%(name, severity)
    start_key = destination + ':start'
    pipe = conn.pipeline()
    end = time.time() + timeout
    while time.time() < end:
        try:
            pipe.watch(start_key)
            now = datetime.utcnow().timetuple()
            hour_start = datetime(*now[:4]).isoformat()

            existing = pipe.get(start_key)
            pipe.multi()
            if existing and existing < hour_start:
                pipe.rename(destination, destination + ':last')
                pipe.rename(start_key, destination + ':pstart')
                pipe.set(start_key, hour_start)
            # add the following two lines
            elif not existing:
                pipe.set(start_key, hour_start)

            pipe.zincrby(destination, message)
            log_recent(pipe, name, message, severity, pipe)
            return
        except redis.exceptions.WatchError:
            continue</pre>
            </blockquote>

            <p>You can see the change in-context by visiting: <a href="https://goo.gl/UN5kMw">https://goo.gl/UN5kMw</a></p>


            <h2><a name=6>Chapter 6</a></h2>
            <h3><a name="6.2.3">Section 6.2.3</a>, Page 121, <a name="l6.9">Listing 6.9</a> (2013/08/24, 2014/02/16, 2014/04/22)</h3>
            <p>There is one printing errata (wrong in the printed version, correct in the source code), three other bugs, and three cleanups in this listing for <tt>purchase_item_with_lock()</tt>.</p>
            <p><strong>Printing errata</strong>: there is a missing line between the two lines that read:</p>
            <blockquote>
<pre>locked = acquire_lock(conn, market)
    return False</pre>
            </blockquote>

            <p>With the missing line replaced, it should read:</p>
            <blockquote>
<pre>locked = acquire_lock(conn, market)
if not locked:
    return False</pre>
            </blockquote>

            <p>You can see the code in-context by visiting: <a href="http://goo.gl/QpcbuC">http://goo.gl/QpcbuC</a></p>

            <p><strong>Bug</strong>: there is an extra <tt>pipe.watch(buyer)</tt> call that breaks the remaining behavior, which can be removed. The lines that read:</p>
            <blockquote>
<pre>try:
    pipe.watch(buyer)
    pipe.zscore("market:", item)
    pipe.hget(buyer, 'funds')</pre>
            </blockquote>

            <p>Should instead read:</p>
            <blockquote>
<pre>try:
    pipe.zscore("market:", item)
    pipe.hget(buyer, 'funds')</pre>
            </blockquote>

            <p>Further, there are missing <tt>'funds'</tt> arguments to the <tt>pipe.hincrby()</tt> calls later, *and* a misnamed argument. The lines that read:</p>
            <blockquote>
<pre>pipe.hincrby(seller, int(price))
pipe.hincrby(buyerid, int(-price))</pre>
            </blockquote>

            <p>Should instead read:</p>
            <blockquote>
<pre>pipe.hincrby(seller, 'funds', int(price))
pipe.hincrby(buyer, 'funds', int(-price))</pre>
            </blockquote>

            <p>Note the addition of the <tt>'funds'</tt> arguments and the renaming of the <tt>buyerid</tt> argument to <tt>buyer</tt>.</p>
            <p>You can see the change in-context by visiting: <a href="http://goo.gl/kQltLd">http://goo.gl/kQltLd</a></p>

            <p><strong>Bug</strong>: the calls to <tt>acquire_lock()</tt> and <tt>release_lock()</tt> reference the variable <tt>market</tt> when they should instead pass the string <tt>'market:'</tt>.

            <p>There are also several additional (but unnecessary) <strong>cleanups</strong> that can be done to this listing. These changes include: 1) removing the <tt>while</tt> loop, 2) removing the <tt>try/except</tt> clause, 3) removing the unnecessary <tt>pipe.unwatch()</tt> call.</p>
            <p>You can see the change for these cleanups and the fixed <tt>market</tt> -&gt; <tt>'market:'</tt> reference inline by visiting: <a href="http://goo.gl/LxGvV8">http://goo.gl/LxGvV8</a></p>

            <hr>
            <h3><a name="6.5.2">Section 6.5.2</a>, Page 145, <a name="l6.28">Listing 6.28</a> (2015/04/10)</h3>
            <p>There are two <strong>bug</strong>s on the last line of the <tt>leave_chat()</tt> function. The first bug is where the <tt>oldest</tt> argument to the <tt>conn.zremrangebyscore()</tt> call should really be <tt>oldest[0][1]</tt> (this was discovered on 2015-04-10). The second bug is what reads as <tt>'chat:'</tt> should read <tt>'msgs:'</tt> (this was discovered on 2015-12-14). The lines that read:</p>
            <blockquote>
<pre>        'chat:' + chat_id, 0, 0, withscores=True)
        conn.zremrangebyscore('chat:' + chat_id, 0, oldest)</pre>
            </blockquote>

            <p>Should instead read:</p>
            <blockquote>
<pre>        'chat:' + chat_id, 0, 0, withscores=True)
        conn.zremrangebyscore('msgs:' + chat_id, 0, oldest[0][1])</pre>
            </blockquote>

            <p>You can see the code in-context by visiting: <a href="https://goo.gl/T8zCp2">https://goo.gl/T8zCp2</a></p>

            <h2><a name=7>Chapter 7</a></h2>
            <h3><a name="7.3.4">Section 7.3.4</a>, Page 177, <a name="l7.15">Listing 7.15</a> (2014/04/22)</h3>
            <p>There is a <strong>printing errata</strong> and <strong>bug</strong> in the <tt>record_click()</tt> function.</p>
            <p><strong>Printing errata</strong> and <strong>bug</strong>: in the printed book, there is a missing <tt>else:</tt> line between the two <tt>pipeline.incr()</tt> calls below, and the first <tt>pipeline.incr()</tt> call is missing a <tt>'%s'</tt> for the string templating to work. So lines 13-15 in the book:</p>
            <blockquote>
<pre>    if action and type == 'cpa':
        pipeline.incr('type:cpa:actions:' % type)
        pipeline.incr('type:%s:clicks:' % type)</pre>
            </blockquote>

            <p>Should instead read:</p>
            <blockquote>
<pre>    if action and type == 'cpa':
        pipeline.incr('type:%s:actions:' % type)
    else:
        pipeline.incr('type:%s:clicks:' % type)</pre>
            </blockquote>

            <p>You can see the change in-context with the <tt>else:</tt> clause by visiting: <a href="http://goo.gl/XXiNTD">http://goo.gl/XXiNTD</a></p>


            <h2><a name=8>Chapter 8</a></h2>
            <h3><a name="8.1.1">Section 8.1.1</a>, Page 187, <a name="l8.1">Listing 8.1</a> (2015/04/13)</h3>
            <p>There is a <strong>bug</strong> in the <tt>create_user()</tt> between lines 7 and 8, where there is a missing <tt>release_lock()</tt> call prior to return. In practice, this may cause a 1 second delay in deleting a status message for a user if someone were trying to create a new account with the same user name.</p>

            <p>Lines 7-8 of <tt>create_user()</tt> originally read:</p>
            <blockquote>
<pre>    if conn.hget('users:', llogin):
        return None</pre>
            </blockquote>

            <p>Lines 7-8 should be replaced with these 3 lines:</p>
            <blockquote>
<pre>    if conn.hget('users:', llogin):
        release_lock(conn, 'user:' + llogin, lock)
        return None</pre>
            </blockquote>

            <p>You can see the code in-context by visiting: <a href="http://goo.gl/quX7ee">http://goo.gl/quX7ee</a></p>

            <hr>
            <h3><a name="8.3">Section 8.3</a>, Page 190-191, <a name="l8.4">Listing 8.4</a> (2015/04/13)</h3>
            <p>There is a <strong>race condition</strong> in the <tt>follow_user()</tt> function, caused by directly setting the size of the following/follower lists instead of incrementally changing them. This is fixed as part of other updates to <tt>follow_user()</tt> in section 10.3.3, listing 10.12, and we are just bringing a couple of these changes back to chapter 8.</p>

            <p>Lines 12-21 originally read:</p>
            <blockquote>
<pre>    pipeline.zadd(fkey1, other_uid, now)
    pipeline.zadd(fkey2, uid, now)
    pipeline.zcard(fkey1)
    pipeline.zcard(fkey2)
    pipeline.zrevrange('profile:%s'%other_uid,
        0, HOME_TIMELINE_SIZE-1, withscores=True)
    following, followers, status_and_score = pipeline.execute()[-3:]

    pipeline.hset('user:%s'%uid, 'following', following)
    pipeline.hset('user:%s'%other_uid, 'followers', followers)</pre>
            </blockquote>

            <p>With 2 lines deleted and 2 lines changed, lines 12-19 should now read:</p>
            <blockquote>
<pre>    pipeline.zadd(fkey1, other_uid, now)
    pipeline.zadd(fkey2, uid, now)
    pipeline.zrevrange('profile:%s'%other_uid,
        0, HOME_TIMELINE_SIZE-1, withscores=True)
    following, followers, status_and_score = pipeline.execute()[-3:]

    pipeline.hincrby('user:%s'%uid, 'following', int(following))
    pipeline.hincrby('user:%s'%other_uid, 'followers', int(followers))</pre>
            </blockquote>

            <p>You can see the code in-context by visiting: <a href="http://goo.gl/mvmxwV">http://goo.gl/mvmxwV</a></p>

            <hr>
            <h3><a name="8.3">Section 8.3</a>, Page 191-192, <a name="l8.5">Listing 8.5</a> (2015/04/13)</h3>
            <p>There is a <strong>race condition</strong> in the <tt>unfollow_user()</tt> function, caused by directly setting the size of the following/follower lists instead of incrementally changing them. This is the exact same bug that occurs in the <tt>follow_user()</tt> function, and has the same fix.</p>

            <p>Lines 12-21 originally read:</p>
            <blockquote>
<pre>    pipeline.zrem(fkey1, other_uid, now)
    pipeline.zrem(fkey2, uid, now)
    pipeline.zcard(fkey1)
    pipeline.zcard(fkey2)
    pipeline.zrevrange('profile:%s'%other_uid,
        0, HOME_TIMELINE_SIZE-1, withscores=True)
    following, followers, status_and_score = pipeline.execute()[-3:]

    pipeline.hset('user:%s'%uid, 'following', following)
    pipeline.hset('user:%s'%other_uid, 'followers', followers)  </pre>
            </blockquote>

            <p>With 2 lines deleted and 2 lines changed, lines 12-19 should now read:</p>
            <blockquote>
<pre>    pipeline.zrem(fkey1, other_uid, now)
    pipeline.zrem(fkey2, uid, now)
    pipeline.zrevrange('profile:%s'%other_uid,
        0, HOME_TIMELINE_SIZE-1, withscores=True)
    following, followers, status_and_score = pipeline.execute()[-3:]

    pipeline.hincrby('user:%s'%uid, 'following', int(following))
    pipeline.hincrby('user:%s'%other_uid, 'followers', int(followers))</pre>
            </blockquote>

            <p>You can see the code in-context by visiting: <a href="http://goo.gl/mxcUqZ">http://goo.gl/mxcUqZ</a></p>

            <hr>
            <h3><a name="8.4">Section 8.4</a>, Page 194, <a name="l8.8">Listing 8.8</a> (2015/04/13)</h3>
            <p>There is a <strong>bug</strong> in the <tt>delete_status()</tt> function between lines 7 and 8, where there is a missing <tt>release_lock()</tt> call prior to return. In practice, this may cause a 1 second delay in deleting a status message for a user if someone tried to delete a status message they didn't own.</p>

            <p>Lines 7-8 of <tt>delete_status(conn, uid, status_id)</tt> originally read:</p>
            <blockquote>
<pre>    if conn.hget(key, 'uid') != str(uid):
        return None</pre>
            </blockquote>

            <p>Lines 7-8 should be replaced with these 3 lines:</p>
            <blockquote>
<pre>    if conn.hget(key, 'uid') != str(uid):
        release_lock(conn, key, lock)
        return None</pre>
            </blockquote>

            <p>You can see the code in-context by visiting: <a href="http://goo.gl/gL6dPG">http://goo.gl/gL6dPG</a></p>

            <hr>
            <h3><a name="8.5.3">Section 8.5.3</a>, Page 201-202, <a name="l8.14">Listing 8.14</a> (2015/04/13)</h3>

            <p>There is a <strong>bug</strong> in the streaming <tt>delete_status()</tt> function between lines 7 and 8, where there is a missing <tt>release_lock()</tt> call prior to return. In practice, this may cause a 1 second delay in deleting a status message for a user if someone tried to delete a status message they didn't own.</p>

            <p>Lines 7-8 of <tt>delete_status(conn, uid, status_id)</tt> originally read:</p>
            <blockquote>
<pre>    if conn.hget(key, 'uid') != str(uid):
        return None</pre>
            </blockquote>

            <p>Lines 7-8 should be replaced with these 3 lines:</p>
            <blockquote>
<pre>    if conn.hget(key, 'uid') != str(uid):
        release_lock(conn, key, lock)
        return None</pre>
            </blockquote>

            <p>You can see the code in-context by visiting: <a href="http://goo.gl/DzweRD">http://goo.gl/DzweRD</a></p>

            <hr>
            <h3>Section 8.5.3, Page 205, <a name="l8.19">Listing 8.19</a> (2014/07/30)</h3>
            <p>There is a <strong>bug</strong> in the <tt>FollowFilter()</tt> function on lines 2, 4, and 10 of the code listing, where the argument <tt>names</tt> is overridden by an empty set, which prevents the <tt>FollowFilter()</tt> call from actually following anyone.</p>

            <p>The function originally read:</p>
            <blockquote>
<pre>def FollowFilter(names):
    names = set()
    for name in names:
        names.add('@' + name.lower().lstrip('@'))

    def check(status):
        message_words = set(status['message'].lower().split())
        message_words.add('@' + status['login'].lower())

        return message_words & names
    return check</pre>
            </blockquote>

            <p>The function (with comments here to show the changes) now reads:</p>
            <blockquote>
<pre>def FollowFilter(names):
    nset = set() # first fix here
    for name in names:
        nset.add('@' + name.lower().lstrip('@')) # second fix here

    def check(status):
        message_words = set(status['message'].lower().split())
        message_words.add('@' + status['login'].lower())

        return message_words & nset # third fix here
    return check</pre>
            </blockquote>

            <p>The diff (which might be easier to read) can be found: <a href="http://goo.gl/Opbp13">http://goo.gl/Opbp13</a></p>
            <p>And the code in-context can be seen: <a href="http://goo.gl/GkqXp5">http://goo.gl/GkqXp5</a></p>


            <h2><a name=11>Chapter 11</a></h2>
            <h3><a name="11.3.2">Section 11.3.2</a>, Page 260-261, <a name="l11.12">Listing 11.12</a> (2015/04/22)</h3>
            <p>This code listing is a copy of the listing from <a href="#6">Chapter 6</a>, <a href="#6.2.3">Section 6.2.3</a>, Page 121, <a href="#l6.9">Listing 6.9</a>, and has all of the same issues except for the printing errata and the <tt>market</tt> variable reference bug.</p>

        </div>
    </body>
</html>