Historical errata for The Cucumber Book

Remove “try to” from “The —dry-run switch tells Cucumber to parse (it will tell you if your Gherkin isn’t valid) but not try to execute the feature.” This seems to indicate an ambiguity in that in some circumstances, the feature may still execute.

“Each feature typically has somewhere between a half and a two dozen sce- narios, each describing different examples of how that feature should behave in different circumstances.” The “a two dozen” is incorrect (the “a” may be removed). Although even after that adjustment, the sentence doesn’t read very fluidly — “a half scenario”.

You have already activated i18n 0.5.0, but your Gemfile requires i18n 0.6.0. Consider using bundle exec. (Gem::LoadError)
This error has blocked me from proceeding with the scenario. I have removed all from the beginning page 130 (8.1) and followed the instruction but have had this same error show up again. How or what have I done to create this problem.

The example shows “Given I have $200 in my account” but all wording about the step uses “$100” value.

All of chapter 4 uses the incorrect version of “withdraw”, opting to use the more medical “withdrawal”.

No text for footnote 1; “walking skeleton”

With the existing page breaks, the blank line in the Feature block is adjacent to the page break between pg 32 and 33.

In the two sample code blocks, the first amount.to_i was replaced with amount, but the second was left as amount.to_i

“you’ll learn how to wok with libraries like FactoryGirl”

Should be “work” instead of “wok”

“are totally irrellevant to our See messages”

Irrelevant is misspelled

In the Matt Says: “really helped thet team to see”

the team

directory name “step_definitions” is misspelled in:

adding.feature
…
step_defintions/
calculator_steps.rb
…

Doesn’t show “1 scenario (1 pending)” because it doesn’t match step definition.

(cash_withdrawal.feature) Given I have $200 in my account
(steps.rb) Given /^I have deposited \\$(\\d+) in my account$/ do |amount|

Typo in “exammples” (extra ‘m’)

P113 talks about running ‘bundle’ but doesn’t mention having to install the Bundler gem first.

For save_and_open_page, this requires the Launchy gem to be installed. It would be useful for people tell new Cucumber users about this.

“The option don’t matter too much” – missing ’s’ after option

require ‘open3’
I am on windows. Seems like open3 is for UNIX/LINUX. I have installed win32-open3 gem. ( Does this book supposed to provide information for windows users regarding what they supposed to do instead of assuming all users are Linux based?) . I changed the above to require ‘win32-open3’ - it did not work, and tried it with require ‘win32/open3’ - it did not work.

In Figure 5 Transforms: How Do They Work? Two instances of Tranform should be Transform.

Why leave the “amount.to_i” in the assertion line for the balance?

This sentence, located mid-way down the page, does not make sense: The Gherkin features call step definitions which provide the link between to the application being built.

The regular expressions on PDF pages 44 to 46 are missing the ^ and $ anchors.

Near the end of the page, the — before _{slow likely should be --tags on the line that reads: cucumber --tags breakfast —}@slow

The ‘Transfer funds from savings into checking account’ scenario code fragment in section 4.3 Multiple Captures has extra ‘I’s after the ’Then’ and ‘And’ clauses. Should be:

Scenario: Transfer funds from savings into checking account
Given I have deposited $10 in my Checking Account
And I have deposited $500 in my Savings Account
When I transfer $500 from my Savings Account into my Checking Account
Then the balance of the Checking Account should be $510
And the balance of the Savings Account should be $0

At the end of the page, typo in and missing full stop at the end.

the command cucumber —i18n

I added the empty Account class in section 6.1 as described and ran cucumber. But instead of the first step succeeding, it fails as follows:

$ cucumber
Feature: Cash Withdrawal

Scenario: Successful withdrawal from an account in credit # features/cash_withdrawal.feature:2
Given I have deposited $100 in my account # features/step_definitions/steps.rb:4
wrong number of arguments (1 for 0) (ArgumentError)
./features/step_definitions/steps.rb:5:in `initialize’
./features/step_definitions/steps.rb:5:in `new’
./features/step_definitions/steps.rb:5:in `/^I have deposited \\$(\\d+) in my account$/’
features/cash_withdrawal.feature:3:in `Given I have deposited $100 in my account’
When I request $20 # features/step_definitions/steps.rb:8
Then $20 should be dispensed # features/step_definitions/steps.rb:12

Failing Scenarios:
cucumber features/cash_withdrawal.feature:2 # Scenario: Successful withdrawal from an account in credit

1 scenario (1 failed)
3 steps (1 failed, 2 skipped)
0m0.003s
—-

It fails in the same way if I use the steps.rb linked in the PDF at that point.

Changing the class definition to the following makes the step pass as expected:

class Account
def initialize(balance)
end
end
—-

For reference:

$ ruby —version
ruby 1.8.7 (2011-02-18 patchlevel 334) [i686-darwin10]

$ cucumber —version
1.0.2

eq(amount.to_i) should be eq(amount) in the step definition (two occurrences on PDF page 92, one on page 93).

From @andypols via a private email:

One minor thing: I have just been reading the Gerkin basics. I think it would be much better if you used the same domain for how not to do it and how to do it. Its confusing when you talk about not writing “If a customer enters a credit card bla bla…” and then you say the way to do it is about changing your username!

The see_messages.feature file has a small typo in the Scenario line:

Scenario: See another users’s messages
->
Scenario: See another user’s messages

I don’t think you need download links on listings of output. On this page, there’s first_taste/01/features/adding.out, but this is only one of many.

Section 3.4, third para, “Each feature typically has somewhere between five and a twenty scenarios,”: “a twenty” should be “twenty”.

Regexes in examples on pages 46-48 have no anchors. On page 56 you say “You’ve probably noticed that the step definitions we’ve been using in our examples always start with a ^ and end with a $.” From erratum #47310 it appears that your policy used to be to omit anchors until they had been explained, but I guess this has changed and pages 46-48 still need correcting.

Adding the ‘launchy’ gem to the Gemfile is talked about and then the code for the Gemfile is shown without the ‘launchy’ gem.

In the box titled “Test Automation is Software Development”, third para:
“Testing and software development are complimentary skills” - “complimentary” should be “complementary”.

The launchy gem isn’t included in the listing at the top of page 123.

“Examining Aruba’s Step Defintions” - missing i in “Definitions”

At the end of the page, you wrote:
“”"
Change the feature to say “Then it should fail with:” instead of “Then it should pass with:”.
“”"
but the example output was made out of the original “Then it should pass with:” so failed to show us how “Aruba failed our step” (following page.)

the text

“my_account.balance.should eq(amount.to_i) ”

should be

“my_account.balance.should eq(amount)”

In other words remove the to_i, because the Transform code is applying the to_i elsewhere.

in three locations on this page.

(BTW loving the book so far :D )

About half way down the page the sentence “… with my_account stored as an instance variable:" is a bit confusing; dropping the possibly makes it clearer: ”… with my_account stored as an instance variable:".

“beskrvive” should be “beskrive” in the phrase “No matter if your users speak Norwegian or Spanish, med Gherkin kan du beskrvive funksjonalitet i et språk de vil forsta”

In the middle of the page there is a feature refactored to use background. The background is given a description, however this description begins with the keyword “When”, which is interpreted as a step.

Running this feature gives the following output (irrelevant bits deleted):

You can implement step definitions for undefined steps with these snippets:

When /^the bank issues new cards to customers, they are supplied with a Personal Identification Number \\(PIN\\) that is randomly generated by the system\\.$/ do
pending # express the regexp above with the code you wish you had
end

More of an inconsistency than an error, really. When defining the Account class directly in the step definition, the first step definition also changes.

In the first version, the step definition is as follows:
Account.new :balance => amount.to_i

In the second, it reads:
Account.new(amount.to_i)

Quote: “As our test suite grows, we can imagine these calls to to_i() littering our step definitions. Even these three characters count as duplication, so let’s stamp them out.”

If underscore is a character, then to_i is actually four characters, not three :)

Probably a question of personal preference: I can’t recall ever seeing to_i called with parentheses in the wild (the same goes for any ruby method that doesn’t take arguments).

Thus “We still have one issue left on our list which is the duplication of the calls to to_i().” reads a bit funny as the parentheses seem a bit out of place.

The last sentence on page 92 (continues on page 93) reads:
“You use it to prepare the environment ready for the rest of your support and step definition code to operate.”

I think you either “prepare for” something or “get ready for” something, not “prepare something ready for”.

Book says “Let’s try to provoke a failure. Change the feature to say Then it should fail with:
instead of Then it should pass with:.” but example cucumber output still shows “Then it should pass with:”

“…Aruba when we wrote those this scenario,…” should read “…Aruba when we wrote those scenario,…”

First Aruba regex example:
“to the STDIN, STOOUT and STDERR” should read “to the STDIN, STDOUT and STDERR”

“for ever,” should read “forever,”

Add
require ‘rubygems’
before
require ‘sinatra’
in nice_bank.rb

without which the web server will not run

As an alternative to my previous comment for this page you could change the text:
ruby lib/nice_bank.rb
to
ruby -rubygems lib/nice_bank.rb

About a third of the way down the page, “and the call to World()” is italicized; possibly only “World()” should be in italics.

In the second paragraph, change “There” to “there” in “… any use yet: There’s no external …”.

Near the bottom of the page, I had to read “We got around the whiteboard with our user experience team, and …” several times before it clicked. How about: “We gathered around the whiteboard with our user experience team and …”?

Near the middle of the page, “For the CashSlot we’ll change our World extension module to read the web application’s setting instead of creating iterations own:” should to be changed to something like “For the CashSlot we’ll change our World extension module to use the web application’s setting instead of creating its own instance:”.

The code silently changes to: @my_account.balance.should eq(amount), “Expected the balance to be #{amount}”
from: my_account.balance.should eq(amount), “Expected the balance to be #{amount} but it was #{my_account.balance}” on page 86pdf, 83book.

I’ve found this a couple of times while reading/playing along. I personally don’t copy and paste the code. So, it is confusing when other parts of the code disappear or change without being mentioned.

The sentence “We’ve moved the code that initializes the Account out of the first step definition into the new method.” makes sense only if you download the code. It is not obvious from reading what was really changed. Not only do you remove the initialization code, but you also revert the instance variables.

Instead of saying “… but we don’t really like leaving instance variables in step definitions like this.” on the previous page, maybe you need to say something to the effect of there is a better way; revert to the version before the instance variables.

Context: As developers, this helps
us stay honest and avoid the tempation to gold-plate our code.

The “gold-plate” reference is somewhat confusing and doesn’t clearly represent what the author may be trying to convey.

The code on the bottom of the page is somewhat cut off in the middle of a block. This makes it somewhat difficult to read with good comprehension. Can there be a page break before the block?

Context: One of the things we love about working outside-in with Cucumber is how it helps us to stay focussed.

On this page you use the UK spelling of the word ‘focused’, but elsewhere you use ‘American’ spellings, like when you use the word ‘color’. Go go gadget consistency. :D

Context: You can even include empty lines, like this one:
In the feature file, this is broken up over two pages just after this line, rendering the example a bit confusing. As a suggestion, either fix the example to be on the same page by moving the context line down on the next, or include line numbers in the code examples. Mmm, that sounds like such a nice feature.

The second User.create! example in the “Joe asks” section uses single quotes around the first four string literals, but double quotes around the last two (“Test” and “User”).

First bullet point in section 8.7: “… in or Rails project …” should be “… in our Rails project …”.

Only suggesting that, in American English, “disoriented” sounds more natural than “diorientated”.

Second line on the page.

Section 4.1 Figure 3 withdrawal is misspelled in Feature string

Footnote contains the typo “youl’ll”.

Got interrupted. Several typos in the footnote.

“youl’ll” should be “you’ll”.
“dependepency” should be “dependency”.
“make a brief encounter” might be better said as “have a brief encounter” or “encounter it briefly”.

“only on User” should be “only one User”, last full paragraph on the page.

“Still, people keep using it to test web application.” should use the plural, “applications”, after the italicized, dirty little secret.

Seems that the line

is missing from the “Then the results should be:” table.

“wes_steps.rb” should be “web_steps.rb” in the “Joe asks:” callout.

The first paragraph ends with

“…we have set ourselves up for failure. The good kind of
failure that tells us what do do next:”

“do do” is a funny typo; should be “to do”.

The last sentence works informally, but it’s not a complete sentence. Guess that’s between you and your editor.

As the chapter develops, it becomes clear that the difference in table lengths between Given and Then is intentional, and used leading up to the discussion of table diffs on p147. I’d suggest that you explain the motivation when the different sized tables are first presented.

I’m reminded of Michael Hartl’s Ruby on Rails tutorial where he tells us that he’s intentionally forgotten to create a view using a generator so that he can show how to add it by hand later on in the chapter.

The attentive reader feels good about noticing that something’s not quite right, and doesn’t fret about it because they know it will be addressed later.

I had to add “require ‘factory_girl’” to factories.rb in order to make this work (legacy app on ruby 1.8.7 & rails 2.3.11).

Elsewhere I’ve mentioned that, ultimately, I need to get cucumber working with a legacy app (ruby 1.8.7/rails 2.3.11). This chapter, which is most important for me, is proving to be a challenge in that environment, so I set up ruby 1.9.2/rails 3.1.0 in order to continue working through the examples.

My scenario does not pass after adding “find(‘ol.results li’)” to search_steps.rb. I have both increased Capybara’s default_wait_time and tried using sleep(), to no effect.

Unable to find css “ol.results li” (Capybara::ElementNotFound)
(eval):2:in `find’

AJAX works as expected when using a browser manually as described at the top of this page.

At the top in the “Joe asks:” box.

“Why are wes_steps.rb” should be “Why are web_steps.rb”

Second sentence, purple callout: “loses it’s abililty” should be “loses its abililty”.

Second paragraph of the main text: “you do want to automated them” should be “you do want to automate them”.

Fourth paragraph: “second scenario on it’s own?” should be “second scenario on its own?”.

Fifth paragraph: “The opposite of this, stateless scenarios each ensure” should probably have a comma between scenarios and each.

Second paragraph under “Shared Environments”: “team, but the tests scripts too.” should be “team, but the test scripts too.”

The second paragraph of “Fixtures” has the sentence: “In a legacy system, especially where the design has evolved organically, where creating the single object actually needed for the test you’re working on means you need to create a huge tree of other dependent objects, you’ll feel like the easiest option…”. The middle bit with the “wheres” is nice, and for readability, I’d suggest separating it from the rest of the sentence with emdashes and adding “one”: “In a legacy system—especially one where the design has evolved organically, where creating the single object actually needed for the test you’re working on means you need to create a huge tree of other dependent objects—you’ll feel like the easiest option…”.

Second paragraph, purple callout: “stored procedudes, views, functions etc.”, much as it made me laugh, should be “stored procedures, views, functions, etc.”. Note, too, the comma between functions and etc.

Top of the page: “other disadvantages that just waiting” should be “other disadvantages than just waiting”.

The “Nightly Build” callout:

First sentence: “have a Slow Features, on page 75 caused” is better without the “a”.

Paragraph before “What We Just Learned”: the sentence “When there’s a problem with the tests, whether that’s an urgent problem like a failing test, or a nagging annoyance like a flickering scenario, put your best people on it, and fix it forever.” might also benefit from emdashes:

When there’s a problem with the tests—whether that’s an urgent problem like a failing test or a nagging annoyance like a flickering scenario—put your best people on it, and fix it forever.

I think you can lose the comma before “or” that way, too.

First full paragraph: “at it’s heart it’s a” should have “at its heart it’s a”.

Purple callout, second paragraph: “individually in you step definition code” should be “individually in your step definition code”. “or have a big fat set” would fit the opening of the sentence better as “or having a big fat set”.

Third paragraph: “our domain model, (see the FactoryGirl documentation for details on how this is done) all”. The parenthetical remark should come before the comma; “model (…), all”.

“stakeholders loose interest” should be “stakholder lose interest” (bottom of p. 76 of the PDF in “Bored Stakeholders”).

“they soon become nothing more than a testing tool” - the referent for “they” is unclear. I’m pretty sure you mean the features, but it could easily mean the stakeholders (as it does a little earlier in the same sentence).

Sentence reads: Have you ever come to a web site and though:

should be “thought”

Figure 10. Don’t capitalize “Of” in “Alternative 2: Out-of-Process”

Since this chapter is structured to contrast the two different approaches to REST API testing, and since section 14.1 is called “In-Process testing of Rack-based REST APIs”, section 14.2 should probably include “Out-of-Process testing” in its heading.

I know you are planning to put installation steps in the Appendices, but, for the benefit of newbies, you should be explicit that sinatra is a gem, and that they need to add it to their Gemfile and to run “bundle install” before attempting to “require” it.

If readers continue to “bootstrap rails” for new projects as they learned in Chapter 11, they will already have a “features/support/env.rb” and the $LOAD_PATH instructions, mid-page, should be about adding to that file rather than creating it.

I doubt if it is worth a purple callout box, but you might consider adding a footnote about the façade pattern when it appears in the middle of the last paragraph on the page.

Your editor would know if it should be “a HTTP server” or “an HTTP server”, but since I read it as “HTTP” and not “hypertext transport protocol”, “an” sounds more correct to my ears.

Oh, and as a complement to #47688, the reader most likely already has the rack-test gem, so they won’t have to add it to their Gemfile and run “bundle install”! (Last paragraph.)

In the “Joe asks:” purple callout box:

In the sentence, “For a graphical user interface, such as a HTML web application we should…”, I would definitely insert a comma between “application” and “we”.

I would also change “a HTML web application” to “an HTML web application”; same reasoning as #47690, which is that “H”, though a consonant, has a vowel sound when read in the context of an acronym. Your editor should be the final judge, though, there must be a style guide for this.

First paragraph of the text:

I would change the link from “Chapter 8, Support Code, on page 121” to “Section 8.4, Using Hooks, on page 135”.

I would also make a more explicit link between the sentence on “using a proper database” and the “Database Cleaner” purple callout on the next page. If they end up on facing pages in the paper version, that might not be necessary, but otherwise, why not make the connection slightly clearer?

“Put the following file in the root directory…”

I created a new project and rvm gemset for this chapter, and ran the cucumber:install generator and followed along in 14.1. I don’t know at what point my config.ru got created, but it’s already there at this point:

1. This file is used by Rack-based servers to start the application.

require ::File.expand_path(‘../config/environment’, FILE)
run RESTwebService::Application

Bottom of page: “you should get a JSON response containing {}”

What I see is a sideways tumbler spilling ice and my drink, plus:

NoMethodError at /fruits
undefined method `to_json’ for nil:NilClass
file: fruit_app.rb location: block in line: 10

and a backtrace, environment settings, and a message that I’ve “enabled the show_exceptions setting”. From your text, I was expecting to literally see “{}” in an unstyled browser window.

“… deleting all the code in the features/support/env.rb.”

I would either add the word “file” to the end of that sentence or delete the word “the” before the filepath.

And, related to #47689, if the reader used the generator, they will have started this chapter with an env.rb that contained “require ‘cucumber/rails’” and settings for Capybara.default_selector, ActionController::Base.allow_rescue, and DatabaseCleaner.

It should be clear that these should be deleted in addition to what was added in 14.1.

The hooks.rb file has a “require ‘httparty’”, so the gem needs to be installed before the Before/After code will work. You do tell readers to install it on page 224, but you need to do this earlier.

Perhaps you should tell people to stop the server, and how to stop the server, before introducing the hooks file.

It is a good example to show how to start and stop the server from within cucumber.

The diagram should really be above or below the code snippets, not in the middle of them. It kinda hinders.

in first_taste/04/features/step_definitions/calculator_steps.rb where you `require ‘open3’ perhaps it would be better to subliminally hint at lazy loading by putting the `require ’open3` inside of `When /^the calculator is run$/ do

The text refers to “We initialize the MessageQueue class” but the class illustrated in the example is the TransactionQueue class.

“To satisfy youself that our new method…” should be “To satisfy yourself that our new method…”

In the first sentence, “simpistic” is not a word.

In the second paragraph, “font-end” is written with a hyphen and “back end” is two words. These two ‘words’ should be consistent in spelling.

his managers thought we was crazy

should be

his managers thought he was crazy

< put come concentrated
—-
> put some concentrated

“we described the over-all flow” - overall shouldn’t be hyphenated

“you will se that the search field is populated with the search” - se -> see

First sentence of the third paragraph of the “Leaky Scenarios” section : “anther earlier scenario” should read “another earlier scenario”

(except if The Cucumber Book has changed into a botanic book :-) — wiktionary: “anther: (n.) The pollen-bearing part of the stamen of a flower.”)

tempation -> temptation
@last line

In “Synchronizing with Delayed Job”, the first paragraphe after the code excerpt: “you’ll make sure your they will wait here” either there are missing words after the “your”, or it can be removed.

“Without Javascript TDD”, third sentence, “Sice Javascript TDD is a big topic”: sice is not a word (well, it is, but not in English, in Czech :-)

After using the Transform to convert the Regex into an integer, the “to_i” call has been left in the step definition.

Given /^I have deposited \\$(#{CAPTURE_A_NUMBER}) in my account$/ do |amount|
my_account = Account.new
my_account.deposit(amount)
my_account.balance.should eq(amount.to_i),
“Expected the balance to be #{amount} but it was #{my_account.balance}”
end

change “show” to “shows” immediately after the link to Figure 9.

Change “them” to “then” at the top of the page in “… whether the amount is a credit or a debit, them writing it to the queue.”

At the bottom of the page in the list of files created by “rails g cucumber:install” is listed “features/support/boot.rb” by which I believe is meant “features/support/env.rb.”

nutshel should be nutshell

How about updating the factory_girl sections to correspond to the current preferred syntax and discussing, even briefly, factory_girl_rails?

FactoryGirl.define do
factory :user do |f|
f.username ‘testuser’
end
end

as opposed to Factory.define

“anther” should be replaced by “another”.

3rd paragraph

When one scenario depends upon state left behind by anther earlier scenario in order for it to pass…

“cruicial” should be “crucial”

“…here to avoid losing users at this cruicial stage of the…”

and

“exiry” should be “expiry”

And I should see a message telling me the exiry data must be wrong

Stay consistent with the use of “(s)”

You use
the context
the event (when there could be more than one)
and
the outcome(s)

“The pending scenarios become our TODO list for the work we’ll do we do when we drop”

In this part

“we’ll do we do when we”

Remove the “we do”

“tell a story, because you reader can keep track of stories much better ”

should be “your reader”

“which is the amounts of money involved”

amount of money

“The ubiquitous language your team uses will driven by the domain”

should be “will be driven”

“It’s a good job we caught this now”

should be “it’s a good thing”

“the team and end up with messy test code”

I think this should be “the team can end up with messy code”

In the Synchronizing with Delayed Job section:

“availlable” should be “available”.

“…but it can be useful to have this availlable for debugging.”

In the specific version of cucumber-rails in our gemfile, the web_steps.rb file has been removed (as you know), and this is a great thing, but it is still mentioned in the book. A user not familiar with previous versions of cucumber-rails may get confused when trying to look for features/step_definitions/web_steps.rb.

“You may wonder why we’re not writing the scenario in such way that we can take advantage of the step definitions in features/step_definitions/web_steps.rb.”

Thoughts?

Not sure if this counts as a typo - reading the epub version, the logo in the “Matt says” section is really large - at least in iBooks on the iPad. Especially in the two-page landscape view, it pushes the text in the column down to just a couple of words.

It’s also not appearing in the header near “Matt says” - it’s on the next page. This is probably because it’s too large.

“It print a message to the console”

should be

“It print*s* a message to the console”

Let’s say your team has agreed to always keep the WIP at or below 3 scenarios. To enforce this, you could run Cucumber with the following option:

$ cucumber —wip @wip:3

This has some interesting side effects. First of all, if Cucumber finds more than 5 scenarios tagged with @wip it will fail immediately without even running any scenarios.

>> I think you mean to say “finds more than 3 scenarios…”

" the best place is look at Aruba’s own Cucumber features"

should be “the best place to look is Aruba’s own…”

fron doing this,
should be:
from doing this,

Guven
should be: Given

s/cruicial/crucial/ in feature description

Never mind, the URL in the book is now working for me.

Last sentence on the page: “it’s” should be “its”.

Last paragraph on the page: “snippet is different to normal for the two step definitions”.

I suspect this may be another case of American usage differing from English usage, but this phrase reads strangely to me. On first pass, “different to normal” sets me up to expect a range, as in “fair to middling”. When I try it a second time, I think it should read “different from normal”. Maybe the solution is to rewrite the sentence entirely.

“Notice the different format for the two snippets where we’re expecting to receive a table.”

“You can read more about Sinatra at http//sinatra.rb.com”

Should be

“You can read more about Sinatra at http//sinatrarb.com”

In the next to last paragraph you have: Tell the story from the beginning, using Guven steps to set up all the state you need for that particular scenario.

Guven should be Given.

Often we’ll want to work with it in that raw form, so we can call the row method on it to do just that. - I believe “row” should be “raw”.

Stateless section.
Last paragraph, 2nd line.
“….using Guven steps….”

The last paragraph in the stateless section has “Guven” I think it meant “Given”. At the top of the page there is a sentence that goes “There is a an additional keyword…” the word “a” is probably a mistake

Middle of the page :

When you’re working with a particular language, you can discover the key-
words by passing the language code (as listed by the command above) to the
—i18 switch. Japanese, for example:

It says —i18 switch, which should be —i18n as stated in the previous and following example.

PS: so far, very good book. I’m glad I bought it!

— “run cucumber again you should that the step has failed”
“run cucumber again you should see that the step has failed”

—" Sue has a different password to Dave"
" Sue has a different password than Dave"

—“with an maintain.”
“with and maintain.”

In the second sentence of the fourth paragraph, “…points out…” should be “…point out….”

In section 1.4, “How Cucumber Works,” the third sentence of the fifth paragraph says “…scenario has having…” when it should say “…scenario as having….”

In the last sentence of the second bullet item in the “Take Care of Naming Scenarios” sidebar, “…you won’t need read the code…” should be “…you won’t need to read the code….”

module MyCusomHelperModule

should be:

module MyCustomHelperModule

context - code/scenario

this:
And I should see a message telling me the exiry data must be wrong
should be replaced with:
And I should see a message telling me the expiry data must be wrong

(Do I Need To Know Ruby?)
Ruby is an … can be installed AND run on all major …

(1.3 Living Documentation)
Cucumber tests share … traditional SPECIFICATION documents …

(5.4 Nesting Steps - in the scenario code shown)
Scenario: SUCCESSFUL login with PIN

(First paragraph.)
… fails to help your team to COMMUNICATE. We’ll start by …

(6.6 Try This - Defection Prevention on Your Team)
Think of three things that are slowing down YOUR team’s production line.

(6.6 Try This - The last paragraph on the page.)
spot the crucial Then step that was missing from the ORIGINAL scenario?

(Second paragraph.)
Now that we’re persisting state TO disk in our TransactionQueue and BalanceStore, …

(Investigating the Flickering)
Run cucumber and give it a try: <— *’d prose incorrectly formatted fixed width (as code).

(paragraph following the Account class code)
… We are also telling ActiveRecord to run migrations in the db/MIGRATE/ directory. …

(Consistency with downloadable file names and the account.rb code, which use “db/migrate”.)

When removing the BalanceStore class: ‘features/support/hooks.rb’ also contains a reference to BalanceStore which must be removed.

(in “Matt says:”)
This gives you the added FLEXIBILITY of being able to actually …

(15.4 Taking Screenshots - second last paragraph)
… We’re calling the save_screenshot METHOD from Selenium to save …
OR
… We’re calling save_screenshot from Selenium to save …

(15.4 Taking Screenshots - second last paragraph - note the apostrophe “t” -> “s”)
… we’re using the scenario object’S id as an identifier …

Scenario flickering didn’t appear with only 10 cucumber runs, I needed many more. Try (non-Windows):

“while /usr/bin/true ; do cucumber -f progress feature/cash_withdrawal.feature ; done”

Scenario flickering didn’t appear with only 10 cucumber runs, I needed many more. Try (non-Windows):

“while /usr/bin/true ; do cucumber -f progress feature/cash_withdrawal.feature ; done”

It’s possibly to tweak the comparison behavior

should be:

It’s possible to tweak the comparison behavior

It says “This is just a taster of what you can do with data tables in Cucumber.” Should say “taste”?

When reading the ePub on my iPad, the “X pages left in chapter” on the bottom right of the page, the page count seems to indicate sections rather than chapters.
For instance, page 136, start of section 5.1, says ‘4 pages left in chapter’ whereas there are far more than 4 pages to go in chapter 5.

After the contents of Gemfile, but before “cd squeaker”.

Need to run ‘bundle’ after editing Gemfile.
Need to run it for the test group.

Since you’re discussing what the cucumber installer does, it would be useful if you could also explain the line cucumber 1.2.1 adds to the database.yml file when it installs:

cucumber:
<<: *test

I assume it’s doing something like appending a cucumber environment to the test environment, but it would be nice to have this confirmed.

name=“criteria”
should be
name=“query”

unnecessary page break before list of platforms

(2nd paragraph) instead of “a reading computer program” you want “reading a computer program”.

The test.feature file should not live in the root dir, but in the features dir. In case the reader just adds on top of what was done in the previous chapter, this will cause cucumber to load all ruby files in the root directory.

Have not got access to page number as reading book on Kindle. However in the Joe Asks: Should I Show Synchronization Points in My Features section, it states ’we’ve used the eventually method to introduce an implicit sync. I cannot find a reference to the eventually method in the preceding section of the chapter.

Scenario: Expiry date invalid
When I enter a card expiry date that’s in the past
And all the other details are correct
And I submit the form
Then the form should be redisplayed
And I should see a message telling me the exiry data must
be wrong

in the last line, instead of “exiry”, it should be “expiry”

The /^I run “(.*)”$/ step definition is deprecated. Please use the `backticks` version.

“Out first two steps” should be “Our first two steps”.

“Spoken languages” should be “Spoken Languages”

“I found myself asking.)” should be “I found myself asking)” (no period)

“bravely refactoring in code that previously” can just be “bravely refactoring code that previously” (delete “in”)

“ActiveRecord was born in the Ruby on Rails framework”. This should make it clear that it was the gem that was born of Rails. The design pattern is older.

If you’re going to use require_relative, you might as well use the new hashrocket-less syntax.

I imagine books will migrate to the Ruby 1.9 syntax over the course of time. It reminds me of the switch to algebraic notation in chess back in the ’70s.

Bundler instructions assume too much prior knowledge of the reader: could be more user friendly.

See forum topic 10355 (The form won’t let me add a link)

“This file must be placed either in your project’s root directory or underneath a directory called config.” Supposed to be “underneath in a directory”?

Unused variable in code: try_again.

The list of files at the top of the page could contain Gemfile and Gemfile.lock.

“These [gems] are installed by adding them to your Gemfile and running bundle update”. Bundle update will work for the example Gemfile, because it specifies all versions. But for the usual project, which should not specify versions (that’s what Gemfile.lock is for), you don’t want to update the whole bundle when adding a gem.

“Feathers’ [recipe]” should be “Feathers’s [recipe]”. There is only one Feathers here.

“we never ran it yet” should be “we haven’t run it yet”

“We’re only entering text … we are not submitting the form by clicking the button or hitting the Enter/Return key”

Instead of “entering”, just say “typing”.

“Capybara’s rack mode”

On page 144, there is an error in the code:
class UserInterface
def withdraw_from…

withdraw_from should be withdrawal_from to match the method name in the Teller class.

Oops. Page 144 doesn’t have an error. Disregard.

I had to change my selector as appearing in the 05 code fro movie_steps/rb to get pass from
When /^I create a movie Caddyshack in the Comedy genre$/

The change I made was to call the HTML id as opposed to the natural text for the release year control.

#select “1980”, :from => “Release Year”
select “1980”, :from => “movie_release_year”

I was running a 1.92 ruby and the 05 gemfile. 1.92 is listed as acceptible in the front of the book and 1.87 was causing problems. This is an RVM environment. I may still be having gem issues. I have to run rails generate cucumber:install in your numbered directories to avoid a problem with the cucumber problem.

“Refactor the feature to make use of the new keyword and compare the new version of the feature with the old one.”
The “new keyword” is not some keyword called “new”, but a new keyword (either Background or Scenario Outline, or possibly Examples) and thus shouldn’t be in a monospaced font.

“It didn’t get chance to update the balance before Cucumber checked it and failed the scenario.”
Should be “didn’t get a chance”.

“This is why we ’ve removed the :balance field that we had before we refactored the class.”
Extra space in “we ’ve”.

“The advantage of truncation that it’s a cleaning strategy that works reliably when we have more than one database connection” is missing a verb.

In “As the number of feature and scenarios grows”, “feature” should be “features”.

“CI systems detect failure by inspecting the exit status of the processes it’s running”: There’s a singular/plural issue (“CI systems” does not match “it’s”).

“Cucumber exits only with an error status (a nonzero value) if there is one or more failing steps”: should be “there are one or more”.

In “They each have their strength and weaknesses”, “strength” should be “strengths”.

Book says:
To get the last step definition working, change the last step definition in
features/step_definitions/calculator_steps.rb to look like this:
Download first_taste/07/features/step_definitions/calculator_steps.rb
Then /^the output should be “([^”]*)"$/ do |expected_output|
@output.should == expected_output
end

However, this does not run. By removing .should from the line “@output.should == expected_output”, the feature passes.

Transform /^(£|\\$|€)(\\d+)$/ do | currency_symbol, digits |
Currency::Money.new(digits, currency_symbol)
end

This piece of code gives an error: “invalid multibyte char (US-ASCII) (SyntaxError)” because the currency symbols (£ and €). To fix this the only way I founded it’s adding “# encoding: utf-8” without the quotes at the very beginning of the file.
Also the transform block should be assigned to CAPTURE_CASH_AMOUNT because it’s used in the following pages.
And because Currency module wasn’t never included in the book, we get an error there too: “uninitialized constant Currency (NameError)”

So, the code should look like this:

1. encoding: utf-8
   CAPTURE_CASH_AMOUNT = Transform /^(£|\\$|€)(\\d+)$/ do | currency_symbol, digits |
   digits.to_i
   end

broken linkto github for Cucumber::Ast::Scenario at the bottom of the page (footnote [4])

“they’re a reading”

“with an features”

According to google translate the word regnskapsførerere does not have an english counterpart. When I remove the last “re” it appears to be accountants. When I remove the last “ere” it appears to be accountant. I believe this is a typo on the norwegian word in the example for other languages.

Book says:

so all we need to add is a field with a name=“criteria” attribute

“criteria” should be “query” ?

Page 253-254 has a code snippet split across a page break. With your formatting tools, can you give some weight to not breaking code snippets across pages?

The “Running Capybara Against Non-Ruby Applications” box…

Does this imply that you could point app_host to your staging or even your production server and drive it to verify they are working correctly?

If that is true, I think I’d mention it. That sounds really powerful to me.

Just after adding the necessary gems in the Gemfile there is the command to generate the cucumber files and folders with cucumber:install:

$ cd squeaker
$ rails generate cucumber:install

Even though it is fairly obvious for experienced rails users, shouldn’t we also add another line before running the cucumber installation script to install the gems as in:

$ bundle install

as if it is not run it would produce the error:

Could not find gem ‘database_cleaner (~> 0.9.1) ruby’ in the gems available on this machine.
Run `bundle install` to install missing gems.

Should the javascript runtime environment gems should also be added to the Gemfile?
If not there is the following error message:

~/.rvm/gems/ruby-1.9.3-p327@rails3_2_9/gems/execjs-1.4.0/lib/execjs/runtimes.rb:51:in `autodetect’: Could not find a JavaScript runtime. See …execjs_url… for a list of available runtimes. (ExecJS::RuntimeUnavailable)

So by adding the following should fix this:

gem “therubyracer”, “~> 0.10.2”

Cucumber scenario always fails when using RackTest driver and the following tag: Withdraw
Possible workaround:
use Selenium driver instead
Fix:
change tag to:

Quick note before description - I wasnt able to select the correct version of the book, the one i am reading is P2.0 August 2012.

Just finishing up for the day on Chapter 8 hooks and then just starting Chap 8.5. The example code support_code/14/features/support/debugging.rb is where i hit a snag.

I had written the code into the new file and it didnt work telling me i needed a version of launchy. Now being at the end of the day and not coming across this specific error message i hadnt quite twigged what could be causing this and after double checking what i had written against your examples i was a bit stumped.

However remember a previous example of where sinatra was used i finally though aha

require ‘launchy’

would be needed, it then worked :3 (Being a pure manual test analyst for a while i havent coded for about 5 years so i am quite gleeful when i figure something out no matter how small!)

However it feels that this bit of code has either been missed off the example or assumed that the reader would automatically put the required code in and could cause confusions like mine.

This is the chapter on testing javascript enabled web pages. The examples all print with the error:

Can’t mass-assign protected attributes: user
(ActiveModel::MassAssignmentSecurity::Error)
./features/step_definitions/user_steps.rb:16

but this is never addressed in the body of the text. I suspect that an “attr_accessible :user” was dropped or missed from the Messages model class.

As of Capybara 2.0, a call to find gives an ambiguous match error when more than one element is found in the page. Because of that, using find here would give an error and fail the scenario. Instead I used sampling with the anticipate library as a workaround. Here’s how my step definition looked like:

include Anticipate

Then /^the results should be:$/ do |expected_results|

trying_every(0.1).seconds.failing_after(200).tries do

page.all(‘ol.results li’).should_not eq(nil)

end

results = [[‘content’]] + page.all(‘ol.results li’).map do |li|

[li.text]

end

expected_results.diff!(results)

end

code snippet capybara/00/features/step_definitions/user_steps.rb at page 248 is Wrong:

Error line:
message_attrs.merge({:user => user})

Correct line:
message_attrs.merge({:user_id => user})

This typo it’s causing subsequents wrong cucumber outputs on lines 249, 251, 252, 254 with the followind content:

Can’t mass-assign protected attributes: user
(ActiveModel::MassAssignmentSecurity::Error)
./features/step_definitions/user_steps.rb:16
features/search.feature:3

Please check it out!.

Capybara 2.0 results in an ambiguous match error when find(‘ol.results li’) is called (as reported in erratum 51206. An easier fix than the one suggested in 51206 is to make the call
find(‘ol.results li:first-child’).

The test examples in Chapter 15, as printed, keep complaining of a Mass Assignment Error. Therefore, we never actually see it working :)

“When the team write their acceptance tests collaboratively”
should be either:

- “the team writes”
or
- “teams write”

There is a list that spans from page 231-2 and “config/cucumber.yml” is the first and last thing in the list, but with different descriptions.

FactoryGirl is a gem and has no gender. Please don’t personify it in your writing. “poor old FactoryGirl complains that she doesn’t know…”

Tocino grueso! (Chunky Bacon!) should be Tocino grueso! (Thick Bacon!). Thick bacon is the proper spanish/english translation.

Rick Flores

The book suggest to write this

Then /^The output should be “([^”]*)"$/ do |expected_output|
@output.should == expected_output
#pending # express the regexp above with the code you wish you had
End

but as the should is a undefined method it doesnt work hence the code must be

Then(/^The output should be “([^”]*)"$/) do |expected_output|
@output == expected_output
#pending # express the regexp above with the code you wish you had
end

With newer FactoryGirl :
Download rails/02/features/step_definitions/user_steps.rb
Given /^there is a User$/ do
FactoryGirl.create(:user)<——-Original“Factory(:user)”
end

Footnote 3

h..p://behaviour-driven.org/

displays :

502 Bad Gateway

Error 404 for footnote 2 h..p://cukes.info/cucumber/api/ruby/latest/Cucumber/Ast/Table.html

Error 404 for footnote 1 h..p://cukes.info/cucumber/api/ruby/latest/Cucumber/RbSupport/RbWorld.html


Missing code when adding instance variable.

Error: @my_account.balance.should eq(amount),“Expected the balance to be #{amount}”

Should be: my_account.balance.should eq(amount), \t"Expected the balance to be #{amount}, but it was #{my_account.balance}"

You should point out that “show-guests” needs to come =before= “home”, otherwise “home” will not be able to see “show-guests”!

Last paragraph “When the team write their acceptance tests collaboratively” replace “write” with “writes” (or “team” with “teams”).

When running cucumber, getting:
DEPRECATION: Using `should` from rspec-expectations’ old `:should` syntax without explicitly enabling the syntax is deprecated. Use the new `:expect` syntax or explicitly enable `:should` instead.

Description of \\b in the “Useful Shorthand Character Classes” top box is wrong. It’s a zero-width match anchor. Sounds like you are describing \\W instead. Either change \\b to \\W, or improve the description so that people don’t think it matches characters.

Between the 2nd and 3rd paragraphs of section 11.5, there’s an example run of cucumber:

$ cucumber —wip @wip:3

That doesn’t work (in v1.3.15, at least): the —wip option doesn’t take arguments, and so cucumber tries to use @wip:3 as a features path. I think the following is correct:

$ cucumber —wip —tags @wip:3

name=“criteria”
should be
name=“query”

Link to Ruby Version Manager (RVM) web site in paragraph 2
is: beginrescueend.com
should be: rvm.io
Reason: Since sometime around April 2012 beginrescueend.com started to redirect to rvm.io. Currently the link redirects to elizamagazine.com.

Footnote 1
is: h..p://beginrescueend.com/gemsets/
should be: h..ps://rvm.io/gemsets/