Fix #4138 fix delivery address #4540

jimmyli97 · 2024-07-19T01:33:01Z

Resolves #4138

Description

Prints no address if delivery_method is pick_up
Prints partner address if partner has no program address and delivery_method is delivery or shipped
Prints partner program address if partner has program address and delivery_method is delivery or shipped
Tests output against expected PDFs
Adds helper test to generate expected PDFs

Type of change

Bug fix (non-breaking change which fixes an issue)

How Has This Been Tested?

Passes test suite
Manual testing

* group examples together with variables * rubocop -A passes on this file without needing to disable ArrayAlignment

* Compare generated against expected PDFs * Add helper test to regenerate expected PDFs * Add helper module

…ethod * Address output prints delivery address if filled in, otherwise partner address * Only does this for delivery/shipped method

cielf · 2024-07-19T11:42:56Z

Hey @jimmyli97 As a matter of process, in the future please get yourself assigned to the issues when you are going to be working on them - or at least comment on them. This helps us avoid duplication of effort. Thanks.

jimmyli97 · 2024-07-19T12:35:57Z

@cielf it automatically unassigned me, should I keep commenting every 7 days?

cielf · 2024-07-19T15:03:20Z

Oops - sorry, my bad (I somehow didn't see that you had been assigned and weren't anymore). Not to worry.

cielf · 2024-07-19T15:05:28Z

But if it's going to lapse while you are still working on it, yeah - commenting makes sense -- otherwise someone else may pick it up. Once you've put in the pull request, it's not as big a deal.

cielf

Hrm. I tried blanking out the two addresses, and got a mess. This shouldn't happen in real life (if it's being delivered, there really should be a delivery address, but I could see it happening in a situation where the bank is very familiar with a new partner, and some staff drops by the partner with the goods on their way home or something ). I'd still like to see it fixed. (You could suppress the "Delivery address" label in this case too.)

* REFACTOR out profile name and email

@dorner

My concerns have been addressed. Over to @dorner for a technical look-see.

cielf · 2024-08-25T01:53:14Z

Hey @dorner -- it looks like the ball on this one is currently in your court?

dorner

Overall looks OK - I had some concerns with the skipped helper method.

dorner · 2024-09-02T18:23:31Z

.gitignore

@@ -49,6 +49,7 @@ dump.rdb
 .DS_Store
 .ruby-gemset
 *.pdf
+!spec/fixtures/files/*.pdf


Why are you adding this if you've added these files? Shouldn't they be tracked?

line 51 tells git to ignore all .pdf files so line 52 specifies an exception (by prefixing an exclamation mark) to track the pdf files in spec/fixtures/files

added a comment to clarify in 0d53723

dorner · 2024-09-02T18:25:14Z

spec/pdfs/distribution_pdf_spec.rb

+        compare_pdf(create_dist(:pick_up), expected_pickup_file)
+      end
+    end
+    # this test is a helper function to regenerate expected PDFs, only commit with it skipped


That's really confusing. If you want a helper method, why not define it as a method, and people can run it from the Rails console?

to generate the pdf files correctly, the helper method generating the file needs to have the same test environment setup as the actual test itself. This way you don't have to repeat code setting up the test environment. I did research but I could not figure out a better way to do it without duplicating code.

you can call this method from the terminal by calling the test's line number e.g. bundle exec rspec spec/pdfs/distribution_pdf_spec.rb:227.

I added a comment clarifying this in 0d53723

alternatively I could delete the method if you want, but I found it super helpful because I could just run that line whenever I made changes to the pdf generation and regenerate the test files. so if anyone in the future makes changes they could do the same thing.

There should definitely be a way to do this without duplicating code. Maybe it's the setup that needs to be extracted to a base method that's called from both places. And worse comes to worst, if you do need to duplicate a few lines of code, it's still a lot less messy than a test that's not actually a test.

I came up with a solution which checks for an environment variable and if it's set to true to overwrite the expected PDFs in 4516951 . I think it's a lot cleaner, let me know if that works.

* Use environment variable for regenerating comparison pdfs

cielf · 2024-09-28T17:04:58Z

@jimmyli97 FYI: We had an urgent fix that required all the senior contributors this week , so we didn't get to look at this again. Hopefully this week will go better.

dorner · 2024-09-30T00:52:25Z

@jimmyli97 this helps a bit but not by much. Essentially:

To generate the PDF, you need some set of environment to be set up.
To run the RSpec, you need the same environment to be set up.

That does not imply that in order to generate the PDF, you should run an RSpec. It means you should extract the behavior that does the environment setup to shared code, and run that setup code both in the RSpec and in a separate script that can be used to generate the PDF.

… comparison PDFs

jimmyli97 · 2024-10-17T22:15:39Z

@dorner I managed to figure out a way to separate out the setup code so you can call a Rails console method which calls that setup code, and also the Rspecs call that setup code. it's a bit messy though, let me know what you think

dorner · 2024-10-22T13:57:26Z

spec/pdfs/distribution_pdf_spec.rb

-  let(:item3) { create(:item, name: "Item 3", value_in_cents: 300) }
-  let(:item4) { create(:item, name: "Item 4", package_size: 25, value_in_cents: 400) }
+  before(:each) do
+    @organization, @storage_location, @item1, @item2, @item3, @item4 = create_organization_storage_items


There's no need to do this via instance variables. Here's how I'd do it to keep the RSpec looking the same:

let(:storage_creation) { create_organization_storage_items } let(:organization) { storage_creation.organization } let(:item1) { storage_creation.items[0] } # etc.

See below where you can change this method to make it easier to work with.

addressed in 6da7702

dorner · 2024-10-22T13:59:07Z

spec/support/distribution_pdf_helper.rb

+    item3 = FactoryBot.create(:item, name: "Item 3", value_in_cents: 300)
+    item4 = FactoryBot.create(:item, name: "Item 4", package_size: 25, value_in_cents: 400)
+
+    [org, storage_location, item1, item2, item3, item4]


I'd define a struct above, something like

StorageCreation = Data.define(:organization, :storage_location, :items)

and here you could return

StorageCreation.new(org, storage_location, [item1, item2, item3, item4])

addressed in 6da7702

dorner · 2024-10-22T13:59:45Z

spec/support/distribution_pdf_helper.rb

-  private def create_profile(program_address1, program_address2, program_city, program_state, program_zip)
-    create(:partner_profile,
+  def create_organization_storage_items
+    org = FactoryBot.create(:organization,


If this setup is being used outside RSpec, it shouldn't be in spec/support and it shouldn't use FactoryBot. It should work more like the seeds file where you create data manually. Probably should live in its own folder under the /lib directory.

addressed in ea7e241

dorner · 2024-10-22T14:00:27Z

spec/support/distribution_pdf_helper.rb

+      organization: organization)
+  end
+
+  def create_file_paths


Are you actually creating file paths here? Or just returning the ones you have specified?

renamed to get_file_paths in 6da7702

…s fire

jimmyli97 · 2024-11-28T03:35:05Z

failing test is a known flaky one

cielf

If the partner should fill in an incomplete delivery address (maybe they just fill in the street address, because that's the part that is different), we get a mess like this:

I checked -- and this has happened 'in the wild'.

Please allow space for the city, state, and zip, even if not filled in.

…ete item units correctly when generating test files

jimmyli97 · 2024-12-03T03:09:00Z

If the partner should fill in an incomplete delivery address (maybe they just fill in the street address, because that's the part that is different), we get a mess like this

fixed in 39e8cb4

jimmyli97 · 2024-12-03T03:24:00Z

(6,1) failing test is known flake
(6, 4) failing test is present on main branch also, unrelated to this PR, think it's flaky and added to list

cielf

We're very, very close!

If there is no partner primary contact info, but there is a delivery address, you get something like this:

I checked production to see if this happens in the wild -- there are about 400 partners with this situation -- probably because the primary contact info wasn't there from day 1.

… blank * Add rspec * Other comparison pdfs change because previously blank phone # skipped line, now blank phone # adds a blank line * Move instance_method call inside function

jimmyli97 · 2024-12-03T20:49:11Z

@cielf I made a solution in 4c6cd65 where if there's no partner primary contact info it fills it in with a blank space, but I could also do it so that it uses the partner's name and email (which are both guaranteed to not be blank) if you want

cielf · 2024-12-04T00:17:57Z

Better! It does feel to me like there is a bit too much space between the Issued To and the Delivery Address, though. Is it possible to tighten that up without side effects?

jimmyli97 · 2024-12-04T03:41:18Z

fixed spacing in b00d1b0

cielf · 2024-12-04T11:05:30Z

Functionality works great! Thank you! Back to @dorner for final tech review.

dorner

Almost there! Had some suggestions around the helpers (I know, this has been going around forever, but I think one last round should do it).

dorner · 2024-12-10T01:05:39Z

lib/test_helpers/pdf_comparison_test_factory.rb

+
+  private def create_comparison_pdf(storage_creation, profile_create_method, expected_file_path, delivery_method)
+    partner = create_partner(storage_creation.organization)
+    profile = PDFComparisonTestFactory.instance_method(profile_create_method).bind_call(Class.new.extend(PDFComparisonTestFactory), partner)


This is a bit over-the-top... can we not just move the possible inputs to this to class methods? Then you can just call PDFComparisonTestFactory.send(profile_create_method).

dorner · 2024-12-10T01:06:42Z

lib/test_helpers/pdf_comparison_test_factory.rb

+    profile.destroy!
+    dist.request.destroy!
+    dist.destroy!
+    partner.destroy!


Probably better to just wrap this in a transaction and roll it back (i.e. raise an ActiveRecord::Rollback) rather than manually calling destroy.

dorner · 2024-12-10T01:07:05Z

lib/test_helpers/pdf_comparison_test_factory.rb

+    storage_creation.items[2].destroy!
+    storage_creation.items[3].request_units.destroy_all
+    storage_creation.items[3].destroy!
+    storage_creation.organization.destroy!


I guess the transaction/rollback can be moved to this method.

dorner · 2024-12-10T01:09:13Z

spec/pdfs/distribution_pdf_spec.rb

+
+RSpec.configure do |c|
+  c.include DistributionPDFHelper
+  c.include PDFComparisonTestFactory


These shouldn't need to be included in the base RSpec config. I generally prefer moving everything to module/class methods instead of instance methods. There aren't any instance variables, so doing that makes everything way easier to work with - so you can just call PDFComparisonTestFactory.create_comparison_pdfs directly, and don't have to pollute existing specs or configs.

jimmyli97 added 3 commits July 19, 2024 01:15

REFACTOR group examples, remove rubocop disable lines,

4bb8332

* group examples together with variables * rubocop -A passes on this file without needing to disable ArrayAlignment

RED add rspec to test address output in distribution PDFs

77b2a55

* Compare generated against expected PDFs * Add helper test to regenerate expected PDFs * Add helper module

GREEN Fix rubyforgood#4138 address output changes based on delivery m…

dcdc0f8

…ethod * Address output prints delivery address if filled in, otherwise partner address * Only does this for delivery/shipped method

cielf requested a review from dorner July 19, 2024 15:10

cielf previously requested changes Jul 19, 2024

View reviewed changes

jimmyli97 added 3 commits July 23, 2024 03:21

BUGFIX pdf now writes when test fails

aa38b4b

RED add rspec for when partner has no addresses

f1d7273

* REFACTOR out profile name and email

GREEN don't print address if no address or delivery address

d64acc4

jimmyli97 requested a review from cielf July 23, 2024 03:39

jimmyli97 mentioned this pull request Jul 25, 2024

[PACKS] # 9 Distribution printout for banks includes custom request units, if applicable #4404

Closed

2 tasks

dorner requested changes Sep 2, 2024

View reviewed changes

jimmyli97 added 3 commits September 12, 2024 00:53

Merge branch 'main' into 4138-fix-delivery-address

4557128

Add comments clarifying gitignore and comparison pdf helper rspec

0d53723

Update bundler version

d722bca

jimmyli97 requested a review from dorner September 12, 2024 01:34

jimmyli97 added 2 commits September 24, 2024 00:15

REFACTOR replace helper test with environment variable

4516951

* Use environment variable for regenerating comparison pdfs

Merge branch 'main' into 4138-fix-delivery-address

ee159ab

jimmyli97 added 4 commits October 17, 2024 01:32

Merge branch 'main' into 4138-fix-delivery-address

cec74ce

update schema date and bundler version

c26f94b

RED remove env variable code, stub out console helper method

df4b975

GREEN add Rails console method for generating comparison pdfs, update…

26ece6e

… comparison PDFs

jimmyli97 force-pushed the 4138-fix-delivery-address branch from 922afc6 to 26ece6e Compare October 17, 2024 22:02

dorner requested changes Oct 22, 2024

View reviewed changes

jimmyli97 added 4 commits November 27, 2024 21:44

Merge branch 'main' into 4138-fix-delivery-address

a21c7f2

Replace instance vars with structs and let

6da7702

FIX destroy request so db is clean, use destroy! instead so exception…

8bc117e

…s fire

Replace FactoryBot with calling ActiveModel.create, move methods to lib/

ea7e241

jimmyli97 force-pushed the 4138-fix-delivery-address branch from 10e5b18 to ea7e241 Compare November 28, 2024 03:21

jimmyli97 requested a review from dorner November 28, 2024 03:35

cielf requested changes Nov 28, 2024

View reviewed changes

Fix formatting when address is incomplete, add rspecs, create and del…

39e8cb4

…ete item units correctly when generating test files

jimmyli97 requested a review from cielf December 3, 2024 03:09

Merge branch 'main' into 4138-fix-delivery-address

29e59df

cielf requested changes Dec 3, 2024

View reviewed changes

FIX replace with space if partner primary contact name/email/phone is…

4c6cd65

… blank * Add rspec * Other comparison pdfs change because previously blank phone # skipped line, now blank phone # adds a blank line * Move instance_method call inside function

jimmyli97 requested a review from cielf December 3, 2024 20:49

FIX spacing between issued to and delivery address

b00d1b0

dorner requested changes Dec 10, 2024

View reviewed changes

Fix #4138 fix delivery address #4540

Are you sure you want to change the base?

Fix #4138 fix delivery address #4540

Conversation

jimmyli97 commented Jul 19, 2024

Description

Type of change

How Has This Been Tested?

cielf commented Jul 19, 2024

jimmyli97 commented Jul 19, 2024

cielf commented Jul 19, 2024 • edited Loading

cielf commented Jul 19, 2024 • edited Loading

cielf left a comment • edited Loading

Choose a reason for hiding this comment

cielf commented Aug 25, 2024

dorner left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jimmyli97 Sep 12, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jimmyli97 Sep 12, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

cielf commented Sep 28, 2024

dorner commented Sep 30, 2024

jimmyli97 commented Oct 17, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jimmyli97 commented Nov 28, 2024

cielf left a comment

Choose a reason for hiding this comment

jimmyli97 commented Dec 3, 2024

jimmyli97 commented Dec 3, 2024 • edited Loading

cielf left a comment • edited Loading

Choose a reason for hiding this comment

jimmyli97 commented Dec 3, 2024

cielf commented Dec 4, 2024

jimmyli97 commented Dec 4, 2024

cielf commented Dec 4, 2024

dorner left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

cielf commented Jul 19, 2024 •

edited

Loading

cielf commented Jul 19, 2024 •

edited

Loading

cielf left a comment •

edited

Loading

jimmyli97 Sep 12, 2024 •

edited

Loading

jimmyli97 Sep 12, 2024 •

edited

Loading

jimmyli97 commented Dec 3, 2024 •

edited

Loading

cielf left a comment •

edited

Loading