Update aid in attachments tests

[Kharchevskyi]

This PR updates aid to attachments tests

close #1218

---

Tests:

• Tests updated

---

To be filled by reviewers

I have reviewed that this PR... (tick whichever items you personally focused on during this review):

• addresses the issue it closes (if any)
• code is readable and understandable
• is accompanied with tests, or tests are not needed
• is free of vulnerabilities
• is documented clearly and usefully, or doesn't need documentation

[Kharchevskyi]

@fcvakintos @sosnovsky @tomholub
Can you guys suggest me your setup for writing/executing Appium UI Tests?
What IDE are you using? Any tools except of appium-inspector?
Do I need to build app bundle exec fastlane build each time for verifing tests? or it is possible to execute test on builded app from Xcode (from DerivedData folder)?

[sosnovsky]

I write appium tests in VS Code, as it supports eslint and highlights code issues.
For testing I just copy FlowCrypt.app from /DerivedData/FlowCrypt-.../Build/Products/Debug-iphonesimulator to appium folder. No need to build app on each test run.

[Kharchevskyi]

@sosnovsky great, thanks. Will try VS Code.

[tomholub]

@Kharchevskyi read steps at https://github.com/FlowCrypt/flowcrypt-ios/tree/master/appium first to get an overview. After that, if the information you needed to know was missing there, please add it (which IDE to use, etc).

In general, this kind of basic knowledge shouldn't be passed only person to person - it should be documented for the long term so that new team members, or our own non-ios developers can get involved easily when needed.

[Kharchevskyi]

@sosnovsky @fcvakintos
Can you also suggest best practices to investigate issues?
Is it possible to enable logs or so?

@tomholub will add this to Readme as well

[tomholub]

Your screenshot is missing reason why the test failed. It's supposed to have appium stack trace and show which test method was run and which selector was attempted etc.

With that information (missing above) it usually becomes clear. If still not clear, I watch the simulator go through the steps while looking at the source code of the tests. So that it becomes clear where it's getting stuck, and usually also why.

Try deleting contents of appium/tmp folder before the run, and then run it and see the files it produces there (else it's hard to find which file to look at). There should be some logs produced by the run, at least.

Looks like your test is failing for some fundamental reason, did it ever produce a simulator and run any steps?

[sosnovsky]

I was also able to get more appium logs by changing logLevel to debug here https://github.com/FlowCrypt/flowcrypt-ios/blob/master/appium/config/wdio.shared.conf.js#L19

[tomholub]

You guys should stop hoarding that kind of information 😆 if it's very useful, document it

[Kharchevskyi]

@tomholub indeed appium stoped even launching a simulator and executing any tests.
Was trying to debug this. tmp folder is empty.

Only changes from yesterday successfull runs is only installed VS Code.

I have pushed latest commit to check whether this issue only locally on my machine, but maybe someone can suggest me where the issue could be?

[tomholub]

It doesn't even launch simulator? try the debug log suggested above

[Kharchevskyi]

[0-9] 2021-12-16T15:59:24.223Z DEBUG @wdio/config:utils: Found 'ts-node' package, auto-compiling TypeScript files
[0-9] 2021-12-16T15:59:24.307Z DEBUG @wdio/local-runner:utils: init remote session
[0-9] 2021-12-16T15:59:24.310Z DEBUG @wdio/utils:initialiseServices: initialise service "appium" as NPM package
[0-9] RUNNING in /Users/antonkharchevskyi/git/flowcrypt-ios/appium/FlowCrypt.app - /tests/specs/live/inbox/ReadAttachmentEmail.spec.ts
[0-9] 2021-12-16T15:59:24.820Z DEBUG @wdio/local-runner:utils: init remote session
[0-9] 2021-12-16T15:59:24.821Z INFO webdriver: Initiate new session using the WebDriver protocol
[0-9] 2021-12-16T15:59:24.904Z INFO webdriver: [POST] http://localhost:4723/session
[0-9] 2021-12-16T15:59:24.904Z INFO webdriver: DATA {
[0-9]   capabilities: {
[0-9]     alwaysMatch: {
[0-9]       platformName: 'iOS',
[0-9]       iosInstallPause: 5000,
[0-9]       deviceName: 'iPhone 13',
[0-9]       platformVersion: '15.0',
[0-9]       automationName: 'XCUITest',
[0-9]       app: '/Users/antonkharchevskyi/git/flowcrypt-ios/appium/FlowCrypt.app',
[0-9]       simpleIsVisibleCheck: true,
[0-9]       newCommandTimeout: 10000,
[0-9]       wdaLaunchTimeout: 300000,
[0-9]       wdaConnectionTimeout: 600000,
[0-9]       wdaStartupRetries: 4,
[0-9]       wdaStartupRetryInterval: 120000
[0-9]     },
[0-9]     firstMatch: [ {} ]
[0-9]   },
[0-9]   desiredCapabilities: {
[0-9]     platformName: 'iOS',
[0-9]     iosInstallPause: 5000,
[0-9]     deviceName: 'iPhone 13',
[0-9]     platformVersion: '15.0',
[0-9]     automationName: 'XCUITest',
[0-9]     app: '/Users/antonkharchevskyi/git/flowcrypt-ios/appium/FlowCrypt.app',
[0-9]     simpleIsVisibleCheck: true,
[0-9]     newCommandTimeout: 10000,
[0-9]     wdaLaunchTimeout: 300000,
[0-9]     wdaConnectionTimeout: 600000,
[0-9]     wdaStartupRetries: 4,
[0-9]     wdaStartupRetryInterval: 120000
[0-9]   }
[0-9] }
[0-9] 2021-12-16T16:00:54.881Z WARN webdriver: Request timed out! Consider increasing the "connectionRetryTimeout" option.

UPD: Removing node_modules and npm install helped once again)
Thanks!

[Kharchevskyi]

@tomholub are you ok with this log level? it is pretty verbose but easier to debug issues

[tomholub]

Surely there could be some options in between silent and debug. Do they have some sort of "normal" verbosity level?

[Kharchevskyi]

debug, info, warn, error

[tomholub]

Let's try info, see if that gives enough clues without being too verbose

[tomholub]

Please pay attention to the whole document holistically.. documentation must be as clean and orderly as code, if not more so.

[Kharchevskyi]

@tomholub are you ok with changing accessibility ids gradually? or do I need to update all of existed?

[sosnovsky]

Looks good 👍

[tomholub]

This looks like you need to take both step 1 and then step 2 in sequence. Which is not true. These two are two alternative approaches, and should be documented as such.

[tomholub]

filed a separate issue for the comment above