Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

AbsoluteUri vs OriginalUri vs IdnMapping #368

Closed
jamesmontemagno opened this issue Jul 10, 2018 · 3 comments
Closed

AbsoluteUri vs OriginalUri vs IdnMapping #368

jamesmontemagno opened this issue Jul 10, 2018 · 3 comments
Assignees

Comments

@jamesmontemagno
Copy link
Collaborator

Currently in our code we are using OriginalUri, however this causes issues for Unicode URL. AbsoluteUri could be better too, but perhaps not in all instances....

https://stackoverflow.com/questions/35875376/uri-absoluteuri-vs-uri-originalstring

One recommendation is to use Idn mapping for Urls...
https://stackoverflow.com/questions/36685160/unicode-url-could-not-initialize-an-instance-of-the-type-foundation-nsurl-t

This directly impacts the Browser API, #356, and #355 and should be assessed before pulling in.

@Mrnikbobjeff
Copy link
Contributor

For the Browser API we definitely should use AbsoluteUri, because OriginalUri does not escape strings. This is especially desired by the browser. Relative URI support is not very important for the browser API which is one of the weaknesses of AbsoluteUri. The benefit of using OriginalUri is that it shifts all responsibility for URI validity to the consumer. OriginalUri does exactly what the name implies, returns the string used to construct the uri.

@Redth
Copy link
Member

Redth commented Jul 18, 2018

So my understanding is we should probably use IDN mapping for hosts, and percent encoding for the path and query.

@Redth
Copy link
Member

Redth commented Jul 18, 2018

I have a PR open for this #393

Basically, the PR fixes up any puny code (IDN) domain names, and then uses the % encoded path and query from the uri (so what AbsoluteUri is doing) to construct a safe encoded version of the URL to pass to each platform.

The platforms seem to all handle this correctly.

jamesmontemagno pushed a commit that referenced this issue Jul 19, 2018
* Allow for unicode in host names and escape query

Previously we were passing the the url as is to each platform and letting it deal with how to parse into the platform specific url.

This handles unicode domain names (IDN Mapping aka PunyCode) for the domain, but also reconstructs the url based on the parsed System.Uri from the original string, and so the PathAndQuery will be the % escaped version which will account for unicode characters in it as well.

Finally, on iOS and Android we are using AbsoluteUri to end up with the fully IDN mapped and % encoded URL to pass to the system URL types.

* Extract Escape Uri logic into its own method

More testable.

* Add uri escaping tests

* Improve uri escape tests

* Update default iOS DeviceTests target

Also don’t allow picking a sim that’s unavailable (this was a bug since we were checking for contains `available` which of course `unavailable` also contains!)
jamesmontemagno added a commit that referenced this issue Jul 31, 2018
* GH-343: Take into consideration VerticalAccuracy on iOS Altitude (#344)

* Fix for #343

Only set altitude if it's available (VerticalAccuracy is negative when it isn't according to iOS doc)

* Use: (double?)

* dotnet foundation CLA broken link (#349)

* Update ns-Xamarin.Essentials.xml

* Update README.md (#340)

* GH-365 Change to generic EventHandlers

* Clean up unit tests.

* GH-363 Enable C# 7.2 (#364)

* GH-363 Enable CD 7.3

* Update Xamarin.Essentials.csproj

* GH-367 Change Ui to UI & Ac to AC (#374)

* GH-367 Change Ui to UI

* Ensure proper Ui for Android

* Fix gryoscope

* Rename Ac to AC

* Update framework indexes

* Create Product Feedback page for Docs (#369)

* Create PRODUCT-FEEDBACK.md

* Update PRODUCT-FEEDBACK.md

* Fixed a typo from "chanaged" to "changed"

* Fixed a typo from "chanaged" to "changed"

* Fixed a typo from "chanaged" to "changed"

* GH-337 Android Low Pass Filter (#354)

* Implemented low pass filter as requested by issue 337

* Added value to sample page.

* Updated docs for ApplyLowPassFilter

* Enabled ApplyLowPassFilter Switch on Android, disabled on all other platforms

* Moved LowPassFilter to shared, removed unnecessary usings.

* Updated Docs

* Sample Project Tweaks and Optimizations (#353)

* Make sure orientation sensor stops on sample.

* Fix up all sensors page to start and stop correctly

* Only check permission if unknown status

* Update to latest sdk.extras

* fix back button press on TTS page. Check for null.

* Add samples sln for CI/CD

* Update readme and samples sln

* Don't use code sign key for simbuilds

* let's try this gain

* Update to projects

* Cleanup sample config

* bump sdk extras

* Add in readme for install with helpful information. Bump MSBuild.Sdk.Extras (#392)

* GH-368: Allow for unicode in host names and escape query (#393)

* Allow for unicode in host names and escape query

Previously we were passing the the url as is to each platform and letting it deal with how to parse into the platform specific url.

This handles unicode domain names (IDN Mapping aka PunyCode) for the domain, but also reconstructs the url based on the parsed System.Uri from the original string, and so the PathAndQuery will be the % escaped version which will account for unicode characters in it as well.

Finally, on iOS and Android we are using AbsoluteUri to end up with the fully IDN mapped and % encoded URL to pass to the system URL types.

* Extract Escape Uri logic into its own method

More testable.

* Add uri escaping tests

* Improve uri escape tests

* Update default iOS DeviceTests target

Also don’t allow picking a sim that’s unavailable (this was a bug since we were checking for contains `available` which of course `unavailable` also contains!)

* Update iOS DeviceInfo.Model to return specifid hw.machine for more control. (#395)

via:
https://github.com/xamarin/xamarin-macios/blob/bc492585d137d8c3d3a2ffc827db3cdaae3cc869/tests/linker/ios/link%20sdk/LinkSdkRegressionTest.cs#L603-L631

Added tests

* GH-380 Use Internal Preferences for Secure Storage consistency. (#386)

* Use Internal Preferences for Secure Storage consistency.
Save if we created key pre-M so we always use pre-M if device upgrades.

* Fix logic for pre-m key check

The logic was slightly off, I think this fixes it, but would be good to have another set of eyes...

1. We check to see if the device is pre-M (if it does _not_ have `M`, or if we already set the fact it's pre-M in the preference - aka from a previous install before an upgrade of the OS)
2. If we aren't pre-M, we can't use Symmetric Key from Keystore
3. If we make it down to using Asymmetric Key, we set the pre-M preference to `true` to persist the value for future invocations, which will make it 'stick' in the event of a pre-M to M+ OS upgrade.

* Address feedback on key naming.

* Added test for secure storage to simulate upgrade

From API < 23 to API >= 23 after storing data with an asymmetric key and then moving to a platform supporting symmetric keys.

* Ensure we always set flags when using specified keygen

* GH-388 If activity is required and null throw null exception with information. (#396)

* If activity is required and null throw null exception with information.

* Remove currentactivity and update getcurrentactivity.

* GH-250 Implement Launcher API (#356) (#405)

* GH-250 Implement Launcher API (#356)

* Initial commit for Launcher Api as discussed in Issue 250

* Fixed and refactored android launch code, fixed unit tests. Fixed namespaces.

* Refactored code, fixed uwp launcher. Added unit test

* Added sample launcher page and viewmodel.

* Fixed launcher Page typo

* Refactored launcher code. after testing exception behaviour I decided on rethrowing instead of swallowing the exception

* Refactored access modifiers. Also changed exception behaviour to enforce uri validity and exception behaviour across platforms.

* Updated docs, stripped even more code.

* Added validation. Adressed all other changes as required in pr

* Removed dead code

* Adressed last issues from pr

* Cleanup launcher

* cleanup iOS tests

* Early morning typos

* Name Alignment for BrowserLaunchMode (#408)

* Rename BrowserLaunchType to BrowserLaunchMode

* update nuget

* Rename parameter

* GH-287 OpenMaps Implementation (#361) (#404)

* GH-287 Maps Implementation (#361)

* Implemented maps as mentioned in branch name issue. also fixed a spelling mistake

* Added samples

* Added doc stubs

* Added comments for map types

* Refactored code

* Formatting of docs

* Added tests

* Changed coordinates to display something recognisable

* Added uri escaping, thus removing duplicate code. also had to turn off a warning due to inconsistent style cop and vs warning behaviour.

* Removed ref until in is added in c# 7.2

* Adressed issues in pr

* Updated launch code

* Adressed method to onliner

* Updated docs

* Removed ClearTop intent, added sample with adress

* Rename to align names. Added extensions and additional paramaters.

* Update tests

* Throw if options are null.

* Add overload for lat/long without options

* Enable multi-target for 25,26,27 (#411)

Alight nuget versions to xamarin.forms.

* No More Shared Projects (#410)

* Change to multi-targeted project for device tests :)

* Cleanup scripts

* Fix build error due to preprocessor directive (#417)

* Sync Sensor Speeds & Reset UWP to 0 (#419)

* Restore the default report interval to release resources while the sensor is not in use

* Cleanup sensor reading times.
Make them all the same and only in 1 place
Compass is still unique

* Create unique preference storage for each feature. (#434)

* Create unique preference storage for each feature.

* update storage per @Redth

* Update version
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants