Smalltalk › Squeak › Squeak - Dev

LanguageEnvironment

_‹ Previous Topic Next Topic _›

Classic

List

Threaded

4 messages Options

Eliot Miranda-2

LanguageEnvironment

Hi All,

in reviewing the tests given recent work leading up to the 5.2 release I see an error in LangEnvBugs>>testIsFontAvailable.

This is the phrase:

(LanguageEnvironment localeID: 'en' ) isFontAvailable

But LanguageEnvironment class>>#localeID: expects a LocaleID as argument. So is the fix to rewrite the test to use

(LanguageEnvironment localeID: (LocaleID isoLanguage: 'en')) isFontAvailable

or to be more friendly:

localeID: localeIDOrISONameString

"LanguageEnvironment localeID: (LocaleID isoString: 'ja-kid')"

"LanguageEnvironment localeID: (LocaleID isoString: 'xx')"

localeIDOrISOName isString ifTrue:

[^self localeID: (LocaleID isoLanguage: localeIDOrISOName)].

^ self knownEnvironments

at: localeIDOrISONameString

ifAbsent: [localeIDOrISONameString hasParent

ifTrue: [self knownEnvironments

at: localeIDOrISONameString parent

ifAbsent: [self localeID: 'en']]

ifFalse: [self localeID: 'en']]

And given that

self ensureInternetConnectionTo: 'http://metatoys.org/pub/'.

fails and

curl http://metatoys.com/pub

<html><head><title>Object moved</title></head><body>

<h2>Object moved to <a href="https://www.hugedomains.com/domain_profile.cfm?d=metatoys&e=com">here</a>.</h2>

</body></html>

shouldn't LocaleTest>>#testIsFontAvailable be n expected failure?

_,,,^..^,,,_

best, Eliot

Levente Uzonyi

Re: LanguageEnvironment

On Wed, 13 Jun 2018, Eliot Miranda wrote:

> Hi All,
> in reviewing the tests given recent work leading up to the 5.2 release I see an error in LangEnvBugs>>testIsFontAvailable.
>
> This is the phrase:
>
> (LanguageEnvironment localeID: 'en' ) isFontAvailable
>
> But LanguageEnvironment class>>#localeID: expects a LocaleID as argument. So is the fix to rewrite the test to use
>
> (LanguageEnvironment localeID: (LocaleID isoLanguage: 'en')) isFontAvailable
>
> or to be more friendly:
>
> localeID: localeIDOrISONameString
>
> "LanguageEnvironment localeID: (LocaleID isoString: 'ja-kid')"
> "LanguageEnvironment localeID: (LocaleID isoString: 'xx')"
> localeIDOrISOName isString ifTrue:
> [^self localeID: (LocaleID isoLanguage: localeIDOrISOName)].
> ^ self knownEnvironments
> at: localeIDOrISONameString
> ifAbsent: [localeIDOrISONameString hasParent
> ifTrue: [self knownEnvironments
> at: localeIDOrISONameString parent
> ifAbsent: [self localeID: 'en']]
> ifFalse: [self localeID: 'en']]
>
> ?

Sounds good.

>
>
> And given that
> self ensureInternetConnectionTo: 'http://metatoys.org/pub/'.
> fails and
> curl http://metatoys.com/pub
> <html><head><title>Object moved</title></head><body>
> <h2>Object moved to <a href="https://www.hugedomains.com/domain_profile.cfm?d=metatoys&e=com">here</a>.</h2>
> </body></html>
>
> shouldn't LocaleTest>>#testIsFontAvailable be n expected failure?

It should. Here is our previous discussion about this test case:
http://forum.world.st/LocaleTest-testIsFontAVailable-td4939155.html

Levente

>
> _,,,^..^,,,_
> best, Eliot
>
>

Tim Johnson-2

Re: LanguageEnvironment

In reply to this post by Eliot Miranda-2

Hi Eliot,

On Jun 13, 2018, at 9:49 AM, Eliot Miranda wrote:

Hi All,

in reviewing the tests given recent work leading up to the 5.2 release I see an error in LangEnvBugs>>testIsFontAvailable.

Thanks for bringing this up. I noticed this one too and spent an hour or two working on it but never sent along my changes... maybe because I had some of the same questions.

This is the phrase:

(LanguageEnvironment localeID: 'en' ) isFontAvailable

But LanguageEnvironment class>>#localeID: expects a LocaleID as argument. So is the fix to rewrite the test to use

[...]

?

Yes, I think so. That meshes with the other users of this system within the image.

And given that
self ensureInternetConnectionTo: 'http://metatoys.org/pub/'.
fails and
curl http://metatoys.com/pub
<html><head><title>Object moved</title></head><body>
<h2>Object moved to <a href="https://www.hugedomains.com/domain_profile.cfm?d=metatoys&e=com">here</a>.</h2>
</body></html>

shouldn't LocaleTest>>#testIsFontAvailable be n expected failure?

Good call. It seems that without this URL being live on the internet, the remote-font-loading-for-locale-switching won't work (Locale class>>#switchAndInstallFontToID:gently:)

We can change the URL to https://web.archive.org/web/20161219105306/http://metatoys.org:80/pub/

...or provide the necessary files on files.squeak.org.

We can also rewrite the test so it does not hardcode the URL but rather takes its URL from LanguageEnvironment>>#fontDownloadUrls

However, I have to ask: was this remote-font-loading system affected by the font system refurbishment during the Squeak license change? Is this remote-font-loading-for-locale-switching system still relevant?

Thanks,

Tim

Tim Johnson-2

Re: LanguageEnvironment

On Jun 17, 2018, at 9:48 AM, Tim Johnson wrote:

> Good call. It seems that without this URL being live on the internet, the remote-font-loading-for-locale-switching won't work (Locale class>>#switchAndInstallFontToID:gently:)
>
> We can change the URL to https://web.archive.org/web/20161219105306/http://metatoys.org:80/pub/
>
> ...or provide the necessary files on files.squeak.org.
>
> We can also rewrite the test so it does not hardcode the URL but rather takes its URL from LanguageEnvironment>>#fontDownloadUrls
>
> However, I have to ask: was this remote-font-loading system affected by the font system refurbishment during the Squeak license change? Is this remote-font-loading-for-locale-switching system still relevant?

Replying to myself: I see Eliot's concerns were addressed in The Trunk: Tests-ul.398.mcz. Thanks to Levente Uzonyi.

I still wonder about the viability of the remote font loading system for locales, and my questions above.

Thanks,
Tim