search

strongloop / strong-globalize

strong-globalize is built on Unicode CLDR and jquery/globalize and implements automatic extraction of strings from JS source code and HTML templates, lint the string resource, machine-translate them in seconds. In runtime, it loads locale and string resource into memory and provides a hook to persistent logging.
Other
25 stars 16 forks source link

Remove 32 char hash from txt file name in -d mode #90

Closed Setogit closed 8 years ago

Setogit commented 8 years ago

connect to https://github.com/strongloop/strong-globalize/issues/88

Setogit commented 8 years ago

@bajtos PTAL

bajtos commented 8 years ago

What are the consequences for existing applications? Is it expected to have the copied resource files submitted to git? If yes, then what will happen to applications that are already using long file paths? Will strong-globalize still load the files with long paths (starting with 32 char hash code prefix)?

There are no tests covering this change, could you please add some?

Last but not least, there are quite few downstream CI failures, are we sure they are not related?

Setogit commented 8 years ago

Quickly examined the CI failures: I see 20 failing, 4 pending, 76 successful checks as of now. Approx. 60% of the 20 is clearly unrelated. 30% is probably not , but not that clear either without module specific knowledge. 10%, the log actually says success, but listed above as failed. No changes in loading side, i.e., if both exist, both will be loaded; if dup, overriden; those are important features. Offline, slt-globalize -l detects duplicates.

TO-DO: document it in readme and add some tests. Consider minor version up.

bajtos commented 8 years ago

Sounds reasonable. I'll trust your judgment, @Setogit.

document it in readme and add some tests

Great. Please let me know when you are done so that I can review again.

Consider minor version up.

Agreed, let's bump the minor version.

Setogit commented 8 years ago

No README changes because it does not mention the 32 char hash in deep-extracted txt file names.

Setogit commented 8 years ago

Moved the obsolete file removal out of the loop.