WebPathResolver - sanitize URL to directory name

[teohhanhui]

No description provided.

[makasim]

@teohhanhui it is not clear what this realy fix. could you give more information on why do we need this?

[teohhanhui]

Before this change Filesystem::dumpFile throws IOException when trying to cache external images, as it's not possible to create a directory with "/" in its name. This is a rudimentary fix.

[makasim]

ok, I see, could you add a test then?

[makasim]

sorry for the delay... I've just found out that you had added some tests.

What about other resolvers, I think this fix has to be applied on a higher level, maybe inside the cache manager

[teohhanhui]

I disagree. Path handling should be left to each resolver as it currently
is (e.g. what makes sense for the filesystem might not make sense for S3).
On Oct 23, 2014 1:56 AM, "Maksim Kotlyar" [email protected] wrote:

sorry for the delay... I've just found out that you had added some tests.

What about other resolvers, I think this fix has to be applied on a higher
level, maybe inside the cache manager

—
Reply to this email directly or view it on GitHub
#480 (comment)
.

[makasim]

I can try to pass an url to s3 too, what would be key on s3 in this case?

[teohhanhui]

That should be left up to the S3 resolver to decide how to normalize the
URL to S3 key... My point is, I don't think there's a one-size-fits-all
solution here.
On Oct 23, 2014 2:44 PM, "Maksim Kotlyar" [email protected] wrote:

I can try to pass an url to s3 too, what would be key on s3 in this case?

—
Reply to this email directly or view it on GitHub
#480 (comment)
.

[makasim]

okay okay