> + if (from.containsKey("size"))
> + builder.size(Long.valueOf(from.get("size")));
> + if (from.containsKey("rawsize"))
> + builder.rawSize(Long.valueOf(from.get("rawsize")));
> + if (from.containsKey("format"))
> + builder.format(ImageConversionType.fromValue(from.get("format")));
> + Map<String, String> metadata = Maps.newLinkedHashMap();
> + for (Entry<String, String> entry : from.entrySet()) {
> + if (entry.getKey().startsWith("user:"))
> +
> metadata.put(entry.getKey().substring(entry.getKey().indexOf(':') + 1),
> entry.getValue());
> + }
> + builder.userMetadata(metadata);
> + try {
> + return builder.build();
> + } catch (NullPointerException e) {
> + logger.trace("entry missing data: %s; %s", e.getMessage(), from);
As far as I understand the code, missing data here (and the `null` returned in
that case) will translate in the StandardDrive (the Image) not being returned.
So I see it as "if some image is missing a field jclouds needs, we're not
returning them and can't be used in the portable abstraction".
This is how it was working before the change. But it's worth an issue if we
feel this is not the right behavior.
---
Reply to this email directly or view it on GitHub:
https://github.com/jclouds/jclouds/pull/362/files#r13054473
