
|
If you were logged in you would be able to see more operations.
|
|
|
| Proposal: |
add to domain, ws + rest
same name siblings is repository specific behavior. cmis does not define uniqueness or how to handle non-unique scenarios.
add to domain, ws + rest
same name siblings is repository specific behavior. cmis does not define uniqueness or how to handle non-unique scenarios.
|
|
reported in various discussions during the PlugFest in Basel, my interpretation may be off.
It would be great to have at least a "get folder by path" that allows to reflect the
hierarchy that we expose through the parent child relationships of folders.
Generally there are a lot of usecases (exposure to filesystem like trees, relative links from document to documents) that would
benefit from a path. This does not stand in contrast with having multiple paths to the same document or folder.
Is anyone opposed to introducing paths into CMIS?
|
|
Description
|
reported in various discussions during the PlugFest in Basel, my interpretation may be off.
It would be great to have at least a "get folder by path" that allows to reflect the
hierarchy that we expose through the parent child relationships of folders.
Generally there are a lot of usecases (exposure to filesystem like trees, relative links from document to documents) that would
benefit from a path. This does not stand in contrast with having multiple paths to the same document or folder.
Is anyone opposed to introducing paths into CMIS? |
Show » |
|
do we need getFolderByPath() or getObjectByPath()?
Issue: same name siblings, a repository might support more than one item having the same name. In this case it could not implement this function. Should it just throw a specific exception in this case?
do we need a mapping in both directions? In the other direction multi-filing would be an issue (no unique path to an object).
Should path navigation be an optional capability of the repository?
Therefore might be tough to specify. However path navigation is such a common pattern (file system, URL, WebDAV, ..) that we really should think about if this isn't the 80% part and the issues are the 20%...
Use Cases: WebDAV mapping, simple CMIS browser based on URLs
All in all it would be good to get a decision from the TC how to deal with this as this issue arises over and over again.