Hi all,
XML feeds are very useful, however it is not very efficient to download the entire iSheet to find the one(or more) records that I want to display on screen.
Has anyone found a workaround to leverage search parameters in the xmls feeds,? (similar to how advanced searches work)
Thanks,
Referencing my duplicate post on this topic (apologies) -- https://collaborate.highq.com/sitepoint/viewBlog.action?metaData.siteID=714&metaData.blogID=7432. I agree that overall it would be better to find a way to leverage making REST API calls from within the page, rather than enhancing the XML feed. That said, having been in software development for a long time, that's going to be a big lift. I suspect that enhancing the XML feed in a two small ways would be fairly easy to implement and sooner, and would have a large impact on many existing projects, echoing what David Corrin said.
1 - Support the same parameters used in Advanced Search (and used for drill down purposes from the charts). My limited testing shows that this seems to work in a very limited and inconsistent fashion. My suspicion is that under the hood there is some common code that would make this possible. Note that this url parameter that can be used with the web URL DOES work to limit what is returned to a single item: &metaData.itemId={ITEMID} . (Another reason I believe the XML and web URL have common code paths.)
2 - Fill in the content gaps in the XML feed. For example, include the fileid for files used in file links (and folder ids for folder links). I have not put together a comprehensive inventory of all of the gaps. I find for most column types the information returned by the XML is sufficient if not complete. The gaps seem fairly small. Andrew Quinn you may know of others? This information is obviously available and I can't imagine adding it to the XML feed could be difficult.
More complex would be supporting a feature like pagination. If we had the ability to filter, that would help in a big way and pagination would not be as important in many situations. I would prioritize the two features above, which are likely the easiest, then consider whether other enhancements make sense.
Mark Salamon - any chance you have experimented much with the +itemid and found a reliable way to download multiple records via this method rather than having to download a huge view?
Is it quicker to download a small view for purely getting their items ids ( 1 or 2 columns) and then target downloading multiple records(1,10,20,100 etc?) individually rather than downloading a view of 500 records with 500 datapoints, or 5k records with 500 datapoints, 1k records with 800 datapoints etc?
The ability to search using a parameter would fix this immediately but wondering if there is some reliable workaround for the time being and what potentially are the boundaries for when it is useful.
David Corrin That's an interesting idea. I have not done that, although technically i don't know why it wouldn't work. I think it depends on the scale of your data. You mentioned having 800 datapoints (columns?) in an iSheet? We have not had anything close to that. In that case, I think it would make sense to do as you propose.
Mark Salamon - I am just to get an idea if there were any boundaries to this essentially (multiple single records v 1 view) and its interaction with small normal iSheets compared to theoretically larger iSheets. Also need to see if loading multiple feeds is reliable as well.
Andrew Quinn - any chance you would have any thoughts on this?
Natasha Miller-Santos Happy near year! Any chance there has been any further progress/discussions made on the above? The ability for advanced search within an export xml would make my year! This exists through the front end by appending &advanceSearch=true&filterValue_406123_1=12345 to the url, so I'm hopeful that a lot of the underlying infrastructure is already present and it wouldn't be a massive undertaking to enable the same functionality on the internal export feed.
Xmlfeed 2.0 or a totally revamped way of working in the backend would be awesome but adding advanced search functionality would add a lot of value with (I am hoping) relatively low effort. Inclusion of missing datapoints (file ids etc) would also be useful but at least these can be retrieved using the injectxml feed.
Hi David Corrin, Happy new year! Not at the moment I'm afraid, our focus remains on improving our api's and its all hands on deck for 5.5. However we have lots of valuable feedback here in this blog, so once the conversations do start I'll include you - thanks and have a great weekend!
David Corrin have you tried the iSheet API - Advanced Search? It will return XML.
Morgan, the difference is that the XML feeds are purely in the browser. You can generate visualizations and other things directly on a Collaborate rich-text page without creating access tokens, etc. when using a REST API call, which can compromise security. But the XML feeds are very limited.
Mark Salamon For some reason I was thinking David was talking about exporting xml from an iSheet (case of Friday lack-of--reading-comprehension skills, clearly). I haven't seen the XML feeds at all but it sounds interesting if not limited.
