Graph support

[wmoore]

Need to know what graph loading is required by clients with a view to supporting this in the new Gateway.

If there are a limited number of graphs we need to load, then we may not need to support lazy loading.

Queries using iQuery in gateway.init.py

• Parent-Child link (select by p.id and c.id)
• Child link + Child + Annotation link + Annotation (select by parent.id, ann.ns, ann.val)
• Child link + Parent + Child (select by child.id, p.id in pIds)
• Child link + Parent + Child (select by parent.id, c.id in cIds)
• Annotation links + Annotation (select by parent.id)
• Annotation (not linked to parent.id)
• Group (select by id in IDs)
• Experimenter (select by id in IDs)
• Experimenter (select by id NOT in IDs)
• Experimenter (select by name like 'start')
• Experimenter (where e.id : g.id exists in map of g.id) "listStaffs"
• Groups (select g.id in gIds)
• Screen + Details (owner & group) select by s.id
• Plate + Details (owner & group) + Screen links + Parent (select by plate.id)
• ...

Queries using iQuery in webclient_gateway.py

• Images + Details (creationEvent, owner, group) (select by IDs)
• Image & Details + Dataset (select by dsId)
• Project + Details
• Orphan Image / Dataset / Plate + Details
• Screen + Details
• Well (& Details) + Plate + Well-Sample + Image (select by Plate ID)
• Well (& Details) + Well-Sample + Image (& Details) (select by well ID)
• Well + Plate + Well-Sample + Image (select by plate name, well column, well row
• Tag (by text, description & ns=null)
• Experimenter (by omeName)
• Experimenter (by email)
• Experimenters (by IDs)
• Group (by name)
• Groups (name not user/system/guest)