Hierarchical export/import to server, single json per item. Moving and deletion allowed

[rber474]

• Adds hierarchycal content export/import. A folder structure will be created and a json file per item.
  This will allow to keep track of changes for each item. Also allow to move o delete content.

• Some fixes for spanish translations

• Modifies import_content and export_content templates to include boostrap classes and change checkboxes to switches.

Export Content

A new checkbox has been added to allow hierarchical export to server. Will generate a tree export with one json per item. For folders content a subdirectory will be added containing children.
Added bootstrap css classes and redesigned checkbox into switches.

The generated exported tree will contain a subdirectory for the exported site with two main directories: content and removed_items

exported_tree/
└── Plone
├── content
│ ├── 11_registration.json
│ ├── 12_schedule.json
│ ├── 19_sprint.json
│ ├── 1_contact.json
│ ├── 20_test1.json
│ ├── 2_events.json
│ ├── 6_intranet.json
│ ├── 7_location-1.json
│ ├── 8_news.json
│ ├── contact
│ ├── events
│ │ ├── 3_deadline-for-talk-submission.json
│ │ ├── 5_second-event-in-events-folder.json
│ │ ├── deadline-for-talk-submission
│ │ │ └── 4_00b2dcca-4e95-4feb-bf83-3789000a017f_1_105_c.jpeg.json
│ │ └── second-event-in-events-folder
│ ├── intranet
│ │ ├── 13_sponsors.json
│ │ ├── sponsors
│ │ ├── 15_training.json
│ │ ├── training
│ │ ├── 16_00b2dcca-4e95-4feb-bf83-3789000a017f_1_105_c.jpeg.json
│ │ ├── 17_document-third-level-inside-training.json
│ │ └── document-third-level-inside-training
│ │ └── 18_00b2dcca-4e95-4feb-bf83-3789000a017f_1_105_c.jpeg.json
│ ├── location-1
│ ├── news
│ │ ├── 10_submit-your-talks.json
│ │ ├── 9_conference-website-online.json
│ │ ├── conference-website-online
│ │ └── submit-your-talks
│ ├── registration
│ ├── schedule
│ ├── sprint
│ └── test1
└── removed_items
└── 14_00b2dcca-4e95-4feb-bf83-3789000a017f_1_105_c.jpeg.json

Import Content

Redesigned form with bootstrap css classes, added fieldsets to separate type of imports.
A new fieldset added for the hierarchical import where you can select the server path to the exported_tree

Moving content

Non folderish item can be moved by just changing the location of the json file.
Folderish object can be moved by changing the location of both json file and subdirectory.

Reordering

Exported json files names include the ordinal number of the original order. At the moment, no further reordering system has been implemented.

Marking content as deleted

Moving json files to removed_itemswill mark them as deleted. In an update import, where site has content, that item will be removed.
Disclaimer: this will be remove the item without taking care of link integrity

[rber474]

@pbauer @fredvd @davisagli
as commented in #223, I have tried to implement a solution for the main requirements made by @fredvd.
Not sure if it covers all of them, but please give a glance and let me know.

[sneridagh]

As I know little about exportimport inner parts, I won't issue an official approval.
However, LGTM.

In @kitconcept/contentcreator we used the approach of using a dotted file, but the folder tree approach seems to me like an improvement.

Question: Which id takes precedence, the one in the filename or the one in the id attribute in the JSON file? (in case of discordance).

I'd love to see this merged soon! Thanks for the effort!

PS: @pbauer how about the PoC that you showed in the PC lightning talk? I guess this supersedes it, right?

[rber474]

@sneridagh the filename is ignored and the process does not modify any file content. So the answer should be both the id and the @id attributes in the file.

The filename is composed by the order number and the object id. The latter with the intention of make it easier reading the files in the exported tree.

Once the importer reads the json file it will modify the parent @id and parent UID extracted from the new position (after a object search) just in memory.

The original import process will update the @id accordanly.

I hope I have explained myself...

            if json_file.parent == Path(os.path.join(self.path, parents)):
                yield item
            else:
                parent_obj = api.content.get(path=parent_path)
                if not parent_obj:
                    # if parent_path is "." or parent_obj doesn't yet exist
                    parent_obj = portal

                # Modify parent data into json to be yield
                # local files won't be modified
                if parent_obj:
                    self.delete_old_if_moved(item.get("UID"))
                    item["@id"] = item.get("@id")
                    json_parent.update(
                        {"@id": parent_obj.absolute_url(), "UID": parent_obj.UID()}
                    )
                    item["parent"] = json_parent
                yield item