Missing details in JSON export

I noticed a few pieces of information that are missing from the JSON export feature.

Missing properties:
* Lexile measure
* Private status - boolean for whether a book is private
* Inventory status - present/absent
* Cover ID - if set by the user

Properties that could be improved:

* Pagination. Currently represented as a string with semicolon-separated list of page counts, e.g. "xiv; 89 ". The numbering type is not included. Suggest a list of objects with page count and type: [{"pages": "xiv", "type": "i"}, {"pages": "123", "type": "1"}]

* Dimensions. Currently only the first set of dimensions (length/width/height) is included. Suggest a list of objects: [{"height": 1, "length": 2, "thickness": 3, "unit": "inch"}]

* Reading dates. Currently only the most recent start/end date are included. Suggest a list of objects: [{"datestarted": "2001-02-03", "dateread": "2004-05-06"}]

* From where. Currently there is no way to tell whether the property represents a venue name or free-text string. Suggest adding a property for venue id.

* Summary; physical description. Currently does not indicate whether the value was auto-generated or specified by the user. Suggest adding a boolean flag for each property to indicate this.

* Languages. The current representation is somewhat strange, for example the primary/secondary languages are included in the original language code field. There also appears to be some unpredictable behavior if there are duplicates among the three language fields. Additionally, there should be a way to distinguish auto-generated values from user-specified values. Suggest an object with a separate property for each field: {"primary": {"name": "English", "code": "eng"}, "secondary": {"name": "Chinese", "code": "chi"}, "original": {"name": "Latin", "code": "lat", "auto": true}}

* Authors. The order of secondary authors should match the book details, currently it does not.

(Edited to add cover id to the list of missing properties.)

(Edited again to add languages.)

(Edited to add authors + note about auto-generated languages.)

(2021-03-14 Edited to strike out private book status, this was a testing error due to https://www.librarything.com/topic/330575)

Is it going to be mentioned that the missing items here are also missing from most of the other exports as well?

Since Lexile is shared data, rather than catalog-level, it doesn't export. See the previous report at https://www.librarything.com/topic/295622#6635127

>3 lorax: To which the obvious rejoinder is that series and award data is also shared data, and yet it is exported.

What I would find useful is some sort of author id to indicate which ones it currently thinks are the same author. Yes, I know it would change from time to time, as work id does now.

>4 antqueen: Lexile data is both shared data, and at least semi-proprietary, unlike series and award data. It's the proprietary bit that probably keeps it out of the running on the export front; there would likely be fees associated with it somewhere.

>5 shadrach_anki: Interesting. I'll admit that the Lexile data in specific isn't something I know anything or even care much about; I was just addressing the shared data comment, which seems to pop up every time someone mentions adding things to the exports.

>2 gilroy: I haven't tested the other formats, but in my opinion JSON is the only one suitable for representing LibraryThing data in a structured, machine-readable way. Excel and TSV are row-based formats that can't represent complex nested data. MARC is a standardized format for cross-system compatibility, but that makes it inflexible; LibraryThing can't alter the standard to match its specific characteristics.

>3 lorax: Lexile score is not a shared value, it can be specified per book. I wonder if there's some legal issue that prevents its inclusion.