SheetXL - v1.0.2
    Preparing search index...

    Interface CacheJSON

    A pivot cache as the workbook carries it, by the id the tables name.

    A cache over a worksheet source is modeled: its fields, their items and the records, which are what a refresh rebuilds and what a pivot is computed from. part, the definition and records as they were read, is present only while the cache is unchanged, and is then what is written. A cache with no fields is carried as its part alone.

    interface CacheJSON {
        cacheId: number;
        fields?: CacheFieldJSON[];
        items?: GridTuple<JSON>;
        part?: IEmbeddedPart.JSON;
        recordCount?: number;
        records?: GridTuple<number>;
        refreshOnLoad?: boolean;
        rest?: string;
        source?: CacheSourceJSON;
    }
    Index
    cacheId: number
    fields?: CacheFieldJSON[]
    items?: GridTuple<JSON>

    Each field's items: the field's index, then its items in order. A string is literal; a missing item is { t: 'z', v: 0 }; a date is its serial number (see CacheFieldJSON.dates and CacheFieldJSON.dateItems).

    The parts as read, written as they are while the cache is unchanged.

    recordCount?: number
    records?: GridTuple<number>

    Each field's records: the field's index, then, per record in order, the index of its item in items. A field with CacheFieldJSON.noRecords has none. Absent when the file saved no records: its tables are not laid out until a refresh from the source.

    refreshOnLoad?: boolean
    rest?: string

    The definition's markup with what is modeled taken out (the fields, the source's cells, the record count and refresh flag): what the cache holds that is not typed yet.

    The cells the records were read from: a sheet and a range, or a table's name. Present for a cache carried as its part too, when it names cells in this workbook.