In array2.spad

\section{domain IARRAY2 IndexedTwoDimensionalArray}
<>=
)abbrev domain IARRAY2 IndexedTwoDimensionalArray
IndexedTwoDimensionalArray(R,mnRow,mnCol):Exports == Implementation where
  ++ An IndexedTwoDimensionalArray is a 2-dimensional array where
  ++ the minimal row and column indices are parameters of the type.
  ++ Rows and columns are returned as IndexedOneDimensionalArray's with
  ++ minimal indices matching those of the IndexedTwoDimensionalArray.
  ++ The index of the first row may be obtained by calling the
  ++ function minRowIndex.  The index of the first column may
  ++ be obtained by calling the function minColIndex.  The index of
  ++ the first element of a Row is the same as the index of the
  ++ first column in an array and vice versa.
  R : Type
  mnRow, mnCol : Integer
  Row ==> IndexedOneDimensionalArray(R,mnCol)
  Col ==> IndexedOneDimensionalArray(R,mnRow)
                                      ^^^^^^
  Exports ==> TwoDimensionalArrayCategory(R,Row,Col)
  Implementation ==>
    InnerIndexedTwoDimensionalArray(R,mnRow,mnCol,Row,Col)

In IndexedOneDimensionalArray?(*)
Col => min index of row (mnRow)
Row => min index of column (mnCol)
Is it a bug ?

No, this is correct. A one-dimensional row vector has columns, and the number of columns need to be specified as mnCol. Similarly, a one-dimensional column vector has rows, and the number of rows need to be specified as mnRow.

mnCol is minimal column index used, for example, in row creation after extraction.

Status: open => closed