Avoid using temporary vars for sparse data arrays

Just use the accessor functions rowvals, nonzeros, getcolptr whenever
needed instead. Especially in the sparse vector case avoid using findnz,
which creates a copy of the data. Use nonzeroinds and nonzeros instead.

src/sparsematrix.jl

@@ -3908,12 +3908,10 @@ function vcat(X::AbstractSparseMatrixCSC...)
         ptr_res = colptr[c]
         for i = 1 : num
             colptrXi = getcolptr(X[i])
-            rowvalXi = rowvals(X[i])
-            nzvalXi = nonzeros(X[i])
             col_length = colptrXi[c + 1] - colptrXi[c]
             ptr_Xi = colptrXi[c]
 
-            stuffcol!(rowval, nzval, ptr_res, rowvalXi, nzvalXi, ptr_Xi, col_length, mX_sofar)
+            stuffcol!(rowval, nzval, ptr_res, rowvals(X[i]), nonzeros(X[i]), ptr_Xi, col_length, mX_sofar)
 
             ptr_res += col_length
             mX_sofar += mX[i]
@@ -3974,23 +3972,19 @@ end
 # Efficient repetition of sparse matrices
 
 function Base.repeat(A::AbstractSparseMatrixCSC, m)
-    colptr_source = getcolptr(A)
-    rowval_source = rowvals(A)
-    nzval_source = nonzeros(A)
-
     nnz_new = nnz(A) * m
-    colptr = similar(colptr_source, length(colptr_source))
-    rowval = similar(rowval_source, nnz_new)
-    nzval = similar(nzval_source, nnz_new)
+    colptr = similar(getcolptr(A), length(getcolptr(A)))
+    rowval = similar(rowvals(A), nnz_new)
+    nzval = similar(nonzeros(A), nnz_new)
 
     colptr[1] = 1
     for c = 1 : size(A, 2)
         ptr_res = colptr[c]
-        ptr_source = colptr_source[c]
-        col_length = colptr_source[c + 1] - ptr_source
+        ptr_source = getcolptr(A)[c]
+        col_length = getcolptr(A)[c + 1] - ptr_source
         for index_repetition = 0 : (m - 1)
             row_offset = index_repetition * size(A, 1)
-            stuffcol!(rowval, nzval, ptr_res, rowval_source, nzval_source, ptr_source,
+            stuffcol!(rowval, nzval, ptr_res, rowvals(A), nonzeros(A), ptr_source,
                       col_length, row_offset)
             ptr_res += col_length
         end

src/sparsevector.jl

@@ -1314,14 +1314,13 @@ function Base.repeat(v::AbstractSparseVector, m)
     nnz_source = nnz(v)
     nnz_new = nnz_source * m
 
-    nzind_source, nzval_source = findnz(v)
-    nzind = similar(nzind_source, nnz_new)
-    nzval = similar(nzval_source, nnz_new)
+    nzind = similar(nonzeroinds(v), nnz_new)
+    nzval = similar(nonzeros(v), nnz_new)
 
     ptr_res = 1
     for index_repetition = 0:(m-1)
         row_offset = index_repetition * length(v)
-        stuffcol!(nzind, nzval, ptr_res, nzind_source, nzval_source, 1, nnz_source, row_offset)
+        stuffcol!(nzind, nzval, ptr_res, nonzeroinds(v), nonzeros(v), 1, nnz_source, row_offset)
         ptr_res += nnz_source
     end
     @assert ptr_res == nnz_new + 1
@@ -1331,11 +1330,9 @@ end
 
 function Base.repeat(v::AbstractSparseVector, m, n)
     w = repeat(v, m)
-    nzind_source, nzval_source = findnz(w)
-
-    colptr = Vector{eltype(nzind_source)}(1 .+ nnz(w) * (0:n))
-    rowval = repeat(nzind_source, n)
-    nzval = repeat(nzval_source, n)
+    colptr = Vector{eltype(nonzeroinds(w))}(1 .+ nnz(w) * (0:n))
+    rowval = repeat(nonzeroinds(w), n)
+    nzval = repeat(nonzeros(w), n)
     SparseMatrixCSC(length(w), n, colptr, rowval, nzval)
 end