From 02dbbddb2b7a132e80ad9e1aeb11f57e77a2ff2d Mon Sep 17 00:00:00 2001
From: Shashi Gowda <shashi@juliacomputing.com>
Date: Wed, 20 Jun 2018 08:24:29 +0530
Subject: [PATCH] fix one-to-many map: fixes JuliaComputing/JuliaDB.jl#193

---
 src/PooledArrays.jl | 24 ++++++++++++++++++++++--
 test/runtests.jl    |  4 ++++
 2 files changed, 26 insertions(+), 2 deletions(-)

diff --git a/src/PooledArrays.jl b/src/PooledArrays.jl
index e41f15a..214b0ab 100644
--- a/src/PooledArrays.jl
+++ b/src/PooledArrays.jl
@@ -178,8 +178,28 @@ Base.find(pdv::PooledVector{Bool}) = find(convert(Vector{Bool}, pdv, false))
 function Base.map{T,R<:Integer}(f, x::PooledArray{T,R})
     ks = collect(keys(x.pool))
     vs = collect(values(x.pool))
-    newpool = Dict(zip(map(f, ks), vs))
-    PooledArray(RefArray(x.refs), newpool)
+    ks1 = map(f, ks)
+    uks = Set(ks1)
+    if length(uks) < length(ks1)
+        # this means some keys have repeated
+        newpool = Dict{eltype(ks), eltype(vs)}()
+        translate = Dict{eltype(vs), eltype(vs)}()
+        i = 1
+        for (k, k1) in zip(ks, ks1)
+            if haskey(newpool, k1)
+                translate[x.pool[k]] = newpool[k1]
+            else
+                newpool[k1] = i
+                translate[x.pool[k]] = i
+                i+=1
+            end
+        end
+        refarray = map(x->translate[x], x.refs)
+    else
+        newpool = Dict(zip(map(f, ks), vs))
+        refarray = copy(x.refs)
+    end
+    PooledArray(RefArray(refarray), newpool)
 end
 
 ##############################################################################
diff --git a/test/runtests.jl b/test/runtests.jl
index f03d271..5de5287 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -31,6 +31,10 @@ let a = rand(10), b = rand(10,10), c = rand(1:10, 1000)
     @test map(identity, pc) == pc
     @test map(x->2x, pc) == map(x->2x, c)
 
+    # case where the outputs are one-to-many
+    pa = PooledArray([1,2,3,4,5,6])
+    @test map(isodd, pa) == [1,0,1,0,1,0]
+
     px = PooledArray(rand(128))
     py = PooledArray(rand(200))