correct graph coloring and variable liveness analysis

.gitignore

@@ -0,0 +1,3 @@
+a.s
+a.out
+.gitignore

README.md

@@ -1,18 +0,0 @@
-# TComp
-A practice of Essential of Complication in Julia
-
-## Dependencies
- - julia
- - [Match.jl](github.com/JuliaServices/Match.jl)
-
-## instruction
-`./src/TComp.jl [.tc file]`
-
-the output assembly code is `./a.c` in AT&T assembly langauge.
-
-to make it executable, please use `gcc`: `gcc ./a.c -o output.out`
-
-the example `.tc` files is in `./test`
-
-## Known issues
- - all connected variable are pathized (I don't understand the tricky method to elimitated the path number)

misc/vertexcoloring.jl

@@ -1,59 +0,0 @@
-graph = [['a', 'b'], ['b', 'c'], ['e', 'd'], ['e', 'a'], ['a', 'c'], ['b','e'], ['e','c']]
-
-
-function vertexColoring(graph)
-    notDefined = -1
-
-    function getColor(v, color)
-        if !(v in keys(color))
-            return -1
-        else
-            return color[v]
-        end
-    end
-
-    vertices = Set(vcat(graph...))
-    verticesList = collect(vertices)
-
-    verticesMapping = map(x -> [x, Set()], verticesList)
-
-    adjacentNodes = Dict(verticesMapping)
-
-    for link in graph
-        a = link[1]
-        b = link[2]
-        push!(adjacentNodes[a], b)
-        push!(adjacentNodes[b], a)
-    end
-
-    sort!(verticesList, by=x -> length(adjacentNodes[x]), rev=true)
-
-    color = Dict()
-
-    println(verticesList)
-
-
-    for i in verticesList
-        i_adjacents = adjacentNodes[i]
-        println(i_adjacents)
-        i_adjacents_color_set = Set(map(x -> getColor(x, color), collect(i_adjacents)))
-        i_adjacents_color_list = sort(collect(i_adjacents_color_set))
-
-        if i_adjacents_color_list == [notDefined]
-            color[i] = 0
-        else
-            tmpId = 0
-            for i in i_adjacents_color_list
-                if tmpId == i
-                    tmpId += 1
-                end
-            end
-            color[i] = tmpId
-        end
-    end
-
-    return color
-
-end
-
-println(vertexColoring(graph))