A method including: parsing a program to identify a plurality of features within the program; performing a first match of the plurality of features to a plurality of code files based on a hierarchical code index, the features corresponding to leaf nodes of the hierarchical code index and the code files corresponding to parent nodes of the leaf nodes; normalizing the first match results to weight against common features; performing a second match of the plurality of code files to one or more code repositories based on the hierarchical code index; normalizing the second match results to weight against common files; and identifying a code repository of the one or more code repositories as being included within the program.