You might look at PROC PLS. This procedure accommodates a CLASS statement (which would cover the binary and ordinal data), and continuous data. That is about all I can help with as it does not accommodate rotations, etc. that seem to be critical to factor analysis. If you are patient, PROC CALIS is available for confirmatory factor analysis, but I have never used it for that, so I don't know if the binary and ordinal variables would be handled the way you want them to be.
You might consider reposting this in the Statistical Procedures forum, as there may be someone over there who could help.
Steve Denham