Hi Xypron, > currently only one data file is allowed in GLPK. This is an unnecessary > restriction.
> One use case for multiple data files is keeping master data (e.g. > machines, products, ...) separated from transactional data (orders, > demands) and from parameters, which comes in handy when calculating > different scenarios for the same model. Another use case arises when > working with SQL. It would be wise to keep database connection > information containing passwords separated from other data. > When I analyzed the code I found that the necessary changes in > glpmpl04.c and glplpx17.c are minor: > http://glpk.dyndns.org/viewvc/svn/glpk/glpk/branches/glpk-4.31-multiple_data_files/src/glpmpl04.c?view=diff&r1=286&r2=285&diff_format=h > http://glpk.dyndns.org/viewvc/svn/glpk/glpk/branches/glpk-4.31-multiple_data_files/src/glplpx17.c?view=diff&r1=286&r2=285&diff_format=h Thank you. Please keep the changes for a while. Currently I need to rewrite glpsol to get rid of using obsolete api routines. BTW, why not to use any alternative way? For example, one could write data; #include "data1" #include "data2" end; and then use the C preprocessor 'gcc -E' to build the data file. _______________________________________________ Help-glpk mailing list [email protected] http://lists.gnu.org/mailman/listinfo/help-glpk
