read multiple files for link_daily

[bstabler]

need to be able to read multiple link files to sum up to daily

[bstabler]

link_daily_file_name: linksMD1.csv

is now:

link_daily_file_names:

• linksMD1.csv
• linksPM2.csv

link_daily_index_fields:

• i
• j

and if one file is specified then a join is not required, but if more than one is, then the join fields are required and the additional table column names have the filename as the column name suffix

[bstabler]

@toliwaga - I think maybe the test data for the ABM example is incorrect for the mce branch? Feel free to clean-up what I added as well before we merge all this into the master. Thanks.

[bstabler]

We need to specify a name to use for the file when we append the tables

link_daily_file_names:

• linksMD1.csv
• linksPM2.csv

becomes

link_daily_file_names: md1 : linksMD1.csv pm2 : linksPM2.csv

and link expressions go from:

links.svol + links.svol_linksPM2

to

either:

links.svol_md1 + links.svol_pm2

or

links.svol + links.svol_pm2 (only the 2nd and subsequent files get the appended file name)

[toliwaga]

We can allow both behaviors by doing this:

# this version has no suffix for linksMD1.csv
link_daily_file_names:
   "" : linksMD1.csv
    pm2 : linksPM2.csv

# this version assigns suffixes to both files
link_daily_file_names:
    md1 : linksMD1.csv
    pm2 : linksPM2.csv

but I think it might be clearer if the example used explicit suffixes for both

[bstabler]

I think we should use explicit suffixes for both

[toliwaga]

resolved by pull request 80