Colors module with custom tables

[NatalieRees]

Describe the bug

For the MESA Summer school we were trying to implement a custom colors table as one of the exercises.
This table has bolometric corrections for a range of Teff and logg at M_div_H=-0.5.
We followed the implementation as demonstrated by the custom_colors test_suite example however found the get_abs_mag_by_name routine was always returning -1d99 so seemed to be extrapolating when our values were definitely in the data range of the table.
I've realised the problem is that when you only have one value for M_div_H, even if the value you supply is exactly equal to it then it will be classed as extrapolation and return -1d99.
To check this I made a fudged data table with duplicated data at both M_div_H=-0.4 and M_div_H=-0.6 which then returns the expected values for M_div_H=-0.5.
I have attached the work directory including the fudged data table which works as well as the original table which doesn't

custom_colors_bug_report.zip
fehm050_Av03.HST_WFPC2.zip