write new mseed that use record_reader for fast and simple data

New in the new mseed to the outside:

• gap_minimum parameter: to collect only gaps/overlaps that greater or equal to gap_minimum
• is_multiplex: if False, can skip the file by checking if the first record meet the requirement, which save a log of time for analyzing non-multiplex data
• on_unittest: skip threading, database code to be easier when writing unittests
• convert data at model level and won't be converted in plotting so we aren't afraid it might be converted twice for case of mass position or zooming.
• remove data flags to avoid mistake: fullData, needConvert, needProcess because data will be calculated once and data are converted only one time at model level.

NOTICE:

• general_data/general_data.py is changed from data_type_model.py
• mseed_data/mseed.py is changed from mseed/mseed.py
• In mseed_data and general_data, the functions copied from existing functions that won't need to be reviewed have doc strings start with 'FROM'; the functions modified from existing functions will have doc strings start with 'CHANGED FROM'.