Transcriber would break if tags differ between files

[ewancook]

For example:

• Tag added midway through a series of files
• Float files contain data different to Tagname

etc.

This would break the fast, maths based transcription - would need to use a dict of tags again...

[ewancook]

Update:

Since (Float).DAT files contain TagIndex and not Tagname, individual (Tagname).DAT files for every float file would be required at some stage in parsing & checking, in order to prevent the above.

If all tag files were provided, float files could be grouped by the tags they contain, allowing tags to be selected for each group. The current fast parsing implementation could then be used.

This would require a GUI redesign and some refactoring.

[ewancook]

Note: since transcriber does not read tags during transcription, errors may occur if an incorrect tag file is used. These could be any of:

• If the tagfile has more tags than the float file. an error will occur on reaching the end of the float file
• If the tagfile has fewer tags than the float file, the transcribed csv file will be incomplete
• If the tagfile has a different number of tags than the float file, values will not be consistent within columns

[ewancook]

Comparing SubclassedDBF.header.numrecords to tag_selecter.presenter.active_tags would allow any (Float.DAT) files with an incorrect number of tags to be identified!