Invalid annotation: singleton character tsd_trial.csv#L1126 and tsd_train.csv#L7895

[GillesJ]

I found two unconnected singleton character annotation which are invalid:

• tsd_trial.csv line 1126, instance 658:

"[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 21]","Ridiculous logic!

G&M sure seem hooked to Real Estate industry cash (propaganda pieces in exchange of ad cash), Trudeau and paying interest on massive Federal debt."

-> corrected label by removing singleton 21: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]

• tsd_train.csv line 7895, instance 4616:

"[94, 95, 96, 97, 98, 241]","He went on a 'traveling the country vacation' there. I hope they have a swift court and swift death penalty. He is immigrated here, non citizen living with parents, in Colorado. 

DO NOT give him back to us. No matter how much Hickenlooper pleads."

-> corrected label by removing singleton 241: [94, 95, 96, 97, 98]

I found these when unitizing annotations from character-level to token-level.
My script found no other singleton characters.

[avidale]

I found similar problems in several more lines.
E.g. line 1143 (668 if newlines in the texts are quoted) in tsd_trial.csv is "[5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20]",Because Trudeau knows that his federal Liberal Party survives on voters with brain damage from recreational pot use., where characters 5 and 6 are a part of the word "Because".
How to deal with such cases?