keywords.csv on the main branch should be reviewed

[jabrah]

While going through the pages that Wax generates, I noticed that some of the data was very odd and traced it to some errors in the original keywords.csv file on the main branch. Sorry if these are already fixed elsewhere! If that's the case, then feel free to ignore.

d0075

• Need to place an extra comma between Jean Baptiste Raguet and "Judicial Acts,Civil Cases
• Duplicate keywords_item
• Need extra comma between keywords and 12
• Duplicate thumbnail: "/img/derivatives/iiif/images/d0075_d0075-01/full/250,/0/default.jpg"
• Not sure if no content in the names_white column was intended

d0111

• No keywords present in this row. Is this intended?

d0281

• Tab instead of comma Spanish\tLouisiana Historical Center should be Spanish,Louisiana Historical Center
• Extra comma between https://docs.google.com/document/d/1RKmmi9fY2d_NLF0iVag1trq5HRW5t9HdV-vcl1ug92c/edit and emancipation|free people of color|collaboration
• Shares the same Order number as d0072 (17)

Note that all of the image information (thumbnail, full, manifest) will be incorrect until the extra or missing commas is addressed, which will fix the row sizes for the CSV. Once that is done, we'll need to rerun the image derivative generation and the pages generation by first fully removing all of them and running the Wax tasks. Removal is important because Wax will likely skip items if files are already in place, even if they are incorrect.

[jabrah]

I would recommend trying to normalize the character used for delimit multivalue fields. I don't think this would break functionality, but could make the digital team's job a bit harder if characters are mixed

• keywords field must use | because of how the faceting and gallery is handled
• names-* shouldn't matter which delimiter is used, as the searching seems to be flexible
  • I can see both comma (,) and semicolon (;) used in different places, which can be confusing