Lets summarize : in Hakuneko we use pdfkit to turn a serie of pictures into a pdf.
I discovered that the problem occurs in the pdf.js library used; when parsing tExt chunks
manga-download/hakuneko#5455
Apparently Druzai fixes is correcting the problem, since the error doesnt occurs anymore if i patch accordingly.