Conversion errors
Note: The document is from 2011
Errors addressed so far (May 2011):
- dårlege originalfiler - gjev ugyldig xml ** desse blir fanga opp i dag
- kodefeil - desse gjev gyldig xml, men meiningslause bokstavar ** utf-som-macroman ** utf-som-latin1 ** utf-som-html-hex ** utf-som-html-entitet
- skannefeil/ocr-feil - desse gjev meiningsfulle bokstavar, men meiningslaus tekst ** đ-som ó, osv.
- bad sentence-delimitation: one real sentence is one fragment in one language, 3 fragments in the other -> alignment goes bunk
- files
freecorpus/converted/sme/admin/others/**STM200420050011000SE_PDFS.pdf.xmlSTM200420050044000SE_PDFS.pdf.xmlhave encoding errors that đ is represented as and the document is full of ’s; thus these files should be deleted ** fileOTP200620070025000SE_PDFS.pdf.xmlhas paragraphs with content ‘——–’ so it should be deleted. ** fileSTM200320040010000SE_PDFA.pdf.xmlhas so many errors, it should be rescanned **uito-ohpenplana.txt.xmlthe original file is corrupted