Cantonese Level 1 - PDF to TSV extraction missing sentences

[awhlam]

I ran the PDF to TSV extraction for Cantonese.

Levels 2 and 3 extracted all 1,000 sentences fine, but level 1 was missing 3 sentences in the output TSV (I modified the Python script to not abort).

Below is an example of a sentence that was missed:

Glossika PDF

Output TSV Extraction:

Sentence #89 - has the Yale for sentence #90 instead Sentence #90 is missing

I think it might be an issue with the regular expressions? Let me know if you have any ideas or if you need more examples.

[awhlam]

The other two missing sentences have a similar issue as above:

Sentence #253 - Are you sitting on the floor? Sentence #255 - Are you feeling all right?

[emesterhazy]

Could you run pdftotext on the PDF and share a snippet that includes the missing sentences?

The command should be something like

pdftotext -layout -enc UTF-8 file_name.pdf output.txt

[awhlam]

@emesterhazy - Sure, here is the snippet of the pdftotext output that corresponds with the screenshot in my first post above.

• It looks like sentence #89 has the translations shifted down by one line.
• Sentences #90-91 look ok to me in this output.

[emesterhazy]

Closing due to archival of project. Best of luck!