getting text from detect_document_text

How can I get the text in natural reading order (left to right) with detect_document_text with line break info?

Example image:

document.text output:
```
quick a brown fox
jumps over the lazy dog
word3
word4
word5 word7
word1 word2
word8word9 word10
```

document.lines output:
`[quick a brown fox, jumps over the lazy dog, word3, word4, word5 word7, word1 word2, word8word9 word10]
`

document.words output:
`[a, brown, the, fox, over, jumps, dog, quick, lazy, word7, word3, word4, word2, word5, word1, word8word9, word10]
`