How to get block wise context for the below image

[nattu22]

I trying to extract a text based on blocks. Currently, I am facing this issue to split this block in to two.

1st block in this image

bottom text needs to be in 2 blocks

Eg. For words, we can specify the spaces between words and line via x_tolerence and y_tolerence.

Is there any options available in fitz. if yes pls suggest something.

[JorjMcKie]

this is a d"Discussions" item, so will be transferred.

[JorjMcKie]

Establishing the hierarchy in TextPage objects "block -> line -> span -> char" fully happens inside MuPDF code and cannot be influenced.
These MuPDF heuristics are quite a complex algorithm which takes a lot of parameters into account ... in an overall effort to deliver good / acceptable results in most cases.

Obviously, this may not always be ideal. So it is left to your wits to find a more adequate separation into text components in situations specific to your data.
To do this, you have all required information - down to single characters if needed. There are the extraction variants "words", "blocks", "dict" and "rawdict" to choose from.

Sorry, but I see no basis for a meaningful enhancement.