Is there any method to remove or ignore the footer and header text?

[hemilparmar]

can we remove header and footer from pdf while extracting the text?

[JorjMcKie]

No, because PDF doesn't know what these things are. It's all just text, and you know better than anyone what to exclude eventually.

[JorjMcKie]

For example, if you know the position of headers / footers on a page, then build a sub-rectangle of the page rectangle and limit extraction to it.

# assume 'top' is the top coordinate of relevant text
# assume 'btm' is the bottom coordinate of relevant text
clip = page.rect
clip.y0 = top
clip.y1 = btm
# all text extraction variants accept the clip parameter:
text = page.get_text(clip=clip)

[hemilparmar]

As you said for parser , it is all text. so is there any way to know the header/footer location using fitz.?

[JorjMcKie]

For every text on the page, you can determine its position. The Page.get_text() method has several ways to get this information in varying levels of detail:

• get_text("blocks") - delivers a list of text blocks, togehter with the rectangle of the block
• get_text("words") - similar list, but on word level
• get_text("dict") - a dictionary with position info for each block, line or text "span"
• get_text("rawdict") - a similar dictionary, but with one more level down to each character

Please see the documentation for more details.

[shiu886]

There might be a separating line between the text and footer. Usually footer would be in a smaller font size.

[JorjMcKie]

@shiu886 - you are right: headers / footers often are written with other font properties than the regular text like smaller, or even a different font.