How is pos calculated for TextWriter? The place seems not correct.

[nissansz]

How is pos calculated for TextWriter? The place seems not correct.

        rect = fitz.Rect(left, top, right, bottom)
        tw.append((left, **bottom**), text,  fontsize=fontsize_fit, font=font, small_caps=True)

tw.append((left, top), text, fontsize=fontsize_fit, font=font, small_caps=True)

tw.append((left, top+fontsize), text, fontsize=fontsize_fit, font=font, small_caps=True)

[nissansz]

below setting seems a little better, but don't know whether it is the best option

fontsize = int((bottom - top)*0.8)
tw.append((left, top+fontsize), text, fontsize=fontsize_fit, font=font, small_caps=True)

[JorjMcKie]

You should try to use the origin of the span that is delivered on text extraction. This value depends on the font. Look at this picture in the documentation.

[nissansz]

How to set origin?

a = span["ascender"]
d = span["descender"]
r = fitz.Rect(span["bbox"])
o = fitz.Point(span["origin"]) # its y-value is the baseline
r.y1 = o.y - span["size"] * d / (a - d)
r.y0 = r.y1 - span["size"]

r now is a rectangle of height 'fontsize'

[nissansz]

I still don't know how to set this origin to TextWriter.

tw.append((int(left), int(top+fontsize)), text, fontsize = fontsize, font=font, small_caps=True)

Change to below?

origin = (bbox.x0, bbox.y1 + font.descender * bbox.height)
tw.append(origin , text, fontsize = fontsize, font=font, small_caps=True)

[JorjMcKie]

origin = (bbox.x0, bbox.y1 + font.descender * bbox.height)

yes, that's my best idea.
If you have the right font.

[nissansz]

result seems wrong.

    origin = (int(left), int(top + font.descender * (bottom-top)))
        tw.append(origin, text,  fontsize = fontsize, font=font, small_caps=True)

[JorjMcKie]

yes, your original calculation is the better approximation - stick with it then.

[nissansz]

This seems better
fontsize = int((bottom - top)0.85)
origin = (int(left), int(bottom+font.descender(bottom-top)))