Alignment changes in some of the pdf file after replace text and replaced text is hidden

[jayakumar0612]

Hi,
I'm trying to find and replace text using below code

doc = fitz.open(input_file_path)
        dict.pop("source_dict")
        dict.pop('irrelevant_fields')
        origin = ""
         fsize = ""
        font = ""



    for field in dict:
        for value in dict[field]:
            for page in doc:
                search_text_list = [" "+value+" ", " "+value+",", " "+value+"."]
                for search_text in search_text_list:
                    rl = page.search_for(search_text)
                    replacement_text = "{{" + str(field) + "}}"
                    #replacement_text = "*" * len(field)
                    for rect in rl:
                        for b in page.get_text("dict", clip=rect)["blocks"]:
                            for l in b["lines"]:
                                for span in l["spans"]:
                                    fsize = span["size"]
                                    origin = fitz.Point(span["origin"])
                                    flags = span["flags"]
                                    try:
                                        font = page.get_text("dict")["blocks"][0]["lines"][0]['spans'][0]['font']
                                    except Exception:
                                        if flags & 2 ** 3:
                                            font = "cour"
                                        else:
                                            font = "helv"
                                    if not font:
                                        if flags & 2 ** 3:
                                            font = "cour"
                                        else:
                                            font = "helv"
                        page.add_redact_annot(rect)
                        page.apply_redactions()

                        # tl = fitz.get_text_length(search_text, fontname=font, fontsize=fsize)
                        # fsize = fsize * rect.width / tl
                        if len(search_text) < len(replacement_text):
                            fsize = fsize / 4
                        if origin and font and fsize:
                            page.insert_text(origin, replacement_text, fontname=font, fontsize=fsize, overlay=False)

    path = os.path.split(input_file_path)[1]
    if os.path.exists(path):
        os.remove(path)
    doc.save(path,garbage=3, deflate=True)
    doc.close()

Some files it is working fine but for some files the replaced text is not visible when opening the file after replace. and when we search for a term it highlights some ir-releant term in pdf viewer software. i have attached source and replaced file for

[JorjMcKie]

I did not try out your files and the code yet, but this sounds like one of the typical issues where pages contain "unprotected" addressing scheme changes - see the docs here.
So a solution may be to immediately after a page has been loaded, do page.clean_contents(). This will clean the page's appearance building instructions and thus take care of correctly positioning new insertions of text and other stuff.

[jayakumar0612]

I have changed overlay=True but the inserted text shows something like this

[JorjMcKie]

I have changed overlay=True but the inserted text shows something like this

instead of what?

[jayakumar0612]

instead of the inserted word "ORGANIZATION" the above code is shown

[JorjMcKie]

Hm, I don't have your full script. So please tell me which word on which page you want to replace with "ORGANIZATION" and will try to reproduce the behavior.

[jayakumar0612]

I'm trying to replace word "Gratuity Act" in below image with "ORGANIZATION"

the below image shows after replacement with overlay=False here "ORGANIZATION" is hidden

below image shows after replacement with overlay=True here "ORGANIZATION" is shown as code.

I have attached two pdf in my first post. one is original file and another one is after replacement with overlay=False

I think The below code snippet should be helpful to reproduce the issue

def replace_text():
  import fitz
  import os

  input_file_path = "replace path to pdf file here"
  doc = fitz.open(input_file_path)
  origin = ""
  fsize = ""
  font = ""
  value = "Gratuity Act"
  field = "ORGANIZATION"
  

  for page in doc:
      search_text_list = [" " + value + " ", " " + value + ",", " " + value + "."]
      for search_text in search_text_list:
          rl = page.search_for(search_text)
          replacement_text = "{{" + str(field) + "}}"
          # replacement_text = "*" * len(field)
          for rect in rl:
              for b in page.get_text("dict", clip=rect)["blocks"]:
                  for l in b["lines"]:
                      for span in l["spans"]:
                          fsize = span["size"]
                          origin = fitz.Point(span["origin"])
                          flags = span["flags"]
                          try:
                              font = page.get_text("dict")["blocks"][0]["lines"][0]['spans'][0]['font']
                          except Exception:
                              if flags & 2 ** 3:
                                  font = "cour"
                              else:
                                  font = "helv"
                          if not font:
                              if flags & 2 ** 3:
                                  font = "cour"
                              else:
                                  font = "helv"
              page.add_redact_annot(rect)
              page.apply_redactions()

              # tl = fitz.get_text_length(search_text, fontname=font, fontsize=fsize)
              # fsize = fsize * rect.width / tl
              if len(search_text) < len(replacement_text):
                  fsize = fsize / 4
              if origin and font and fsize:
                  page.insert_text(origin, replacement_text, fontname=font, fontsize=fsize, overlay=False)

  path = os.path.split(input_file_path)[1]
  if os.path.exists(path):
      os.remove(path)
  doc.save(path, garbage=3, deflate=True)
  doc.close()

[JorjMcKie]

Something must be wrong with your script. Here is what I did:

page=doc[0]
rl=page.search_for(stext)

spans=[]
for r in rl:
    for b in page.get_text("dict",clip=r)["blocks"]:
        for l in b["lines"]:
            for s in l["spans"]:
                spans.append(s)

page.apply_redactions()

for s in spans:
    bbox=s["bbox"]
    fsize=s["size"]
    origin=s["origin"]
    page.insert_text(origin, text,fontsize=fsize)

doc.save("x.pdf")

And it does work!
The following also work, but the text is fairly small because it had to be shrinked to fit the old place.

page=doc[0]
page.clean_contents()
for r in rl:
    page.add_redact_annot(r, text)

page.apply_redactions()

doc.save("y.pdf")

In neither case appears your funny text.

[jayakumar0612]

Thanks for your support @JorjMcKie the above code snippet works fine for me