page.search_for return extra height

[ashifaliclientpoint]

Hello
I am using pymypdf(1.19.6) to search string from a pdf file. And doing redaction. But it returning extra height in compare of original pdf file. And also removing the string which overlapped by this string.
Please help me for this strange issue.
I am also attaching the original file and converted file.
DRAFT_Executive.pdf
highlighted_file.pdf

Reproduce step
import re
import fitz
import sys, json
file_path = "DRAFT_Executive.pdf"

pattern = r'[\s*([s|c|d|i|t]):([a-z]):([o|r])\s*]' # Replace with your desired regex pattern
doc = fitz.open(file_path)
resultOutput = []
tagsPerPage = {}
addedTags = set()

for page in doc:
text = page.get_text()
tagsPerPage[page.number]=[]
matches = re.finditer(pattern, text, re.IGNORECASE | re.MULTILINE | re.DOTALL)
if matches:
for match in matches:
start, end = match.span()
coordinates = page.search_for(match.group())
tempDict={}
firstCoordStr = ""
singleTagArr = []
needleStarted=0
for rect in coordinates:
x1, y2, x2, y1 = rect

            height = y2 - y1

            y1 = page.rect.height - y1
            y2 = page.rect.height - y2  


            currCoordsStr = str(y2)+"_"+str(y1)+"_"+str(x2)+"_"+str(x1)+"_"+str(page.number)
            currText = page.get_text("text",clip=rect)
            currTextTrimmed = ''.join(currText.split())  
            tagComplete = re.match(pattern, currTextTrimmed, re.IGNORECASE | re.MULTILINE | re.DOTALL)
            if(tagComplete):
                if currCoordsStr in addedTags:                        
                    pass     
                else:            
                    addedTags.add(currCoordsStr)
                    tagsPerPage[page.number].append({'y2':y2, 'y1':y1, 'tag':match.group(), 'x2':x2, 'x1':x1, 'page':(page.number+1) })
                    resultOutput.append({'y2':y2, 'y1':y1, 'tag':match.group(), 'x2':x2, 'x1':x1, 'page':(page.number+1) })
            else:
                if needleStarted == 0 and currTextTrimmed.find('[')!=-1:
                    needleStarted = 1  
                    tempDict = {}     
                    if "x2" in tempDict and x2 < tempDict["x2"]:
                        x2 = tempDict["x2"]
                    if "x1" in tempDict and x1 > tempDict["x1"]:
                        x1 = tempDict["x1"]  
                    tempDict={'y2':y2, 'y1':y1, 'tag':match.group(), 'x2':x2, 'x1':x1, 'page':(page.number+1) }                                                   
                elif needleStarted == 1 and currTextTrimmed.find(']')!=-1:
                    needleStarted = 0       
                    currCoordsStr = str(tempDict['y2'])+"_"+str(tempDict['y1'])+"_"+str(tempDict['x2'])+"_"+str(tempDict['x1'])+"_"+str(page.number)                        
                    if currCoordsStr in addedTags:     
                        tempDict = {}                   
                        pass     
                    else:            
                        if "x2" in tempDict and x2 < tempDict["x2"]:
                            x2 = tempDict["x2"]
                        if "x1" in tempDict and x1 > tempDict["x1"]:
                            x1 = tempDict["x1"]  
                        tempDict={'y2':y2, 'y1':y1, 'tag':match.group(), 'x2':x2, 'x1':x1, 'page':(page.number+1) }     
                        addedTags.add(currCoordsStr)
                        tagsPerPage[page.number].append(tempDict)
                        resultOutput.append(tempDict)
                        tempDict = {}
                else:
                    if "x2" in tempDict and x2 < tempDict["x2"]:
                        x2 = tempDict["x2"]
                    if "x1" in tempDict and x1 > tempDict["x1"]:
                        x1 = tempDict["x1"]  
                    tempDict={'y2':y2, 'y1':y1, 'tag':match.group(), 'x2':x2, 'x1':x1, 'page':(page.number+1) }             

if tagsPerPage:
for page in doc:
if tagsPerPage[page.number]:
for item in tagsPerPage[page.number]:
currPage= page.number+1
if item['page']==currPage:
y1 = page.rect.height - item['y1']
y2 = page.rect.height - item['y2']

                page.add_redact_annot(fitz.Rect(item['x1'], y2, item['x2'], y1),text_color=(0, 0, 0),cross_out=True)

            page.apply_redactions()

doc.save("highlighted_file.pdf", garbage=3, deflate=True)
doc.close()

Configuration
OS ubuntu
Python 3.8
PyMuPDF 1.19.6

[JorjMcKie]

This is a typical "Discussions" item. Let me first transfer it.

[JorjMcKie]

This is a common thing to happen!
Background:
Fonts have their own "natural" line height, which often is larger than the font size used. For example, the popular Helvetica font has a line height 37.4% larger than font size.
If the PDF creator inserted text lines with a closer distance than the natural height, extracted line bboxes will overlap of course.
This in turn will lead to redacting away unintended text (above and / or below current line), because redaction removes all characters with a bbox overlapping the redaction rectangle.

Try to avoid this by extracting / searching text that deliver bboxes with minimal (font size) heights, by setting fitz.TOOLS.set_small_glyph_heights(True) before extracting / searching.
This often helps.
Occasionally, PDF creators may have chosen to use even smaller line heights than the font size value. In these extreme cases you must use your own wits and create a bbox with a tiny height. For example 20% of the font size and located around the horizontal line through the middle of the original bbox, or so.

[ashifaliclientpoint]

@JorjMcKie I am getting text cutting issue in some files. Can you check the reason.

[JorjMcKie]

@JorjMcKie I am getting text cutting issue in some files. Can you check the reason.

This is not possible without a reproducing file. Please submit another issue and attach an example to it.