Highlight text content in pdf files using python and save a screenshot - python

I have a list of pdf files and I need to highlight specific text on each page of these files and save a snapshot for each of the text instances.
So far I am able to highlight the text and save the entire page of a pdf file as a snapshot. But, I want to find the position of highlighted text and take a zoomed in the snapshot which will be more detailed compared to the full page snapshot.
I'm pretty sure there must be a solution to this problem. I am new to Python and hence I am not able to find it. I would be really grateful if someone can help me out with this.
I have tried using PyPDF2, Pymupdf libraries but I couldn't figure out the solution. I also tried highlighting by providing coordinates which works but couldn't find a way to get these coordinates as output.
[![Sample snapshot from the code[![\]\[1\]][1]][1]][1]
#import PyPDF2
import os
import fitz
from wand.image import Image
import csv
#import re
#from pdf2image import convert_from_path
check = r'C:\Users\Pradyumna.M\Desktop\Pradyumna\Automation\Intel Bytes\Create Source Docs\Sample Check 8 Apr 2019'
dir1 = check + '\\Source Docs\\'
dir2 = check + '\\Output\\'
dir = [dir1, dir2]
for x in dir:
try:
os.mkdir(x)
except FileExistsError:
print("Directory ", x, " already exists")
### READ PDF FILE
with open('upload1.csv', newline='') as myfile:
reader = csv.reader(myfile)
for row in reader:
rowarray = '; '.join(row)
src = rowarray.split("; ")
file = check + '\\' + src[4] + '.pdf'
print(file)
#pdfFileObj = open(file,'rb')
#pdfReader = PyPDF2.PdfFileReader(pdfFileObj)
#print("Total number of pages: " + str(pdfReader.numPages))
doc = fitz.open(file)
print(src[5])
for i in range(int(src[5])-1, int(src[5])):
i = int(i)
page = doc[i]
print("Processing page: " + str(i))
text = src[3]
#SEARCH TEXT
print("Searching: " + text)
text_instances = page.searchFor(text)
for inst in text_instances:
highlight = page.addHighlightAnnot(inst)
file1 = check + '\\Output\\' + src[4] + '_output.pdf'
print(file1)
doc.save(file1, garbage=4, deflate=True, clean=True)
### Screenshot
with(Image(filename=file1, resolution=150)) as source:
images = source.sequence
newfilename = check + "\\Source Docs\\" + src[0] + '.jpeg'
Image(images[i]).save(filename=newfilename)
print("Screenshot of " + src[0] + " saved")

"couldn't find a way to get these coordinates as output"
- you can get the coordinates out by doing this:
for inst in text_instances:
print(inst)
inst are fitz.Rect objects which contain the top left and bottom right coordinates of the piece of text that was found. All the information is available in the docs.
I managed to highlight points and also save a cropped region using the following snippet of code. I am using python 3.7.1 and my output for fitz.version is ('1.14.13', '1.14.0', '20190407064320').
import fitz
doc = fitz.open("foo.pdf")
inst_counter = 0
for pi in range(doc.pageCount):
page = doc[pi]
text = "hello"
text_instances = page.searchFor(text)
five_percent_height = (page.rect.br.y - page.rect.tl.y)*0.05
for inst in text_instances:
inst_counter += 1
highlight = page.addHighlightAnnot(inst)
# define a suitable cropping box which spans the whole page
# and adds padding around the highlighted text
tl_pt = fitz.Point(page.rect.tl.x, max(page.rect.tl.y, inst.tl.y - five_percent_height))
br_pt = fitz.Point(page.rect.br.x, min(page.rect.br.y, inst.br.y + five_percent_height))
hl_clip = fitz.Rect(tl_pt, br_pt)
zoom_mat = fitz.Matrix(2, 2)
pix = page.getPixmap(matrix=zoom_mat, clip = hl_clip)
pix.writePNG(f"pg{pi}-hl{inst_counter}.png")
doc.close()
I tested this on a sample pdf that i peppered with "hello":
Some of the outputs from the script:
I composed the solution out of the following pages of the documentation:
Tutorial page to get introduced into the library
page.searchFor to figure out the return type of the searchFor method
fitz.Rect to understand what the returned objects from page.searchFor are
Collection of Recipes page (called faq in the URL) to figure out how to crop and save part of a pdf page

Related

Python: how do I change page size of a Word document and then print it into PDF?

I have a bunch of Word documents that I need to reset to the Letter page size (they are all currently 11"x17"). Is there a Python library that I can use to:
Load a Word doc
Set its page size to Letter
Print it into PDF
Seems like docx2pdf can be used to do 1 and 3 but what about 2? And if it can do it, how? Other options?
TIA!
Here's what worked for me at the end:
from docx import Document
from docx.shared import Inches
import os
from docx2pdf import convert
for root, dirs, files in os.walk(r"c:\rootdir"):
for file in files:
if file.endswith('.docx'):
inDocxFN = os.path.join(root, file)
base = os.path.splitext(inDocxFN)[0]
outDocxFN = base + '.2.docx'
outPdfFN = base + '.pdf'
print("%s TO %s" % (inDocxFN,outPdfFN))
document = Document(inDocxFN)
section = document.sections[0]
section.page_width = 7772400
section.page_height = 10058400
# don't want to modify the originals, so create a copy and then generate PDF from it, then delete the copy
#
document.save(outDocxFN)
try:
convert(outDocxFN,outPdfFN)
except:
print("Failed converting %s" % outDocxFN)
try:
os.remove(outDocxFN)
except:
print("Failed deleting %s" % outDocxFN)

picture in header hide first picture in docx with python-docx

In a Document of docx in python,
when an image is added to the header, the first next picture does not appear.
The image is visible with LibreOffice (7.0) but not with MS Office(365). And MS Office asks to repair the file after a modification in the file.
from docx import Document as DocumentDocx # for creating docx files
from docx.shared import Cm # section parameters
file_list = ['i1', 'i2', 'i3', 'i4']
document = DocumentDocx()
section = document.sections[0]
header = section.header
p = header.add_paragraph('')
r = p.add_run()
r.add_picture('logo.png', height=Cm(1))
p = document.add_paragraph()
p.alignment = 1
run = p.add_run()
run.add_text('1')
for ii, name_file in enumerate(file_list):
run.add_text(str(ii))
run.add_picture(name_file + '.png', width=Cm(12))
document.save('file' + '.docx')
How to correctly add a picture in header and show all pictures ?
The missing image is due to Word (version greater than 2105), and only if there is image in header section of the existing word document
as eamars say it.

How to make my Tesseract-OCR conversion code run faster

I have a conversion script, which converts pdf files and image files to text files. But it takes forever to run my script. It took me almost 48 hours to finished 2000 pdf documents. Right now, I have a pool of documents (around 12000+) that I need to convert. Based on my previous rate, I can't imagine how long will it take to finish the conversion using my code. I am wondering is there anything I can do/change with my code to make it run faster?
Here is the code that I used.
def tesseractOCR_pdf(pdf):
filePath = pdf
pages = convert_from_path(filePath, 500)
# Counter to store images of each page of PDF to image
image_counter = 1
# Iterate through all the pages stored above
for page in pages:
# Declaring filename for each page of PDF as JPG
# For each page, filename will be:
# PDF page 1 -> page_1.jpg
# PDF page 2 -> page_2.jpg
# PDF page 3 -> page_3.jpg
# ....
# PDF page n -> page_n.jpg
filename = "page_"+str(image_counter)+".jpg"
# Save the image of the page in system
page.save(filename, 'JPEG')
# Increment the counter to update filename
image_counter = image_counter + 1
# Variable to get count of total number of pages
filelimit = image_counter-1
# Create an empty string for stroing purposes
text = ""
# Iterate from 1 to total number of pages
for i in range(1, filelimit + 1):
# Set filename to recognize text from
# Again, these files will be:
# page_1.jpg
# page_2.jpg
# ....
# page_n.jpg
filename = "page_"+str(i)+".jpg"
# Recognize the text as string in image using pytesserct
text += str(((pytesseract.image_to_string(Image.open(filename)))))
text = text.replace('-\n', '')
#Delete all the jpg files that created from above
for i in glob.glob("*.jpg"):
os.remove(i)
return text
def tesseractOCR_img(img):
filePath = img
text = str(pytesseract.image_to_string(filePath,lang='eng',config='--psm 6'))
text = text.replace('-\n', '')
return text
def Tesseract_ALL(docDir, txtDir, troubleDir):
if docDir == "": docDir = os.getcwd() + "\\" #if no docDir passed in
for doc in os.listdir(docDir): #iterate through docs in doc directory
try:
fileExtension = doc.split(".")[-1]
if fileExtension == "pdf":
pdfFilename = docDir + doc
text = tesseractOCR_pdf(pdfFilename) #get string of text content of pdf
textFilename = txtDir + doc + ".txt"
textFile = open(textFilename, "w") #make text file
textFile.write(text) #write text to text file
else:
# elif (fileExtension == "tif") | (fileExtension == "tiff") | (fileExtension == "jpg"):
imgFilename = docDir + doc
text = tesseractOCR_img(imgFilename) #get string of text content of img
textFilename = txtDir + doc + ".txt"
textFile = open(textFilename, "w") #make text file
textFile.write(text) #write text to text file
except:
print("Error in file: "+ str(doc))
shutil.move(os.path.join(docDir, doc), troubleDir)
for filename in os.listdir(txtDir):
fileExtension = filename.split(".")[-2]
if fileExtension == "pdf":
os.rename(txtDir + filename, txtDir + filename.replace('.pdf', ''))
elif fileExtension == "tif":
os.rename(txtDir + filename, txtDir + filename.replace('.tif', ''))
elif fileExtension == "tiff":
os.rename(txtDir + filename, txtDir + filename.replace('.tiff', ''))
elif fileExtension == "jpg":
os.rename(txtDir + filename, txtDir + filename.replace('.jpg', ''))
docDir = "/drive/codingstark/Project/pdf/"
txtDir = "/drive/codingstark/Project/txt/"
troubleDir = "/drive/codingstark/Project/trouble_pdf/"
Tesseract_ALL(docDir, txtDir, troubleDir)
Does anyone know how can I edit my code to make it run faster?
I think a process pool would be perfect for your case.
First you need to figure out parts of your code that can run independent of each other, than you wrap it into a function.
Here is an example
from concurrent.futures import ProcessPoolExecutor
def do_some_OCR(filename):
pass
with ProcessPoolExecutor() as executor:
for file in range(file_list):
_ = executor.submit(do_some_OCR, file)
The code above will open a new process for each file and start processing things in parallel.
You can find the oficinal documentation here: https://docs.python.org/3/library/concurrent.futures.html#concurrent.futures.ProcessPoolExecutor
There is also an really awesome video that shows step-by-step how to use processes for exactly this: https://www.youtube.com/watch?v=fKl2JW_qrso
Here is a compact version of the function removing the file write stuff. I think this should work based on what I was reading on the APIs but I haven't tested this.
Note that I changed from string to list because adding to a list is MUCH less costly than appending to a string (See this about join vs concatenation
How slow is Python's string concatenation vs. str.join?) TLDR is that string concat makes a new string every time you are concatenating so with large strings you start having to copy many times.
Also, when you were calling replace each iteration on the string after concatenation, it was doing again creating a new string. So I moved that to operate on each string that is generated. Note that if for some reason that string '-\n' is an artifact that occured due to the concatenation previously, then it should be removed from where it is and placed here: return ''.join(pageText).replace('-\n','') but realize putting it there will be creating a new string with the join, then creating a whole new string from the replace.
def tesseractOCR_pdf(pdf):
pages = convert_from_path(pdf, 500)
# Counter to store images of each page of PDF to image
# Create an empty list for storing purposes
pageText = []
# Iterate through all the pages stored above will be a PIL Image
for page in pages:
# Recognize the text as string in image using pytesserct
# Add the text to a list while removing the -\n characters.
pageText.append(str(pytesseract.image_to_string(page)).replace('-\n',''))
return ''.join(pageText)
An even more compact one-liner version
def tesseractOCR_pdf(pdf):
#This takes each page of the pdf, extracts the text, removing -\n and combines the text.
return ''.join([str(pytesseract.image_to_string(page)).replace('-\n', '') for page in convert_from_path(pdf, 500)])

How to split a PDF with PyMuPDF (with a loop)?

I'd like to use PyMuPDF : I'd like to split a pdf, with for each splitted file, a file named with the name of the bookmark, with only page
I've succefully my files, for exemple 4 PDF files for a 4 pages PDF source.... but in the several pdf, I don't have one page but with a random number of page ?
import sys, fitz
file = '/home/ilyes/Bulletins_Originaux.pdf'
bookmark = ''
try:
doc = fitz.open(file)
toc = doc.getToC(simple = True)
except Exception as e:
print(e)
for i in range(len(toc)):
documentPdfCible=toc[i][1]
documentPdfCibleSansSlash=documentPdfCible.replace("/","-")
numeroPage=toc[i][2]
pagedebut=numeroPage
pagefin=numeroPage + 1
print (pagedebut)
print (pagefin)
doc2 = fitz.open(file)
doc2.insertPDF(doc, from_page = pagedebut, to_page = pagefin, start_at = 0)
doc2.save('/home/ilyes/' + documentPdfCibleSansSlash + ".pdf")
doc2.close
Could you tell me what's wrong ?
Maybee because I use always "doc2" in the loop ?
Thanks you,
Abou Ilyès
Seems weird, that you open the same document twice.
You open your pdf file at doc = fitz.open(file) and again at doc2 = fitz.open(file).
Then you insert pages into the same file by doc2.insertPDF(doc, from_page = pagedebut, to_page = pagefin, start_at = 0).
Of course the doc files toc will get messed up completely by "randomly" inserting pages.
I recommend to replace doc2 = fitz.open(file) with doc2 = fitz.open()
This will create an empty "in memory" pdf (see the documentation), in which you can then insert the pages you need from doc. Then save this as a new pdf by its bookmark title by running
doc2.save('/home/ilyes/' + documentPdfCibleSansSlash + ".pdf")

Copying .docx and preserving images

I am trying to copy elements of a doc from one doc file to other. The text part is easy, the images is where it gets tricky.
Attaching an image to explain the structure of the doc: Just some text and 1 image.
from docx import Document
import io
doc = Document('/Users/neha/Desktop/testing.docx')
new_doc = Document()
for elem in doc.element.body:
new_doc.element.body.append(elem)
new_doc.save('/Users/neha/Desktop/out.docx')
This gets me the whole structure of the doc in the new_doc but the image is still blank. Image below:
Good thing is I have the blank image in the right place so I thought of getting the byte level data from the previous image and insert it in the new doc. Here is how I extended the above code:
from docx import Document
import io
doc = Document('/Users/neha/Desktop/testing.docx')
new_doc = Document()
for elem in doc.element.body:
new_doc.element.body.append(elem)
im = doc.inline_shapes[0]
blip = im._inline.graphic.graphicData.pic.blipFill.blip
rId = blip.embed
doc_part = doc.part
image_part = doc_part.related_parts[rId]
bytes = image_part._blob #Here I get the byte level data for the image
im2 = new_doc.inline_shapes[0]
blip2 = im2._inline.graphic.graphicData.pic.blipFill.blip
rId2 = blip2.embed
document_part2 = new_doc.part
document_part2.related_parts[rId2]._blob = bytes
new_doc.save('/Users/neha/Desktop/out.docx')
But the image still shows empty in the new_doc. What should I do from here?
I figured out a solution a couple of days back. However the text loses formatting using this way, but the images are correctly placed.
So the idea is, for para in paras for the source doc, if there is text, I write it to dest doc. And if there is an inline image present, I add a unique identifier at that place in the dest doc (refer here to see how these identifiers work, and contexts in docxtpl). These identifiers and docxtpl proved to be particularly useful here. And then using those unique identifiers I create a 'context' (as shown below) which is basically a map mapping the unique identifier to its particular InlineImage, and finally I render this context..
Below is my code (Apologies for the unnecessary indentation, I copied it directly from my text editor, and shift+tab doesn't work here :P)
from docxtpl import DocxTemplate, InlineImage
import Document
import io
import xml.etree.ElementTree as ET
dest = DocxTemplate()
source = Document(source_path)
context = {}
ims = [im for im in source.inline_shapes]
im_addresses = []
im_streams = []
count = 0
for im in ims:
blip = im._inline.graphic.graphicData.pic.blipFill.blip
rId = blip.embed
doc_part = source.part
image_part = doc_part.related_parts[rId]
byte_data = image_part._blob
image_stream = io.BytesIO(byte_data)
im_streams.append(image_stream)
image_name = self.img_path+"img_"+"_"+str(count)+".jpeg"
with open(image_name, "wb") as fh:
fh.write(byte_data)
fh.close()
im_addresses.append(image_name)
count += 1
paras = source.paragraphs
im_idx = 0
for para in paras:
p = dest.add_paragraph()
r = p.add_run()
if(para.text):
r.add_text(para.text)
root = ET.fromstring(para._p.xml)
namespace = {'wp':"http://schemas.openxmlformats.org/drawingml/2006/wordprocessingDrawing"}
inlines = root.findall('.//wp:inline',namespace)
if(len(inlines) > 0):
uid = "img_"+str(im_idx)
r.add_text("{{ " + uid + " }}")
context[uid] = InlineImage(dest,im_addresses[im_idx])
im_idx += 1
try:
dest.render(context)
except Exception as e:
print(e)
dest.save(dest_path)
PS: If a paragraph has two images, this code will prove to be sub-optimal.. One will have to make some change in the following:
if(len(inlines) > 0):
uid = "img_"+str(im_idx)
r.add_text("{{ " + uid + " }}")
context[uid] = InlineImage(dest,im_addresses[im_idx])
im_idx += 1
Will have to add a for loop inside the if statement as well. Since I didn't need as usually my images were big enough, so they always came in different paragraphs. Just a side note for anyone who may need it..
Cheers!
You could try:
Extracting the images from the first document by unzipping the .docx file (per How can I search a word in a Word 2007 .docx file?)
Save those images to the file system (as foo.png, for instance)
Generate the new .docx file with Python and add the .png file using document.add_picture('foo.png').
This problem is solved by this package https://docxtpl.readthedocs.io/en/latest/

Categories