pdftextract 0.0.5

a very fast and efficient text and image pdf extractor.

PdfTextract

A very fast and efficient python PDF text & images extractor that uses the xpdf c++ library.

Features

• Several times fatser then any python based pdf text extractor
• very easy and simple to use
• Extract text while maintaining original document layout
• Trys to automaticaly extract tables if they exist (still in beta)
• No local server setup required
• No dependencies needed

Instalation

Install via PyPi:

pip install pdftextract

or via github:

1. first clone the repo:

git clone https://github.com/Bnilss/pdftextract.git

2. then run

python setup.py

Usage

1. Importing the package

from pdftextract import XPdf

file_path = "examples/pubmed_example.pdf"
pdf = XPdf(file_path)

2. Get the PDF meta-data

print(pdf.info) # this will return a dict of pdf metadata (author, size, pages..)
# to get the number of pages for example
print(pdf.info['Pages'])

3. Extracting text from all pages in a PDF and return it as string

txt = pdf.to_text()
print(txt)

4. Extracting a single page only, to get the 3rd page for example

# we can extrat using the previous method (start_index=1)
txt = pdf.to_text(just_one=3)
# or use the bracket notation (start_index=0)
txt = pdf[2]

5. Extracting text from a single page (page 7) and saving it to .txt file

pdf.to_text("page7.txt", just_one=7)

6. Extracting text from page 1 to 5

txt = pdf.to_text(start=1, end=5)
# or
txt = pdf[:5]

7. Extract tables

pdf = XPdf("examples/table_sample.pdf")
txt = pdf.to_text(table=True)
# the use a regex or something to parse the text ..
# or try automatic paring (still in beta)
tables = pdf.table[:]
print(len(tables)) # 3
print(tables[0]) # print formated content of table 1
#Number of Coils | Number of Paperclips
#______________________________________
#       5        |       3, 5, 4
#      10        |       7, 8, 6
#      15        |      11, 10, 12
#      20        |      15, 13, 14
table1_data = tables[0].data # will return all rows in table except headers

OS support

by default the package support windows.

to use it in linux or mac:

1. download xpdf files for linux or mac
2. extract the files in got to os version (32/64 bit)
3. copy these files: pdftotext, pdfinfo, pdfimages
4. Got to the packages-directory/pdftextract/xpdf
5. past the file in that directory

Credits

• xpdf c++ by Derek Noonburg

License

pdftextract is licensed under the GNU General Public License (GPL), version 3.