add pdfplumber input module (#404)

README.md

@@ -8,7 +8,7 @@ A command line tool and Python library to support your accounting
 process.
 
 1. extracts text from PDF files using different techniques, like
-   `pdftotext`, `text`, `pdfminer` or OCR -- `tesseract`, `tesseract4` or
+   `pdftotext`, `text`, `pdfminer`, `pdfplumber` or OCR -- `tesseract`, `tesseract4` or
    `gvision` (Google Cloud Vision).
 2. searches for regex in the result using a YAML-based template system
 3. saves results as CSV, JSON or XML or renames PDF files to match the content.
@@ -58,6 +58,7 @@ Choose any of the following input readers:
 - pdftotext `invoice2data --input-reader text invoice.txt`
 - tesseract `invoice2data --input-reader tesseract invoice.pdf`
 - pdfminer.six `invoice2data --input-reader pdfminer invoice.pdf`
+- pdfplumber `invoice2data --input-reader pdfplumber invoice.pdf`
 - tesseract4 `invoice2data --input-reader tesseract4 invoice.pdf`
 - gvision `invoice2data --input-reader gvision invoice.pdf` (needs `GOOGLE_APPLICATION_CREDENTIALS` env var)
 

setup.cfg

@@ -41,7 +41,7 @@ install_requires =
   unidecode
 
 [options.extras_require]
-test = pytest; pytest-cov; flake8; pdfminer.six; tox
+test = pytest; pytest-cov; flake8; pdfminer.six; pdfplumber; tox
 
 [options.entry_points]
 console_scripts =

src/invoice2data/input/pdfplumber.py

@@ -0,0 +1,48 @@
+# -*- coding: utf-8 -*-
+
+import logging
+
+logger = logging.getLogger(__name__)
+
+
+def to_text(path):
+    """Wrapper around `pdfplumber`.
+    Parameters
+    ----------
+    path : str
+        path of electronic invoice in PDF
+    Returns
+    -------
+    str : str
+        returns extracted text from pdf
+    """
+    try:
+        import pdfplumber
+    except ImportError:
+        logger.debug("Cannot import pdfplumber")
+
+    raw_text = ""
+    raw_text = raw_text.encode(encoding='UTF-8')
+    with pdfplumber.open(path, laparams={"detect_vertical": True}) as pdf:
+        pages = []
+        for pdf_page in pdf.pages:
+            pages.append(
+                pdf_page.extract_text(
+                    layout=True, use_text_flow=True, x_tolerance=6, y_tolerance=4, keep_blank_chars=True
+                )  # y_tolerance=6, dirty Fix for html table problem
+            )
+        res = {
+            "all": "\n\n".join(pages),
+            "first": pages and pages[0] or "",
+        }
+    logger.debug("Text extraction made with pdfplumber")
+
+    raw_text = res_to_raw_text(res)
+    return raw_text.encode("utf-8")
+
+
+def res_to_raw_text(res):
+    # we need to convert result to raw text:
+    raw_text_dict = res
+    raw_text = (raw_text_dict["first"] or raw_text_dict["all"])
+    return raw_text

src/invoice2data/main.py

@@ -11,6 +11,7 @@
 
 from .input import pdftotext
 from .input import pdfminer_wrapper
+from .input import pdfplumber
 from .input import tesseract
 from .input import tesseract4
 from .input import gvision
@@ -30,6 +31,7 @@
     "tesseract": tesseract,
     "tesseract4": tesseract4,
     "pdfminer": pdfminer_wrapper,
+    "pdfplumber": pdfplumber,
     "gvision": gvision,
     "text": text,
 }

tests/test_lib.py

@@ -19,7 +19,7 @@
 import unittest
 
 from invoice2data.main import extract_data
-from invoice2data.input import pdftotext, tesseract, pdfminer_wrapper
+from invoice2data.input import pdftotext, tesseract, pdfminer_wrapper, pdfplumber
 from invoice2data.output import to_csv, to_json, to_xml
 from .common import get_sample_files
 
@@ -87,6 +87,11 @@ def test_extract_data_pdfminer(self):
                 self.assertTrue(False, "pdfminer is not installed")
                 self.assertTrue(type(res) is str, "return is not a string")
 
+    def test_extract_data_pdfplumber(self):
+        pdf_files = get_sample_files('.pdf')
+        for file in pdf_files:
+            extract_data(file, None, pdfplumber)
+
     def test_tesseract_for_return(self):
         png_files = get_sample_files('.png')
         for file in png_files: