Merge pull request #190 from camelot-dev/vinayak/2025-01-03-fix-deprecated-camelot-methods

Fix deprecated camelot methods

Author: vinayak-mehta

.github/workflows/unittests.yml renamed to .github/workflows/tests.yml

@@ -1,7 +1,4 @@
-# This workflow will install Python dependencies, run tests
-# For more information see: https://help.github.com/actions/language-and-framework-guides/using-python-with-github-actions
-
-name: Run Unittests
+name: Tests
 
 on:
   push:
@@ -26,7 +23,7 @@ jobs:
     - name: Install dependencies
       run: |
         python -m pip install --upgrade pip
-        python -m pip install ".[all]"
+        python -m pip install ".[dev]"
     - name: Test with pytest
       run: |
         python -m pytest

docs/user/install.rst

@@ -8,13 +8,10 @@ This part of the documentation covers the steps to install Excalibur.
 Using pip
 ---------
 
-After installing `ghostscript`_, which is one of the requirements for Camelot (See `install instructions`_), you can simply use pip to install Excalibur::
+You can simply use pip to install Excalibur::
 
     $ pip install excalibur-py
 
-.. _ghostscript: https://www.ghostscript.com/
-.. _install instructions: https://camelot-py.readthedocs.io/en/master/user/install-deps.html
-
 From the source code
 --------------------
 

excalibur/cli.py

@@ -3,12 +3,12 @@
 
 import click
 
-from . import settings, __version__
+from . import __version__, settings
 from . import configuration as conf
-from .tasks import split, extract
-from .www.app import create_app
-from .utils.database import reset_database, initialize_database
 from .operators.python_operator import PythonOperator
+from .tasks import extract, split
+from .utils.database import initialize_database, reset_database
+from .www.app import create_app
 
 
 def abort_if_false(ctx, param, value):

excalibur/models.py

@@ -1,15 +1,15 @@
-import json
 import datetime as dt
+import json
 from typing import Any  # noqa
 
 from sqlalchemy import (
-    Text,
-    Column,
-    String,
     Boolean,
-    Integer,
+    Column,
     DateTime,
     ForeignKey,
+    Integer,
+    String,
+    Text,
 )
 from sqlalchemy.ext.declarative import declarative_base
 

excalibur/tasks.py

@@ -7,7 +7,7 @@
 
 import camelot
 import pandas as pd
-from camelot.backends.ghostscript_backend import GhostscriptBackend
+from camelot.backends.pdfium_backend import PdfiumBackend
 from camelot.core import TableList
 from camelot.parsers import Lattice, Stream
 
@@ -43,23 +43,8 @@ def split(file_id):
             imagepath = os.path.join(conf.PDFS_FOLDER, file_id, imagename)
 
             # convert single-page PDF to PNG
-            try:
-                backend = GhostscriptBackend()
-                backend.convert(filepath, imagepath, 300)
-            except OSError:
-                gs_command = [
-                    "gs",
-                    "-q",
-                    "-sDEVICE=png16m",
-                    f"-o{imagepath}",
-                    "-r300",
-                    filepath,
-                ]
-                try:
-                    subprocess.run(gs_command, check=True, capture_output=True)
-                except subprocess.CalledProcessError as e:
-                    logging.error(f"Ghostscript conversion failed: {e.stderr.decode()}")
-                    raise
+            backend = PdfiumBackend()
+            backend.convert(filepath, imagepath, 300)
 
             filenames[page] = filename
             filepaths[page] = filepath
@@ -123,7 +108,7 @@ def extract(job_id):
             if flavor.lower() == "lattice":
                 kwargs.pop("columns", None)
 
-            t = camelot.read_pdf(filepaths[p], **kwargs, backend="poppler")
+            t = camelot.read_pdf(filepaths[p], **kwargs)
             for _t in t:
                 _t.page = int(p)
             tables.extend(t)

excalibur/utils/task.py

@@ -1,8 +1,12 @@
 import os
 
 import cv2
+from camelot.utils import (
+    get_image_char_and_text_objects,
+    get_page_layout,
+    get_rotation,
+)
 from PyPDF2 import PdfReader, PdfWriter
-from camelot.utils import get_rotation, get_page_layout, get_text_objects
 
 
 def get_pages(filename, pages, password=""):
@@ -62,9 +66,9 @@ def save_page(filepath, page_number):
     froot, fext = os.path.splitext(outpath)
     layout, __ = get_page_layout(outpath)
     # fix rotated PDF
-    chars = get_text_objects(layout, ltype="char")
-    horizontal_text = get_text_objects(layout, ltype="horizontal_text")
-    vertical_text = get_text_objects(layout, ltype="vertical_text")
+    images, chars, horizontal_text, vertical_text = get_image_char_and_text_objects(
+        layout
+    )
     rotation = get_rotation(chars, horizontal_text, vertical_text)
     if rotation != "":
         outpath_new = "".join([froot.replace("page", "p"), "_rotated", fext])

excalibur/www/static/js/workspace.js

@@ -110,7 +110,7 @@ const onSavedRuleClick = function (e) {
 
         if (ruleOptions['flavor'].toLowerCase() == 'lattice') {
           document.getElementById('process-background').value = ruleOptions['process_background'];
-          document.getElementById('line-size-scaling').value = ruleOptions['line_size_scaling'];
+          document.getElementById('line-scale').value = ruleOptions['line_scale'];
           document.getElementById('split-text-l').value = ruleOptions['split_text'];
           document.getElementById('flag-size-l').value = ruleOptions['flag_size'];
         } else if (ruleOptions['flavor'].toLowerCase() == 'stream') {
@@ -176,14 +176,14 @@ const getRuleOptions = function () {
     switch(flavor.toString().toLowerCase()) {
       case 'lattice': {
         ruleOptions['process_background'] = $("#process-background").val() ? true : false;
-        ruleOptions['line_size_scaling'] = $('#line-size-scaling').val() ? Number($('#line-size-scaling').val()) : 15;
+        ruleOptions['line_scale'] = $('#line-scale').val() ? Number($('#line-scale').val()) : 15;
         ruleOptions['split_text'] = $("#split-text-l").val() ? true : false;
         ruleOptions['flag_size'] = $("#flag-size-l").val() ? true : false;
         break;
       }
       case 'stream': {
-        ruleOptions['row_close_tol'] = $('#row-close-tol').val() ? Number($('#line-size-scaling').val()) : 2;
-        ruleOptions['col_close_tol'] = $('#col-close-tol').val() ? Number($('#line-size-scaling').val()) : 0;
+        ruleOptions['row_close_tol'] = $('#row-close-tol').val() ? Number($('#line-scale').val()) : 2;
+        ruleOptions['col_close_tol'] = $('#col-close-tol').val() ? Number($('#line-scale').val()) : 0;
         ruleOptions['split_text'] = $("#split-text-s").val() ? true : false;
         ruleOptions['flag_size'] = $("#flag-size-s").val() ? true : false;
         break;