Add CLI

camelot/cli.py

@@ -1 +1,84 @@
+# -*- coding: utf-8 -*-
+from pprint import pprint
+
 import click
+
+from .io import read_pdf
+from .plotting import plot_geometry
+from .utils import validate_input, remove_extra
+
+
+class Mutex(click.Option):
+    def handle_parse_result(self, ctx, opts, args):
+        mesh = opts.get('mesh', False)
+        geometry_type = opts.get('geometry_type', False)
+        validate_input(opts, mesh=mesh, geometry_type=geometry_type)
+        return super(Mutex, self).handle_parse_result(ctx, opts, args)
+
+
+@click.command()
+@click.option("-p", "--pages", default="1", help="")
+@click.option("-o", "--output", help="")
+@click.option("-f", "--format",
+              type=click.Choice(["csv", "json", "excel", "html"]), help="")
+@click.option("-z", "--zip", is_flag=True, help="")
+@click.option("-m", "--mesh", is_flag=True, help="Whether or not to"
+              "use Lattice method of parsing. Stream is used by default.")
+@click.option("-G", "--geometry_type",
+              type=click.Choice(["text", "table", "contour", "joint", "line"]),
+              help="Plot geometry found on pdf page for debugging.")
+@click.option("-T", "--table_area", default=[], multiple=True,
+              help="")
+@click.option("-split", "--split_text", is_flag=True, help="")
+@click.option("-flag", "--flag_size", is_flag=True, help="")
+@click.option("-M", "--margins", nargs=3, default=(1.0, 0.5, 0.1),
+              help="")
+@click.option("-C", "--columns", default=[], multiple=True, cls=Mutex,
+              help="")
+@click.option("-r", "--row_close_tol", default=2, cls=Mutex, help="")
+@click.option("-c", "--col_close_tol", default=0, cls=Mutex, help="")
+@click.option("-back", "--process_background", is_flag=True, cls=Mutex,
+              help="Use with --mesh")
+@click.option("-scale", "--line_size_scaling", default=15, cls=Mutex,
+              help="Use with --mesh")
+@click.option("-copy", "--copy_text", default=[], cls=Mutex,
+              help="Use with --mesh")
+@click.option("-shift", "--shift_text", default=["l", "t"], cls=Mutex,
+              help="Use with --mesh")
+@click.option("-l", "--line_close_tol", default=2, cls=Mutex,
+              help="Use with --mesh")
+@click.option("-j", "--joint_close_tol", default=2, cls=Mutex,
+              help="Use with --mesh")
+@click.option("-block", "--threshold_blocksize", default=15, cls=Mutex,
+              help="Use with --mesh")
+@click.option("-const", "--threshold_constant", default=-2, cls=Mutex,
+              help="Use with --mesh")
+@click.option("-I", "--iterations", default=0, cls=Mutex,
+              help="Use with --mesh")
+@click.argument("filepath", type=click.Path(exists=True))
+def cli(*args, **kwargs):
+    pages = kwargs.pop("pages")
+    output = kwargs.pop("output")
+    f = kwargs.pop("format")
+    compress = kwargs.pop("zip")
+    mesh = kwargs.pop("mesh")
+    geometry_type = kwargs.pop("geometry_type")
+    filepath = kwargs.pop("filepath")
+
+    table_area = list(kwargs['table_area'])
+    kwargs['table_area'] = None if not table_area else table_area
+    columns = list(kwargs['columns'])
+    kwargs['columns'] = None if not columns else columns
+
+    kwargs = remove_extra(kwargs, mesh=mesh)
+    if geometry_type is None:
+        tables = read_pdf(filepath, pages=pages, mesh=mesh, **kwargs)
+        click.echo(tables)
+        if output is None:
+            raise click.UsageError("Please specify an output filepath using --output")
+        if f is None:
+            raise click.UsageError("Please specify an output format using --format")
+        tables.export(output, f=f, compress=compress)
+    else:
+        plot_geometry(filepath, pages=pages, mesh=mesh,
+                      geometry_type=geometry_type, **kwargs)

camelot/io.py

@@ -1,4 +1,5 @@
 from .handlers import PDFHandler
+from .utils import validate_input, remove_extra
 
 
 def read_pdf(filepath, pages='1', mesh=False, **kwargs):
@@ -78,17 +79,14 @@ def read_pdf(filepath, pages='1', mesh=False, **kwargs):
         PDFMiner margins. (char_margin, line_margin, word_margin)
 
         For for information, refer `PDFMiner docs <https://euske.github.io/pdfminer/>`_.
-    debug : bool, optional (default: False)
-        Whether or not to return all text objects on the page
-        which can be used to generate a matplotlib plot, to get
-        values for table_area(s) and debugging.
 
     Returns
     -------
     tables : camelot.core.TableList
 
     """
-    # validate kwargs?
+    validate_input(kwargs, mesh=mesh)
     p = PDFHandler(filepath, pages)
+    kwargs = remove_extra(kwargs, mesh=mesh)
     tables, __ = p.parse(mesh=mesh, **kwargs)
     return tables

camelot/plotting.py

@@ -3,9 +3,10 @@ import matplotlib.pyplot as plt
 import matplotlib.patches as patches
 
 from .handlers import PDFHandler
+from .utils import validate_input, remove_extra
 
 
-def plot_geometry(filepath, pages='1', mesh=False, geometry_type='text', **kwargs):
+def plot_geometry(filepath, pages='1', mesh=False, geometry_type=None, **kwargs):
     """Plot geometry found on pdf page based on type specified,
     useful for debugging and playing with different parameters to get
     the best output.
@@ -23,7 +24,7 @@ def plot_geometry(filepath, pages='1', mesh=False, geometry_type='text', **kwarg
     mesh : bool (default: False)
         Whether or not to use Lattice method of parsing. Stream
         is used by default.
-    geometry_type : str, optional (default: 'text')
+    geometry_type : str, optional (default: None)
         'text' : Plot text objects found on page, useful to get
                  table_area and columns coordinates.
         'table' : Plot parsed table.
@@ -91,15 +92,12 @@ def plot_geometry(filepath, pages='1', mesh=False, geometry_type='text', **kwarg
         PDFMiner margins. (char_margin, line_margin, word_margin)
 
         For for information, refer `PDFMiner docs <https://euske.github.io/pdfminer/>`_.
-    debug : bool, optional (default: False)
-        Whether or not to return all text objects on the page
-        which can be used to generate a matplotlib plot, to get
-        values for table_area(s) and debugging.
 
     """
-    # validate kwargs?
+    validate_input(kwargs, mesh=mesh, geometry_type=geometry_type)
     p = PDFHandler(filepath, pages)
-    debug = True if geometry_type else False
+    kwargs = remove_extra(kwargs, mesh=mesh)
+    debug = True if geometry_type is not None else False
     kwargs.update({'debug': debug})
     __, geometry = p.parse(mesh=mesh, **kwargs)
 
@@ -140,8 +138,6 @@ def plot_geometry(filepath, pages='1', mesh=False, geometry_type='text', **kwarg
                                      [cell.lb[1], cell.rb[1]])
             plt.show()
     elif geometry_type == 'contour':
-        if not mesh:
-            raise ValueError("Use mesh=True")
         for img, table_bbox in geometry.images:
             for t in table_bbox.keys():
                 cv2.rectangle(img, (t[0], t[1]),
@@ -149,8 +145,6 @@ def plot_geometry(filepath, pages='1', mesh=False, geometry_type='text', **kwarg
             plt.imshow(img)
             plt.show()
     elif geometry_type == 'joint':
-        if not mesh:
-            raise ValueError("Use mesh=True")
         for img, table_bbox in geometry.images:
             x_coord = []
             y_coord = []
@@ -164,8 +158,6 @@ def plot_geometry(filepath, pages='1', mesh=False, geometry_type='text', **kwarg
             plt.imshow(img)
             plt.show()
     elif geometry_type == 'line':
-        if not mesh:
-            raise ValueError("Use mesh=True")
         for v_s, h_s in geometry.segments:
             for v in v_s:
                 plt.plot([v[0], v[2]], [v[1], v[3]])

camelot/utils.py

@@ -20,6 +20,53 @@ from pdfminer.layout import (LAParams, LTAnno, LTChar, LTTextLineHorizontal,
                              LTTextLineVertical)
 
 
+stream_kwargs = [
+    'columns',
+    'row_close_tol',
+    'col_close_tol'
+]
+lattice_kwargs = [
+    'process_background',
+    'line_size_scaling',
+    'copy_text',
+    'shift_text',
+    'line_close_tol',
+    'joint_close_tol',
+    'threshold_blocksize',
+    'threshold_constant',
+    'iterations'
+]
+
+
+def validate_input(kwargs, mesh=False, geometry_type=False):
+    def check_intersection(parser_kwargs, input_kwargs, message_bool):
+        isec = set(parser_kwargs).intersection(set(input_kwargs.keys()))
+        if isec:
+            raise ValueError("{} can not be used with mesh set to {}".format(
+                             ",".join(sorted(isec)), message_bool))
+
+    if mesh:
+        check_intersection(stream_kwargs, kwargs, True)
+    else:
+        check_intersection(lattice_kwargs, kwargs, False)
+    if geometry_type:
+        if not mesh and geometry_type in ['contour', 'joint', 'line']:
+            raise ValueError("Use geometry_type={} with mesh set to True".format(
+                             geometry_type))
+
+
+def remove_extra(kwargs, mesh=False):
+    if mesh:
+        for key in kwargs.keys():
+            if key in stream_kwargs:
+                kwargs.pop(key)
+    else:
+        for key in kwargs.keys():
+            if key in lattice_kwargs:
+                kwargs.pop(key)
+    return kwargs
+
+
 # https://stackoverflow.com/a/22726782
 class TemporaryDirectory(object):
     def __enter__(self):

setup.py

@@ -49,7 +49,12 @@ def setup_package():
                     author_email=AUTHOR_EMAIL,
                     license=LICENSE,
                     packages=['camelot'],
-                    install_requires=reqs)
+                    install_requires=reqs,
+                    entry_points={
+                        'console_scripts': [
+                            'camelot = camelot.cli:cli',
+                        ],
+                    })
 
     try:
         from setuptools import setup