ocrd.processor.base.run_processor() - Code Metrics - Inspection of ":memo: changelog" - OCR-D/core - Measure and Improve Code Quality continuously with Scrutinizer

Passed

Push — master ( 77064b...da9bd1 )

by Konstantin

created 2020-08-22 12:59 UTC

ocrd.processor.base.run_processor() B

↳ Parent: ocrd.processor.base

Complexity

Conditions

Size

Total Lines	59
Code Lines	48

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
eloc	48
dl	0
loc	59
rs	8.7018
c	0
b	0
f	0
cc	4
nop	12

How to fix Long Method Many Parameters

"""
Processor base class and helper functions
"""

__all__ = ['Processor', 'generate_processor_help', 'run_cli', 'run_processo']

import os
import json
from ocrd_utils import getLogger, VERSION as OCRD_VERSION, MIMETYPE_PAGE
from ocrd_validators import ParameterValidator

# XXX imports must remain for backwards-compatibilty
from .helpers import run_cli, run_processor, generate_processor_help # pylint: disable=unused-import

log = getLogger('ocrd.processor')

class Processor():
    """
    A processor runs an algorithm based on the workspace, the mets.xml in the
    workspace (and the input files defined therein) as well as optional
    parameter.
    """

    def __init__(
            self,
            workspace,
            ocrd_tool=None,
            parameter=None,
            # TODO OCR-D/core#274
            # input_file_grp=None,
            # output_file_grp=None,
            input_file_grp="INPUT",
            output_file_grp="OUTPUT",
            page_id=None,
            show_help=False,
            show_version=False,
            dump_json=False,
            version=None
    ):
        if parameter is None:
            parameter = {}
        if dump_json:
            print(json.dumps(ocrd_tool, indent=True))
            return
        self.ocrd_tool = ocrd_tool
        if show_help:
            self.show_help()
            return
        self.version = version
        if show_version:
            self.show_version()
            return
        self.workspace = workspace
        # FIXME HACK would be better to use pushd_popd(self.workspace.directory)
        # but there is no way to do that in process here since it's an
        # overridden method. chdir is almost always an anti-pattern.
        if self.workspace:
            os.chdir(self.workspace.directory)
        self.input_file_grp = input_file_grp
        self.output_file_grp = output_file_grp
        self.page_id = None if page_id == [] or page_id is None else page_id
        parameterValidator = ParameterValidator(ocrd_tool)
        report = parameterValidator.validate(parameter)
        if not report.is_valid:
            raise Exception("Invalid parameters %s" % report.errors)
        self.parameter = parameter

    def show_help(self):
        print(generate_processor_help(self.ocrd_tool))

    def show_version(self):
        print("Version %s, ocrd/core %s" % (self.version, OCRD_VERSION))

    def verify(self):
        """
        Verify that the input fulfills the processor's requirements.
        """
        return True

    def process(self):
        """
        Process the workspace
        """
        raise Exception("Must be implemented")

    @property
    def input_files(self):
        """
        List the input files.

        - If there's a PAGE-XML for the page, take it (and forget about all
          other files for that page)
        - Else if there's only one image, take it (and forget about all other
          files for that page)
        - Otherwise raise an error (complaining that only PAGE-XML warrants

          having multiple images for a single page)
        (https://github.com/cisocrgroup/ocrd_cis/pull/57#issuecomment-656336593)
        """
        ret = self.workspace.mets.find_files(
            fileGrp=self.input_file_grp, pageId=self.page_id, mimetype=MIMETYPE_PAGE)
        if ret:
            return ret
        ret = self.workspace.mets.find_files(
            fileGrp=self.input_file_grp, pageId=self.page_id, mimetype="//image/.*")
        if self.page_id and len(ret) > 1:
            raise ValueError("No PAGE-XML %s in fileGrp '%s' but multiple images." % (
                "for page '%s'" % self.page_id if self.page_id else '',
                self.input_file_grp
                ))
        return ret


1			"""
2			Processor base class and helper functions
3			"""
4
5			__all__ = ['Processor', 'generate_processor_help', 'run_cli', 'run_processo']
6
7			import os
8			import json
9			from ocrd_utils import getLogger, VERSION as OCRD_VERSION, MIMETYPE_PAGE
10			from ocrd_validators import ParameterValidator
11
12			# XXX imports must remain for backwards-compatibilty
13			from .helpers import run_cli, run_processor, generate_processor_help # pylint: disable=unused-import
14
15			log = getLogger('ocrd.processor')
16
17			class Processor():
18			"""
19			A processor runs an algorithm based on the workspace, the mets.xml in the
20			workspace (and the input files defined therein) as well as optional
21			parameter.
22			"""
23
24			def __init__(
25			self,
26			workspace,
27			ocrd_tool=None,
28			parameter=None,
29			# TODO OCR-D/core#274
30			# input_file_grp=None,
31			# output_file_grp=None,
32			input_file_grp="INPUT",
33			output_file_grp="OUTPUT",
34			page_id=None,
35			show_help=False,
36			show_version=False,
37			dump_json=False,
38			version=None
39			):
40			if parameter is None:
41			parameter = {}
42			if dump_json:
43			print(json.dumps(ocrd_tool, indent=True))
44			return
45			self.ocrd_tool = ocrd_tool
46			if show_help:
47			self.show_help()
48			return
49			self.version = version
50			if show_version:
51			self.show_version()
52			return
53			self.workspace = workspace
54			# FIXME HACK would be better to use pushd_popd(self.workspace.directory)
55			# but there is no way to do that in process here since it's an
56			# overridden method. chdir is almost always an anti-pattern.
57			if self.workspace:
58			os.chdir(self.workspace.directory)
59			self.input_file_grp = input_file_grp
60			self.output_file_grp = output_file_grp
61			self.page_id = None if page_id == [] or page_id is None else page_id
62			parameterValidator = ParameterValidator(ocrd_tool)
63			report = parameterValidator.validate(parameter)
64			if not report.is_valid:
65			raise Exception("Invalid parameters %s" % report.errors)
66			self.parameter = parameter
67
68			def show_help(self):
69			print(generate_processor_help(self.ocrd_tool))
70
71			def show_version(self):
72			print("Version %s, ocrd/core %s" % (self.version, OCRD_VERSION))
73
74			def verify(self):
75			"""
76			Verify that the input fulfills the processor's requirements.
77			"""
78			return True
79
80			def process(self):
81			"""
82			Process the workspace
83			"""
84			raise Exception("Must be implemented")
85
86			@property
87			def input_files(self):
88			"""
89			List the input files.
90
91			- If there's a PAGE-XML for the page, take it (and forget about all
92			other files for that page)
93			- Else if there's only one image, take it (and forget about all other
94			files for that page)
95			- Otherwise raise an error (complaining that only PAGE-XML warrants
96
97			having multiple images for a single page)
98			(https://github.com/cisocrgroup/ocrd_cis/pull/57#issuecomment-656336593)
99			"""
100			ret = self.workspace.mets.find_files(
101			fileGrp=self.input_file_grp, pageId=self.page_id, mimetype=MIMETYPE_PAGE)
102			if ret:
103			return ret
104			ret = self.workspace.mets.find_files(
105			fileGrp=self.input_file_grp, pageId=self.page_id, mimetype="//image/.*")
106			if self.page_id and len(ret) > 1:
107			raise ValueError("No PAGE-XML %s in fileGrp '%s' but multiple images." % (
108			"for page '%s'" % self.page_id if self.page_id else '',
109			self.input_file_grp
110			))
111			return ret
112

OCR-D / core

Push — master ( 77064b...da9bd1 )

ocrd.processor.base.run_processor() B

Complexity

Size

Duplication

Importance

How to fix Long Method Many Parameters

Long Method

Many Parameters

Duplication Side-by-Side

Filter issues like