ocrd 3.5.1__py3-none-any.whl → 3.7.0__py3-none-any.whl

ocrd_utils/__init__.py

@@ -8,11 +8,11 @@ Utility functions and constants usable in various circumstances.
     levels below page (i.e. region, line, word, glyph) between relative coordinates
     w.r.t. a corresponding image and absolute coordinates w.r.t. the top-level image.
     This includes rotation and offset correction, based on affine transformations.
-    (Used by :py:class:`ocrd.workspace.Workspace` methods 
-    :py:meth:`ocrd.workspace.Workspace.image_from_page` and 
+    (Used by :py:class:`ocrd.workspace.Workspace` methods
+    :py:meth:`ocrd.workspace.Workspace.image_from_page` and
     :py:meth:`ocrd.workspace.Workspace.image_from_segment`.)
 
-* :py:func:`rotate_coordinates`, 
+* :py:func:`rotate_coordinates`,
   :py:func:`scale_coordinates`,
   :py:func:`shift_coordinates`,
   :py:func:`transpose_coordinates`,
@@ -23,7 +23,7 @@ Utility functions and constants usable in various circumstances.
     used to pass down the coordinate system along with images (both invariably sharing
     the same operations context) when traversing the element hierarchy top to bottom.
     (Used by :py:class:`ocrd.workspace.Workspace` methods
-    :py:meth:`ocrd.workspace.Workspace.image_from_page` and 
+    :py:meth:`ocrd.workspace.Workspace.image_from_page` and
     :py:meth:`ocrd.workspace.Workspace.image_from_segment`.)
 
 * :py:func:`rotate_image`,

ocrd_utils/config.py

@@ -16,9 +16,11 @@ from textwrap import fill, indent
 def _validator_boolean(val):
     return isinstance(val, bool) or str.lower(val) in ('true', 'false', '0', '1')
 
+
 def _parser_boolean(val):
     return bool(val) if isinstance(val, (int, bool)) else str.lower(val) in ('true', '1')
 
+
 class OcrdEnvVariable():
 
     def __init__(self, name, description, parser=str, validator=lambda _: True, default=[False, None]):
@@ -61,7 +63,7 @@ class OcrdEnvVariable():
                 desc += ' '
             desc += f'(Default: "{default}")'
         ret = ''
-        ret  = f'{self.name}\n'
+        ret = f'{self.name}\n'
         if wrap_text:
             desc = fill(desc, width=50)
         if indent_text:
@@ -69,6 +71,7 @@ class OcrdEnvVariable():
             desc = indent(desc, '    ')
         return ret + desc
 
+
 class OcrdEnvConfig():
 
     def __init__(self):
@@ -83,7 +86,7 @@ class OcrdEnvConfig():
         return self._variables[name]
 
     def has_default(self, name):
-        if not name in self._variables:
+        if name not in self._variables:
             raise ValueError(f"Unregistered env variable {name}")
         return self._variables[name].has_default
 
@@ -99,13 +102,13 @@ class OcrdEnvConfig():
                 pass
 
     def describe(self, name, *args, **kwargs):
-        if not name in self._variables:
+        if name not in self._variables:
             raise ValueError(f"Unregistered env variable {name}")
         return self._variables[name].describe(*args, **kwargs)
 
     def __getattr__(self, name):
         # will be called if name is not accessible (has not been added directly yet)
-        if not name in self._variables:
+        if name not in self._variables:
             raise AttributeError(f"Unregistered env variable {name}")
         var_obj = self._variables[name]
         try:
@@ -120,39 +123,44 @@ class OcrdEnvConfig():
         return var_obj.parser(raw_value)
 
     def is_set(self, name):
-        if not name in self._variables:
+        if name not in self._variables:
             raise ValueError(f"Unregistered env variable {name}")
         return name in environ
 
     def raw_value(self, name):
-        if not name in self._variables:
+        if name not in self._variables:
             raise ValueError(f"Unregistered env variable {name}")
         return environ[name]
 
+
 config = OcrdEnvConfig()
 
 config.add('OCRD_METS_CACHING',
-    description='If set to `true`, access to the METS file is cached, speeding in-memory search and modification.',
-    validator=_validator_boolean,
-    parser=_parser_boolean)
+           description='If set to `true`, access to the METS file is cached, speeding in-memory search and modification.',
+           validator=_validator_boolean,
+           parser=_parser_boolean)
 
 config.add('OCRD_MAX_PROCESSOR_CACHE',
-    description="Maximum number of processor instances (for each set of parameters) to be kept in memory (including loaded models) for processing workers or processor servers.",
-    parser=int,
-    default=(True, 128))
+           description="Maximum number of processor instances (for each set of parameters) to be kept in memory "
+           "(including loaded models) for processing workers.",
+           parser=int,
+           default=(True, 128))
 
 config.add('OCRD_MAX_PARALLEL_PAGES',
-    description="Maximum number of processor workers for page-parallel processing (within each Processor's selected page range, independent of the number of Processing Workers or Processor Servers). If set >1, then a METS Server must be used for METS synchronisation.",
-    parser=int,
-    default=(True, 1))
+           description="Maximum number of processor workers for page-parallel processing (within "
+           "each Processor's selected page range, independent of the number of Processing "
+           "Workers). If set >1, then a METS Server must be used for METS synchronisation.",
+           parser=int,
+           default=(True, 1))
 
 config.add('OCRD_PROCESSING_PAGE_TIMEOUT',
-    description="Timeout in seconds for processing a single page. If set >0, when exceeded, the same as OCRD_MISSING_OUTPUT applies.",
-    parser=int,
-    default=(True, 0))
+           description="Timeout in seconds for processing a single page. If set >0, when exceeded, "
+           "the same as OCRD_MISSING_OUTPUT applies.",
+           parser=int,
+           default=(True, 0))
 
 config.add("OCRD_PROFILE",
-    description="""\
+           description="""\
 Whether to enable gathering runtime statistics
 on the `ocrd.profile` logger (comma-separated):
 \b
@@ -161,16 +169,18 @@ on the `ocrd.profile` logger (comma-separated):
 - `PSS`: also yields peak memory (proportional set size)
 \b
 """,
-  validator=lambda val : all(t in ('', 'CPU', 'RSS', 'PSS') for t in val.split(',')),
-  default=(True, ''))
+           validator=lambda val: all(t in ('', 'CPU', 'RSS', 'PSS') for t in val.split(',')),
+           default=(True, ''))
 
 config.add("OCRD_PROFILE_FILE",
-    description="If set, then the CPU profile is written to this file for later peruse with a analysis tools like snakeviz")
+           description="If set, then the CPU profile is written to this file for later peruse "
+           "with a analysis tools like snakeviz")
 
 config.add("OCRD_DOWNLOAD_RETRIES",
-    description="Number of times to retry failed attempts for downloads of resources or workspace files.",
-    validator=int,
-    parser=int)
+           description="Number of times to retry failed attempts for downloads of resources or workspace files.",
+           validator=int,
+           parser=int)
+
 
 def _ocrd_download_timeout_parser(val):
     timeout = val.split(',')
@@ -180,18 +190,19 @@ def _ocrd_download_timeout_parser(val):
         timeout = float(timeout[0])
     return timeout
 
+
 config.add("OCRD_DOWNLOAD_TIMEOUT",
-    description="Timeout in seconds for connecting or reading (comma-separated) when downloading.",
-    parser=_ocrd_download_timeout_parser)
+           description="Timeout in seconds for connecting or reading (comma-separated) when downloading.",
+           parser=_ocrd_download_timeout_parser)
 
 config.add("OCRD_DOWNLOAD_INPUT",
-    description="Whether to download files not present locally during processing",
-    default=(True, True),
-    validator=_validator_boolean,
-    parser=_parser_boolean)
+           description="Whether to download files not present locally during processing",
+           default=(True, True),
+           validator=_validator_boolean,
+           parser=_parser_boolean)
 
 config.add("OCRD_MISSING_INPUT",
-    description="""\
+           description="""\
 How to deal with missing input files
 (for some fileGrp/pageId) during processing:
 \b
@@ -199,12 +210,12 @@ How to deal with missing input files
  - `ABORT`: throw :py:class:`.MissingInputFile`
 \b
 """,
-    default=(True, 'SKIP'),
-    validator=lambda val: val in ['SKIP', 'ABORT'],
-    parser=str)
+           default=(True, 'SKIP'),
+           validator=lambda val: val in ['SKIP', 'ABORT'],
+           parser=str)
 
 config.add("OCRD_MISSING_OUTPUT",
-    description="""\
+           description="""\
 How to deal with missing output files
 (for some fileGrp/pageId) during processing:
 \b
@@ -213,17 +224,18 @@ How to deal with missing output files
  - `ABORT`: re-throw whatever caused processing to fail
 \b
 """,
-    default=(True, 'SKIP'),
-    validator=lambda val: val in ['SKIP', 'COPY', 'ABORT'],
-    parser=str)
+           default=(True, 'SKIP'),
+           validator=lambda val: val in ['SKIP', 'COPY', 'ABORT'],
+           parser=str)
 
 config.add("OCRD_MAX_MISSING_OUTPUTS",
-    description="Maximal rate of skipped/fallback pages among all processed pages before aborting (decimal fraction, ignored if negative).",
-    default=(True, 0.1),
-    parser=float)
+           description="Maximal rate of skipped/fallback pages among all processed pages before aborting "
+           "(decimal fraction, ignored if negative).",
+           default=(True, 0.1),
+           parser=float)
 
 config.add("OCRD_EXISTING_OUTPUT",
-    description="""\
+           description="""\
 How to deal with already existing output files
 (for some fileGrp/pageId) during processing:
 \b
@@ -232,13 +244,13 @@ How to deal with already existing output files
  - `ABORT`: re-throw :py:class:`FileExistsError`
 \b
 """,
-    default=(True, 'SKIP'),
-    validator=lambda val: val in ['SKIP', 'OVERWRITE', 'ABORT'],
-    parser=str)
+           default=(True, 'SKIP'),
+           validator=lambda val: val in ['SKIP', 'OVERWRITE', 'ABORT'],
+           parser=str)
 
 config.add("OCRD_NETWORK_SERVER_ADDR_PROCESSING",
-        description="Default address of Processing Server to connect to (for `ocrd network client processing`).",
-        default=(True, ''))
+           description="Default address of Processing Server to connect to (for `ocrd network client processing`).",
+           default=(True, ''))
 
 config.add("OCRD_NETWORK_CLIENT_POLLING_SLEEP",
            description="How many seconds to sleep before trying again.",
@@ -251,27 +263,25 @@ config.add("OCRD_NETWORK_CLIENT_POLLING_TIMEOUT",
            default=(True, 3600))
 
 config.add("OCRD_NETWORK_SERVER_ADDR_WORKFLOW",
-        description="Default address of Workflow Server to connect to (for `ocrd network client workflow`).",
-        default=(True, ''))
+           description="Default address of Workflow Server to connect to (for `ocrd network client workflow`).",
+           default=(True, ''))
 
 config.add("OCRD_NETWORK_SERVER_ADDR_WORKSPACE",
-        description="Default address of Workspace Server to connect to (for `ocrd network client workspace`).",
-        default=(True, ''))
+           description="Default address of Workspace Server to connect to (for `ocrd network client workspace`).",
+           default=(True, ''))
 
 config.add("OCRD_NETWORK_RABBITMQ_CLIENT_CONNECT_ATTEMPTS",
            description="Number of attempts for a RabbitMQ client to connect before failing.",
            parser=int,
            default=(True, 3))
 
-config.add(
-    name="OCRD_NETWORK_RABBITMQ_HEARTBEAT",
-    description="""
-    Controls AMQP heartbeat timeout (in seconds) negotiation during connection tuning. An integer value always overrides the value 
-    proposed by broker. Use 0 to deactivate heartbeat.
-    """,
-    parser=int,
-    default=(True, 0)
-)
+config.add(name="OCRD_NETWORK_RABBITMQ_HEARTBEAT",
+           description="""
+Controls AMQP heartbeat timeout (in seconds) negotiation during connection tuning. An integer value always overrides the value
+proposed by broker. Use 0 to deactivate heartbeat.
+           """,
+           parser=int,
+           default=(True, 0))
 
 config.add(name="OCRD_NETWORK_SOCKETS_ROOT_DIR",
            description="The root directory where all mets server related socket files are created",
@@ -296,24 +306,24 @@ except PermissionError:
     pass
 
 config.add("HOME",
-    description="Directory to look for `ocrd_logging.conf`, fallback for unset XDG variables.",
-    # description="HOME directory, cf. https://specifications.freedesktop.org/basedir-spec/basedir-spec-latest.html",
-    validator=lambda val: Path(val).is_dir(),
-    parser=lambda val: Path(val),
-    default=(True, lambda: Path.home()))
+           description="Directory to look for `ocrd_logging.conf`, fallback for unset XDG variables.",
+           # description="HOME directory, cf. https://specifications.freedesktop.org/basedir-spec/basedir-spec-latest.html",
+           validator=lambda val: Path(val).is_dir(),
+           parser=lambda val: Path(val),
+           default=(True, lambda: Path.home()))
 
 config.add("XDG_DATA_HOME",
-    description="Directory to look for `./ocrd-resources/*` (i.e. `ocrd resmgr` data location)",
-    parser=lambda val: Path(val),
-    default=(True, lambda: Path(config.HOME, '.local/share')))
+           description="Directory to look for `./ocrd-resources/*` (i.e. `ocrd resmgr` data location)",
+           parser=lambda val: Path(val),
+           default=(True, lambda: Path(config.HOME, '.local/share')))
 
 config.add("XDG_CONFIG_HOME",
-    description="Directory to look for `./ocrd/resources.yml` (i.e. `ocrd resmgr` user database)",
-    parser=lambda val: Path(val),
-    default=(True, lambda: Path(config.HOME, '.config')))
+           description="Directory to look for `./ocrd/resources.yml` (i.e. `ocrd resmgr` user database)",
+           parser=lambda val: Path(val),
+           default=(True, lambda: Path(config.HOME, '.config')))
 
 config.add("OCRD_LOGGING_DEBUG",
-    description="Print information about the logging setup to STDERR",
-    default=(True, False),
-    validator=_validator_boolean,
-    parser=_parser_boolean)
+           description="Print information about the logging setup to STDERR",
+           default=(True, False),
+           validator=_validator_boolean,
+           parser=_parser_boolean)

ocrd_utils/deprecate.py

@@ -1,9 +1,11 @@
 import functools
 import warnings
 
+
 def deprecation_warning(msg, stacklevel=2):
     warnings.warn(msg, DeprecationWarning, stacklevel)
 
+
 def deprecated_alias(**aliases):
     """
     Deprecate a kwarg in favor of another kwarg
@@ -16,6 +18,7 @@ def deprecated_alias(**aliases):
         return wrapper
     return deco
 
+
 def rename_kwargs(func_name, kwargs, aliases):
     """
     https://stackoverflow.com/questions/49802412/how-to-implement-deprecation-in-python-with-argument-alias