ocrd 3.4.0__tar.gz → 3.4.1__tar.gz

{ocrd-3.4.0/src/ocrd.egg-info → ocrd-3.4.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: ocrd
-Version: 3.4.0
+Version: 3.4.1
 Summary: OCR-D framework
 Author-email: Konstantin Baierer <unixprog@gmail.com>
 License: Apache License 2.0

ocrd-3.4.1/VERSION

@@ -0,0 +1 @@
+3.4.1

{ocrd-3.4.0 → ocrd-3.4.1}/src/ocrd/cli/__init__.py

@@ -67,6 +67,12 @@ Variables:
 \b
 {config.describe('OCRD_EXISTING_OUTPUT', wrap_text=False)}
 \b
+{config.describe('OCRD_MAX_MISSING_OUTPUTS')}
+\b
+{config.describe('OCRD_MAX_PARALLEL_PAGES')}
+\b
+{config.describe('OCRD_PROCESSING_PAGE_TIMEOUT')}
+\b
 {config.describe('OCRD_METS_CACHING')}
 \b
 {config.describe('OCRD_MAX_PROCESSOR_CACHE')}

{ocrd-3.4.0 → ocrd-3.4.1}/src/ocrd/decorators/ocrd_cli_options.py

@@ -56,7 +56,7 @@ def ocrd_cli_options(f):
         # subcommands. So we have to work around that by creating a
         # pseudo-subcommand handled in ocrd_cli_wrap_processor
         argument('subcommand', nargs=1, required=False,
-                 type=click.Choice([AgentType.PROCESSING_WORKER, AgentType.PROCESSOR_SERVER])),
+                 type=click.Choice(list(map(str, AgentType)))),
     ]
     for param in params:
         param(f)

{ocrd-3.4.0 → ocrd-3.4.1}/src/ocrd/processor/base.py

@@ -29,8 +29,7 @@ from frozendict import frozendict
 # this is where the fixes came from:
 from loky import Future, ProcessPoolExecutor
 import multiprocessing as mp
-from threading import Timer
-from _thread import interrupt_main
+from multiprocessing.pool import ThreadPool
 
 from click import wrap_text
 from deprecated import deprecated
@@ -783,11 +782,16 @@ class Processor():
         page_id = input_files[input_pos].pageId
         self._base_logger.info("processing page %s", page_id)
         for i, input_file in enumerate(input_files):
+            grp = self.input_file_grp.split(',')[i]
             if input_file is None:
-                grp = self.input_file_grp.split(',')[i]
                 self._base_logger.debug(f"ignoring missing file for input fileGrp {grp} for page {page_id}")
                 continue
             assert isinstance(input_file, get_args(OcrdFileType))
+            if not input_file.local_filename:
+                self._base_logger.error(f'No local file exists for page {page_id} in file group {grp}')
+                if config.OCRD_MISSING_INPUT == 'ABORT':
+                    raise MissingInputFile(grp, page_id, input_file.mimetype)
+                continue
             self._base_logger.debug(f"parsing file {input_file.ID} for page {page_id}")
             try:
                 page_ = page_from_file(input_file)
@@ -796,6 +800,9 @@ class Processor():
             except ValueError as err:
                 # not PAGE and not an image to generate PAGE for
                 self._base_logger.error(f"non-PAGE input for page {page_id}: {err}")
+        if not any(input_pcgts):
+            self._base_logger.warning(f'skipping page {page_id}')
+            return
         output_file_id = make_file_id(input_files[input_pos], self.output_file_grp)
         if input_files[input_pos].fileGrp == self.output_file_grp:
             # input=output fileGrp: re-use ID exactly
@@ -1107,7 +1114,11 @@ class Processor():
             self._base_logger.critical(f"Could not find any files for selected pageId {self.page_id}.\n"
                                        f"compare '{self.page_id}' with the output of 'orcd workspace list-page'.")
         ifts = []
-        for page, ifiles in pages.items():
+        # use physical page order
+        for page in self.workspace.mets.physical_pages:
+            if page not in pages:
+                continue
+            ifiles = pages[page]
             for i, ifg in enumerate(ifgs):
                 if not ifiles[i]:
                     # could be from non-unique with on_error=skip or from true gap
@@ -1150,18 +1161,15 @@ def _page_worker(timeout, *input_files):
     """
     page_id = next((file.pageId for file in input_files
                     if hasattr(file, 'pageId')), "")
-    if timeout > 0:
-        timer = Timer(timeout, interrupt_main)
-        timer.start()
+    pool = ThreadPool(processes=1)
     try:
-        _page_worker_processor.process_page_file(*input_files)
+        #_page_worker_processor.process_page_file(*input_files)
+        async_result = pool.apply_async(_page_worker_processor.process_page_file, input_files)
+        async_result.get(timeout or None)
         _page_worker_processor.logger.debug("page worker completed for page %s", page_id)
-    except KeyboardInterrupt:
+    except mp.TimeoutError:
         _page_worker_processor.logger.debug("page worker timed out for page %s", page_id)
-        raise TimeoutError()
-    finally:
-        if timeout > 0:
-            timer.cancel()
+        raise
 
 def generate_processor_help(ocrd_tool, processor_instance=None, subcommand=None):
     """Generate a string describing the full CLI of this processor including params.

{ocrd-3.4.0 → ocrd-3.4.1}/src/ocrd/workspace.py

@@ -777,16 +777,14 @@ class Workspace():
             raise Exception('Found no AlternativeImage that satisfies all requirements ' +
                             'filename="%s" in page "%s"' % (
                                 filename, page_id))
-        if not all(feature in page_coords['features']
-                   for feature in feature_selector.split(',') if feature):
-            raise Exception('Found no AlternativeImage that satisfies all requirements ' +
-                            'selector="%s" in page "%s"' % (
-                                feature_selector, page_id))
-        if any(feature in page_coords['features']
-               for feature in feature_filter.split(',') if feature):
-            raise Exception('Found no AlternativeImage that satisfies all requirements ' +
-                            'filter="%s" in page "%s"' % (
-                                feature_filter, page_id))
+        if (not all(feature in page_coords['features']
+                    for feature in feature_selector.split(',') if feature) or
+            any(feature in page_coords['features']
+                for feature in feature_filter.split(',') if feature)):
+            raise Exception('Found no AlternativeImage that satisfies all requirements' +
+                            ' selector="%s"' % feature_selector +
+                            ' filter="%s"' % feature_filter +
+                            ' in page "%s"' % page_id)
         # ensure DPI will be set in image meta-data again
         if 'DPI' in page_coords:
             dpi = page_coords['DPI']
@@ -1038,16 +1036,14 @@ class Workspace():
             raise Exception('Found no AlternativeImage that satisfies all requirements ' +
                             'filename="%s" in segment "%s"' % (
                                 filename, segment.id))
-        if not all(feature in segment_coords['features']
-                   for feature in feature_selector.split(',') if feature):
+        if (not all(feature in segment_coords['features']
+                    for feature in feature_selector.split(',') if feature) or
+            any(feature in segment_coords['features']
+                for feature in feature_filter.split(',') if feature)):
             raise Exception('Found no AlternativeImage that satisfies all requirements' +
-                            'selector="%s" in segment "%s"' % (
-                                feature_selector, segment.id))
-        if any(feature in segment_coords['features']
-               for feature in feature_filter.split(',') if feature):
-            raise Exception('Found no AlternativeImage that satisfies all requirements ' +
-                            'filter="%s" in segment "%s"' % (
-                                feature_filter, segment.id))
+                            ' selector="%s"' % feature_selector +
+                            ' filter="%s"' % feature_filter +
+                            ' in segment "%s"' % segment.id)
         # ensure DPI will be set in image meta-data again
         if 'DPI' in segment_coords:
             dpi = segment_coords['DPI']

{ocrd-3.4.0 → ocrd-3.4.1/src/ocrd.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: ocrd
-Version: 3.4.0
+Version: 3.4.1
 Summary: OCR-D framework
 Author-email: Konstantin Baierer <unixprog@gmail.com>
 License: Apache License 2.0

{ocrd-3.4.0 → ocrd-3.4.1}/src/ocrd_network/constants.py

@@ -11,12 +11,16 @@ OCRD_ALL_TOOL_JSON = "ocrd-all-tool.json"
 SERVER_ALL_PAGES_PLACEHOLDER = "all_pages"
 
 
-class AgentType(str, Enum):
+class StrEnum(str, Enum):
+    def __str__(self):
+        return self.value
+
+class AgentType(StrEnum):
     PROCESSING_WORKER = "worker"
     PROCESSOR_SERVER = "server"
 
 
-class DeployType(str, Enum):
+class DeployType(StrEnum):
     # Deployed by the Processing Server config file
     DOCKER = "docker"
     NATIVE = "native"
@@ -26,7 +30,7 @@ class DeployType(str, Enum):
 
 
 # TODO: Make the states uppercase
-class JobState(str, Enum):
+class JobState(StrEnum):
     # The processing job is cached inside the Processing Server requests cache
     cached = "CACHED"
     # The processing job was cancelled due to failed dependencies
@@ -43,7 +47,7 @@ class JobState(str, Enum):
     unset = "UNSET"
 
 
-class NetworkLoggingDirs(str, Enum):
+class NetworkLoggingDirs(StrEnum):
     METS_SERVERS = "mets_servers"
     PROCESSING_JOBS = "processing_jobs"
     PROCESSING_SERVERS = "processing_servers"
@@ -51,7 +55,7 @@ class NetworkLoggingDirs(str, Enum):
     PROCESSOR_SERVERS = "processor_servers"
 
 
-class ServerApiTags(str, Enum):
+class ServerApiTags(StrEnum):
     ADMIN = "admin"
     DISCOVERY = "discovery"
     PROCESSING = "processing"

{ocrd-3.4.0 → ocrd-3.4.1}/src/ocrd_utils/os.py

@@ -254,7 +254,7 @@ def guess_media_type(input_file : str, fallback : str = None, application_xml :
     if mimetype is None:
         mimetype = EXT_TO_MIME.get(''.join(Path(input_file).suffixes), fallback)
     if mimetype is None:
-        raise ValueError("Could not determine MIME type of input_file must")
+        raise ValueError(f"Could not determine MIME type of {input_file}")
     if mimetype == 'application/xml':
         mimetype = application_xml
     return mimetype

ocrd-3.4.0/VERSION

@@ -1 +0,0 @@
-3.4.0