Remove all trailing whitespace

.circleci/config.yml

@@ -27,7 +27,7 @@ jobs:
     steps:
       - checkout
       - run: sudo apt-get -y update
-      - run: sudo make deps-ubuntu 
+      - run: sudo make deps-ubuntu
       - run: make install deps-test
       - run: make test benchmark
 

.github/workflows/docker-image.yml

@@ -38,4 +38,3 @@ jobs:
       run: |
         docker push ${{ env.DOCKER_TAG }}:latest
         docker push ${{ env.DOCKER_TAG }}-cuda:latest
-

CHANGELOG.md

@@ -1294,7 +1294,7 @@ Added:
   * Workspace validation will check cardinality of images per file is 1, #243, OCR-D/spec#132
 
 Changed:
-  
+
   * bashlib will no longer warn about "non-conformant" file group names, #365
   * Invalid `file:/` URL will now raise exceptions, #373
   * image_from_*: increase tolerance for size mismatch after rotation to 2px, #371

Makefile

@@ -264,7 +264,7 @@ pyclean:
 .PHONY: docker docker-cuda
 
 # Additional arguments to docker build. Default: '$(DOCKER_ARGS)'
-DOCKER_ARGS = 
+DOCKER_ARGS =
 
 # Build docker image
 docker: DOCKER_BASE_IMAGE = ubuntu:20.04
@@ -277,7 +277,7 @@ docker-cuda: DOCKER_FILE = Dockerfile.cuda
 
 docker-cuda: docker
 
-docker docker-cuda: 
+docker docker-cuda:
 	docker build --progress=plain -f $(DOCKER_FILE) -t $(DOCKER_TAG) --build-arg BASE_IMAGE=$(DOCKER_BASE_IMAGE) $(DOCKER_ARGS) .
 
 # Build wheels and source dist and twine upload them

ocrd/ocrd/cli/workspace.py

@@ -56,7 +56,7 @@ def workspace_cli(ctx, directory, mets, mets_basename, mets_server_url, backup):
 
     A workspace comprises a METS file and a directory as point of reference.
 
-    Operates on the file system directly or via a METS server 
+    Operates on the file system directly or via a METS server
     (already running via some prior `server start` subcommand).
     """
     initLogging()

ocrd/ocrd/processor/base.py

@@ -44,7 +44,7 @@ class Processor():
     for run-time data processing. That is, it executes a single workflow step,
     or a combination of workflow steps, on the workspace (represented by local METS).
     It reads input files for all or requested physical pages of the input fileGrp(s),
-    and writes output files for them into the output fileGrp(s). It may take 
+    and writes output files for them into the output fileGrp(s). It may take
     a number of optional or mandatory parameters.
     """
 
@@ -166,12 +166,12 @@ def verify(self):
 
     def process(self):
         """
-        Process the :py:attr:`workspace` 
+        Process the :py:attr:`workspace`
         from the given :py:attr:`input_file_grp`
         to the given :py:attr:`output_file_grp`
         for the given :py:attr:`page_id`
         under the given :py:attr:`parameter`.
-        
+
         (This contains the main functionality and needs to be overridden by subclasses.)
         """
         raise Exception("Must be implemented")
@@ -282,7 +282,7 @@ def input_files(self):
         - Otherwise raise an error (complaining that only PAGE-XML warrants
           having multiple images for a single page)
         Algorithm <https://github.com/cisocrgroup/ocrd_cis/pull/57#issuecomment-656336593>_
-        
+
         Returns:
             A list of :py:class:`ocrd_models.ocrd_file.OcrdFile` objects.
         """

ocrd/ocrd/resolver.py

@@ -175,7 +175,7 @@ def workspace_from_url(
             src_baseurl (string, None): Base URL for resolving relative file locations
             **kwargs (): Passed on to ``OcrdMets.find_files`` if download == True
 
-        Download (clone) :py:attr:`mets_url` to ``mets.xml`` in :py:attr:`dst_dir`, unless 
+        Download (clone) :py:attr:`mets_url` to ``mets.xml`` in :py:attr:`dst_dir`, unless
         the former is already local and the latter is ``none`` or already identical to its directory name.
 
         Returns:

ocrd/ocrd/resource_manager.py

@@ -158,9 +158,9 @@ def list_installed(self, executable=None):
                     resdict = resdict_list[0]
                 elif str(res_filename.parent) == moduledir:
                     resdict = {
-                        'name': res_name, 
-                        'url': str(res_filename), 
-                        'description': 'Found at module', 
+                        'name': res_name,
+                        'url': str(res_filename),
+                        'description': 'Found at module',
                         'type': res_type,
                         'size': res_size
                     }

ocrd_models/README.md

@@ -36,7 +36,7 @@ Let's say you want to add a method `get_FirstTextRegion` on the `pc:Page` elemen
 
     Would add the method `exportChildren` from a file `exportChildren_PageType.py`.
 
-    > **Note**: 
+    > **Note**:
     > The method name in the file must match the method name passed to
     > `_add_method`. This is *not* checked automatically, so double-check manually!
 

ocrd_models/ocrd_models/ocrd_file.py

@@ -235,7 +235,7 @@ def __init__(self, el, mimetype=None, pageId=None, loctype='OTHER', local_filena
             mimetype (string): ``@MIMETYPE`` of this ``mets:file``
             pageId (string): ``@ID`` of the physical ``mets:structMap`` entry corresponding to this ``mets:file``
             loctype (string): ``@LOCTYPE`` of this ``mets:file``
-            url (string): ignored XXX the remote/original file once we have proper mets:FLocat bookkeeping 
+            url (string): ignored XXX the remote/original file once we have proper mets:FLocat bookkeeping
             local_filename (): ``@xlink:href`` of this ``mets:file`` - XXX the local file once we have proper mets:FLocat bookkeeping
             ID (string): ``@ID`` of this ``mets:file``
         """

ocrd_models/ocrd_models/ocrd_mets.py

@@ -137,7 +137,7 @@ def _clear_caches(self):
         self._file_cache = None
         self._page_cache = None
         self._fptr_cache = None
-        
+
     def refresh_caches(self):
         if self._cache_flag:
             # Cache for the files (mets:file) - two nested dictionaries
@@ -158,11 +158,11 @@ def refresh_caches(self):
             # The inner dictionary's Key: 'fptr.FILEID'
             # The inner dictionary's Value: a 'fptr' object at some memory location
             self._fptr_cache = {}
-            
+
             # Note, if the empty_mets() function is used to instantiate OcrdMets
             # Then the cache is empty even after this operation
             self._fill_caches()
-        
+
     @property
     def unique_identifier(self):
         """
@@ -173,7 +173,7 @@ def unique_identifier(self):
             found = self._tree.getroot().find('.//mods:identifier[@type="%s"]' % t, NS)
             if found is not None:
                 return found.text
-        
+
     @unique_identifier.setter
     def unique_identifier(self, purl):
         """
@@ -268,8 +268,8 @@ def find_files(
             local_filename (string) : ``@xlink:href`` local/cached filename of ``mets:Flocat`` of ``mets:file``
             mimetype (string) : ``@MIMETYPE`` of ``mets:file``
             local (boolean) : Whether to restrict results to local files in the filesystem
-            include_fileGrp (list[str]) : Whitelist of allowd file groups 
-            exclude_fileGrp (list[str]) : Blacklist of disallowd file groups 
+            include_fileGrp (list[str]) : Whitelist of allowd file groups
+            exclude_fileGrp (list[str]) : Blacklist of disallowd file groups
         Yields:
             :py:class:`ocrd_models:ocrd_file:OcrdFile` instantiations
         """
@@ -303,7 +303,7 @@ def find_files(
             mimetype = re.compile(mimetype[REGEX_PREFIX_LEN:])
         if url and url.startswith(REGEX_PREFIX):
             url = re.compile(url[REGEX_PREFIX_LEN:])
-            
+
         candidates = []
         if self._cache_flag:
             if fileGrp:
@@ -315,7 +315,7 @@ def find_files(
                 candidates = [el_file for id_to_file in self._file_cache.values() for el_file in id_to_file.values()]
         else:
             candidates = self._tree.getroot().xpath('//mets:file', namespaces=NS)
-            
+
         for cand in candidates:
             if ID:
                 if isinstance(ID, str):
@@ -404,7 +404,7 @@ def rename_file_group(self, old, new):
         if el_fileGrp is None:
             raise FileNotFoundError("No such fileGrp '%s'" % old)
         el_fileGrp.set('USE', new)
-        
+
         if self._cache_flag:
             self._file_cache[new] = self._file_cache.pop(old)
 
@@ -452,7 +452,7 @@ def remove_file_group(self, USE, recursive=False, force=False):
 
         if self._cache_flag:
             # Note: Since the files inside the group are removed
-            # with the 'remove_one_file' method above, 
+            # with the 'remove_one_file' method above,
             # we should not take care of that again.
             # We just remove the fileGrp.
             del self._file_cache[el_fileGrp.get('USE')]
@@ -591,7 +591,7 @@ def physical_pages(self):
         """
         if self._cache_flag:
             return list(self._page_cache.keys())
-            
+
         return [str(x) for x in self._tree.getroot().xpath(
             'mets:structMap[@TYPE="PHYSICAL"]/mets:div[@TYPE="physSequence"]/mets:div[@TYPE="page"]/@ID',
             namespaces=NS)]
@@ -604,7 +604,7 @@ def get_physical_pages(self, for_fileIds=None):
         if for_fileIds is None:
             return self.physical_pages
         ret = [None] * len(for_fileIds)
-        
+
         if self._cache_flag:
             for pageId in self._fptr_cache.keys():
                 for fptr in self._fptr_cache[pageId].keys():
@@ -657,14 +657,14 @@ def set_physical_page_for_file(self, pageId, ocrd_file, order=None, orderlabel=N
         if el_seqdiv is None:
             el_seqdiv = ET.SubElement(el_structmap, TAG_METS_DIV)
             el_seqdiv.set('TYPE', 'physSequence')
-        
+
         el_pagediv = None
         if self._cache_flag:
             if pageId in self._page_cache:
                 el_pagediv = self._page_cache[pageId]
         else:
             el_pagediv = el_seqdiv.find('mets:div[@ID="%s"]' % pageId, NS)
-        
+
         if el_pagediv is None:
             el_pagediv = ET.SubElement(el_seqdiv, TAG_METS_DIV)
             el_pagediv.set('TYPE', 'page')
@@ -676,10 +676,10 @@ def set_physical_page_for_file(self, pageId, ocrd_file, order=None, orderlabel=N
             if self._cache_flag:
                 # Create a new entry in the page cache
                 self._page_cache[pageId] = el_pagediv
-                # Create a new entry in the fptr cache and 
+                # Create a new entry in the fptr cache and
                 # assign an empty dictionary to hold the fileids
                 self._fptr_cache[pageId] = {}
-                
+
         el_fptr = ET.SubElement(el_pagediv, TAG_METS_FPTR)
         el_fptr.set('FILEID', ocrd_file.ID)
 
@@ -756,7 +756,7 @@ def remove_physical_page_fptr(self, fileId):
         if self._cache_flag:
             for page_id in self._fptr_cache.keys():
                 if fileId in self._fptr_cache[page_id].keys():
-                    mets_fptrs.append(self._fptr_cache[page_id][fileId]) 
+                    mets_fptrs.append(self._fptr_cache[page_id][fileId])
         else:
             mets_fptrs = self._tree.getroot().xpath(
                 'mets:structMap[@TYPE="PHYSICAL"]/mets:div[@TYPE="physSequence"]/mets:div[@TYPE="page"]/mets:fptr[@FILEID="%s"]' % fileId, namespaces=NS)