make sure to retain pageId (physical structMap) in output (#9)

cisocrgroup · May 8, 2019 · 5d7ceb0 · 5d7ceb0
1 parent 12ea2f9
commit 5d7ceb0
Show file tree

Hide file tree

Showing 5 changed files with 5 additions and 0 deletions.
diff --git a/ocrd_cis/align/aligner.py b/ocrd_cis/align/aligner.py
@@ -39,6 +39,7 @@ def process(self):
             out = self.workspace.add_file(
                 ID=ID,
                 file_grp=self.output_file_grp,
+                pageId=ift[0].input_file.pageId,
                 basename=basename,
                 local_filename=os.path.join(self.output_file_grp, basename),
                 mimetype=MIMETYPE_PAGE,

diff --git a/ocrd_cis/div/clean.py b/ocrd_cis/div/clean.py
@@ -65,6 +65,7 @@ def process(self):
             out = self.workspace.add_file(
                 ID=ID,
                 file_grp=self.output_file_grp,
+                pageId=input_file.pageId,
                 basename=self.output_file_grp + '-' + input_file.basename,
                 mimetype=MIMETYPE_PAGE,
                 content=to_xml(pcgts),

diff --git a/ocrd_cis/div/importer.py b/ocrd_cis/div/importer.py
@@ -108,6 +108,7 @@ def process(self):
             out = self.workspace.add_file(
                 ID=ID,
                 file_grp=self.output_file_grp,
+                pageId=input_file.pageId,
                 basename=self.output_file_grp + '-' + input_file.basename,
                 mimetype=MIMETYPE_PAGE,
                 content=to_xml(pcgts),

diff --git a/ocrd_cis/ocropy/recognize.py b/ocrd_cis/ocropy/recognize.py
@@ -316,6 +316,7 @@ def process(self):
             out = self.workspace.add_file(
                 ID=ID,
                 file_grp=self.output_file_grp,
+                pageId=input_file.pageId,
                 basename=input_file.basename,
                 local_filename=os.path.join(self.output_file_grp, input_file.basename),
                 mimetype=MIMETYPE_PAGE,

diff --git a/ocrd_cis/profile/profiler.py b/ocrd_cis/profile/profiler.py
@@ -27,6 +27,7 @@ def process(self):
                 mimetype=MIMETYPE_PAGE,
                 content=to_xml(pcgts),
                 file_grp=self.output_file_grp,
+                pageId=ifile.pageId,
                 basename=ifile.basename,
             )
         self.workspace.save_mets()