dinglehopper: Support disabling metrics in the OCR-D interface

pull/23/head
Gerber, Mike 5 years ago
parent f699697eb3
commit 668de758a0

@ -57,6 +57,13 @@ This generates HTML and JSON reports in the `OCR-D-OCR-TESS-EVAL` filegroup.
![dinglehopper displaying metrics and character differences](.screenshots/dinglehopper.png?raw=true) ![dinglehopper displaying metrics and character differences](.screenshots/dinglehopper.png?raw=true)
You may also want to disable metrics and the green-red color scheme by
parameter:
~~~
ocrd-dinglehopper -I ABBYY-FULLTEXT,OCR-D-OCR-CALAMARI -O OCR-D-OCR-COMPARE-ABBYY-CALAMARI -p '{"metrics": false}'
~~~
Testing Testing
------- -------
Use `pytest` to run the tests in [the tests directory](qurator/dinglehopper/tests): Use `pytest` to run the tests in [the tests directory](qurator/dinglehopper/tests):

@ -16,7 +16,14 @@
], ],
"steps": [ "steps": [
"recognition/text-recognition" "recognition/text-recognition"
] ],
"parameters": {
"metrics": {
"type": "boolean",
"default": true,
"description": "Enable/disable metrics and green/red"
}
}
} }
} }
} }

@ -34,6 +34,7 @@ class OcrdDinglehopperEvaluate(Processor):
return file_id return file_id
def process(self): def process(self):
metrics = self.parameter['metrics']
gt_grp, ocr_grp = self.input_file_grp.split(',') gt_grp, ocr_grp = self.input_file_grp.split(',')
for n, page_id in enumerate(self.workspace.mets.physical_pages): for n, page_id in enumerate(self.workspace.mets.physical_pages):
gt_file = self.workspace.mets.find_files(fileGrp=gt_grp, pageId=page_id)[0] gt_file = self.workspace.mets.find_files(fileGrp=gt_grp, pageId=page_id)[0]
@ -50,7 +51,12 @@ class OcrdDinglehopperEvaluate(Processor):
os.mkdir(self.output_file_grp) os.mkdir(self.output_file_grp)
except FileExistsError: except FileExistsError:
pass pass
cli_process(gt_file.local_filename, ocr_file.local_filename, report_prefix) cli_process(
gt_file.local_filename,
ocr_file.local_filename,
report_prefix,
metrics=metrics
)
# Add reports to the workspace # Add reports to the workspace
for report_suffix, mimetype in \ for report_suffix, mimetype in \

Loading…
Cancel
Save