|
|
|
@ -1,61 +1,55 @@
|
|
|
|
|
import json
|
|
|
|
|
from functools import cached_property
|
|
|
|
|
import os
|
|
|
|
|
from typing import Optional
|
|
|
|
|
|
|
|
|
|
import click
|
|
|
|
|
from ocrd_models import OcrdFileType
|
|
|
|
|
from ocrd import Processor
|
|
|
|
|
from ocrd.decorators import ocrd_cli_options, ocrd_cli_wrap_processor
|
|
|
|
|
from ocrd_utils import assert_file_grp_cardinality, getLogger, make_file_id
|
|
|
|
|
from pkg_resources import resource_string
|
|
|
|
|
from ocrd_utils import make_file_id
|
|
|
|
|
|
|
|
|
|
from .cli import process as cli_process
|
|
|
|
|
|
|
|
|
|
OCRD_TOOL = json.loads(resource_string(__name__, "ocrd-tool.json").decode("utf8"))
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
@click.command()
|
|
|
|
|
@ocrd_cli_options
|
|
|
|
|
def ocrd_dinglehopper(*args, **kwargs):
|
|
|
|
|
return ocrd_cli_wrap_processor(OcrdDinglehopperEvaluate, *args, **kwargs)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
class OcrdDinglehopperEvaluate(Processor):
|
|
|
|
|
def __init__(self, *args, **kwargs):
|
|
|
|
|
kwargs["ocrd_tool"] = OCRD_TOOL["tools"]["ocrd-dinglehopper"]
|
|
|
|
|
super(OcrdDinglehopperEvaluate, self).__init__(*args, **kwargs)
|
|
|
|
|
|
|
|
|
|
def process(self):
|
|
|
|
|
assert_file_grp_cardinality(self.input_file_grp, 2, "GT and OCR")
|
|
|
|
|
assert_file_grp_cardinality(self.output_file_grp, 1)
|
|
|
|
|
@cached_property
|
|
|
|
|
def executable(self):
|
|
|
|
|
return 'ocrd-dinglehopper'
|
|
|
|
|
|
|
|
|
|
log = getLogger("processor.OcrdDinglehopperEvaluate")
|
|
|
|
|
def process_page_file(self, *input_files: Optional[OcrdFileType]) -> None:
|
|
|
|
|
|
|
|
|
|
assert self.parameter
|
|
|
|
|
metrics = self.parameter["metrics"]
|
|
|
|
|
textequiv_level = self.parameter["textequiv_level"]
|
|
|
|
|
gt_grp, ocr_grp = self.input_file_grp.split(",")
|
|
|
|
|
|
|
|
|
|
input_file_tuples = self.zip_input_files(on_error="abort")
|
|
|
|
|
for n, (gt_file, ocr_file) in enumerate(input_file_tuples):
|
|
|
|
|
# wrong number of inputs: let fail
|
|
|
|
|
gt_file, ocr_file = input_files
|
|
|
|
|
# missing on either side: skip (zip_input_files already warned)
|
|
|
|
|
if not gt_file or not ocr_file:
|
|
|
|
|
# file/page was not found in this group
|
|
|
|
|
continue
|
|
|
|
|
gt_file = self.workspace.download_file(gt_file)
|
|
|
|
|
ocr_file = self.workspace.download_file(ocr_file)
|
|
|
|
|
page_id = gt_file.pageId
|
|
|
|
|
return
|
|
|
|
|
# missing download (i.e. OCRD_DOWNLOAD_INPUT=false):
|
|
|
|
|
if not gt_file.local_filename:
|
|
|
|
|
if config.OCRD_MISSING_INPUT == 'ABORT':
|
|
|
|
|
raise MissingInputFile(gt_file.fileGrp, gt_file.pageId, gt_file.mimetype)
|
|
|
|
|
return
|
|
|
|
|
if not ocr_file.local_filename:
|
|
|
|
|
if config.OCRD_MISSING_INPUT == 'ABORT':
|
|
|
|
|
raise MissingInputFile(ocr_file.fileGrp, ocr_file.pageId, ocr_file.mimetype)
|
|
|
|
|
return
|
|
|
|
|
|
|
|
|
|
log.info("INPUT FILES %i / %s↔ %s", n, gt_file, ocr_file)
|
|
|
|
|
page_id = gt_file.pageId
|
|
|
|
|
|
|
|
|
|
file_id = make_file_id(ocr_file, self.output_file_grp)
|
|
|
|
|
report_prefix = os.path.join(self.output_file_grp, file_id)
|
|
|
|
|
|
|
|
|
|
# Process the files
|
|
|
|
|
try:
|
|
|
|
|
os.mkdir(self.output_file_grp)
|
|
|
|
|
except FileExistsError:
|
|
|
|
|
pass
|
|
|
|
|
cli_process(
|
|
|
|
|
gt_file.local_filename,
|
|
|
|
|
ocr_file.local_filename,
|
|
|
|
|
report_prefix,
|
|
|
|
|
file_id,
|
|
|
|
|
self.output_file_grp,
|
|
|
|
|
metrics=metrics,
|
|
|
|
|
textequiv_level=textequiv_level,
|
|
|
|
|
)
|
|
|
|
@ -65,12 +59,16 @@ class OcrdDinglehopperEvaluate(Processor):
|
|
|
|
|
[".html", "text/html"],
|
|
|
|
|
[".json", "application/json"],
|
|
|
|
|
]:
|
|
|
|
|
output_file_id = file_id + report_suffix
|
|
|
|
|
output_file = next(self.workspace.mets.find_files(ID=output_file_id), None)
|
|
|
|
|
if output_file and config.OCRD_EXISTING_OUTPUT != 'OVERWRITE':
|
|
|
|
|
raise FileExistsError(f"A file with ID=={output_file_id} already exists {output_file} and neither force nor ignore are set")
|
|
|
|
|
self.workspace.add_file(
|
|
|
|
|
file_id=file_id + report_suffix,
|
|
|
|
|
file_id=output_file_id,
|
|
|
|
|
file_grp=self.output_file_grp,
|
|
|
|
|
page_id=page_id,
|
|
|
|
|
mimetype=mimetype,
|
|
|
|
|
local_filename=report_prefix + report_suffix,
|
|
|
|
|
local_filename=file_id + report_suffix,
|
|
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|