Пример #1
0
def process_dataset(parameters):
    global extractorName, workerScript, inputDirectory, outputDirectory

    # Find input files in dataset
    files = get_all_files(parameters)

    # Download files to input directory
    for fileExt in files:
        files[fileExt]['path'] = extractors.download_file(
            channel=parameters['channel'],
            header=parameters['header'],
            host=parameters['host'],
            key=parameters['secretKey'],
            fileid=files[fileExt]['id'],
            # What's this argument for?
            intermediatefileid=files[fileExt]['id'],
            ext=fileExt)
        # Restore temp filenames to original - script requires specific name formatting so tmp names aren't suitable
        files[fileExt]['old_path'] = files[fileExt]['path']
        files[fileExt]['path'] = os.path.join(inputDirectory,
                                              files[fileExt]['filename'])
        os.rename(files[fileExt]['old_path'], files[fileExt]['path'])
        print 'found %s file: %s' % (fileExt, files[fileExt]['path'])

    # Invoke terraref.sh
    outFilePath = os.path.join(outputDirectory,
                               get_output_filename(files['_raw']['filename']))
    print 'invoking terraref.sh to create: %s' % outFilePath
    returncode = subprocess.call([
        "bash", workerScript, "-d", "1", "-I", inputDirectory, "-O",
        outputDirectory
    ])
    print 'done creating output file (%s)' % (returncode)

    if returncode != 0:
        print 'terraref.sh encountered an error'

    # Verify outfile exists and upload to clowder
    if os.path.exists(outFilePath):
        print 'output file detected'
        if returncode == 0:
            print 'uploading output file...'
            extractors.upload_file_to_dataset(filepath=outFilePath,
                                              parameters=parameters)
            print 'done uploading'
        # Clean up the output file.
        os.remove(outFilePath)
    else:
        print 'no output file was produced'

    print 'cleaning up...'
    # Clean up the input files.
    for fileExt in files:
        os.remove(files[fileExt]['path'])
    print 'done cleaning'
def process_dataset(parameters):
    global outputDir

    metafile, bin_file, metadata = None, None, None

    # Get left/right files and metadata
    for f in parameters['files']:
        # First check metadata attached to dataset in Clowder for item of interest
        if f.endswith('_dataset_metadata.json'):
            all_dsmd = getFlir.load_json(f)
            for curr_dsmd in all_dsmd:
                if 'content' in curr_dsmd and 'lemnatec_measurement_metadata' in curr_dsmd[
                        'content']:
                    metafile = f
                    metadata = curr_dsmd['content']
        # Otherwise, check if metadata was uploaded as a .json file
        elif f.endswith('_metadata.json') and f.find(
                '/_metadata.json') == -1 and metafile is None:
            metafile = f
            metadata = getFlir.load_json(metafile)
        elif f.endswith('_ir.bin'):
            bin_file = f
    if None in [metafile, bin_file, metadata]:
        getFlir.fail('Could not find all of ir.bin/metadata.')
        return

    print("...bin_file: %s" % bin_file)
    print("...metafile: %s" % metafile)
    dsname = parameters["datasetInfo"]["name"]
    if dsname.find(" - ") > -1:
        timestamp = dsname.split(" - ")[1]
    else:
        timestamp = "dsname"
    if timestamp.find("__") > -1:
        datestamp = timestamp.split("__")[0]
    else:
        datestamp = ""
    out_dir = os.path.join(outputDir, datestamp, timestamp)
    print("...output directory: %s" % out_dir)
    if not os.path.exists(out_dir):
        os.makedirs(out_dir)

    #Determine output paths
    binbase = os.path.basename(bin_file)[:-7]
    png_path = os.path.join(out_dir, binbase + '.png')
    tiff_path = os.path.join(out_dir, binbase + '.tif')
    print("...png: %s" % (png_path))
    print("...tif: %s" % (tiff_path))

    print("Creating png image")
    raw_data = getFlir.load_flir_data(bin_file)  # get raw data from bin file
    im_color = getFlir.create_png(raw_data, png_path)  # create png
    print("Uploading output PNGs to dataset")
    extractors.upload_file_to_dataset(png_path, parameters)

    print("getting information from json file for geoTIFF")
    center_position, scan_time, fov = getFlir.parse_metadata(metadata)
    if center_position is None or scan_time is None or fov is None:
        print("error getting metadata; skipping geoTIFF")
    else:
        gps_bounds = getFlir.get_bounding_box(
            center_position,
            fov)  # get bounding box using gantry position and fov of camera

        print("Creating geoTIFF images")
        tc = getFlir.rawData_to_temperature(raw_data, scan_time,
                                            metadata)  # get temperature
        getFlir.create_geotiff_with_temperature(im_color, tc, gps_bounds,
                                                tiff_path)  # create geotiff
        print("Uploading output geoTIFFs to dataset")
        extractors.upload_file_to_dataset(tiff_path, parameters)

    # Tell Clowder this is completed so subsequent file updates don't daisy-chain
    metadata = {
        "@context": {
            "@vocab":
            "https://clowder.ncsa.illinois.edu/clowder/assets/docs/api/index.html#!/files/uploadToDataset"
        },
        "dataset_id": parameters["datasetId"],
        "content": {
            "status": "COMPLETED"
        },
        "agent": {
            "@type": "cat:extractor",
            "extractor_id":
            parameters['host'] + "/api/extractors/" + extractorName
        }
    }
    extractors.upload_dataset_metadata_jsonld(mdata=metadata,
                                              parameters=parameters)
def process_dataset(parameters):
    metafile, img_left, img_right, metadata = None, None, None, None

    # Get left/right files and metadata
    for f in parameters['files']:
        # First check metadata attached to dataset in Clowder for item of interest
        if f.endswith('_dataset_metadata.json'):
            all_dsmd = bin2tiff.load_json(f)
            for curr_dsmd in all_dsmd:
                if 'content' in curr_dsmd and 'lemnatec_measurement_metadata' in curr_dsmd['content']:
                    metafile = f
                    metadata = curr_dsmd['content']
        # Otherwise, check if metadata was uploaded as a .json file
        elif f.endswith('_metadata.json') and f.find('/_metadata.json') == -1 and metafile is None:
            metafile = f
            metadata = bin2tiff.load_json(metafile)
        elif f.endswith('_left.bin'):
            img_left = f
        elif f.endswith('_right.bin'):
            img_right = f
    if None in [metafile, img_left, img_right, metadata]:
        bin2tiff.fail('Could not find all of left/right/metadata.')
        return

    print("img_left: %s" % img_left)
    print("img_right: %s" % img_right)
    print("metafile: %s" % metafile)
    temp_out_dir = metafile.replace(os.path.basename(metafile), "")

    print("Determining image shapes")
    left_shape = bin2tiff.get_image_shape(metadata, 'left')
    right_shape = bin2tiff.get_image_shape(metadata, 'right')

    center_position = bin2tiff.get_position(metadata) # (x, y, z) in meters
    fov = bin2tiff.get_fov(metadata, center_position[2], left_shape) # (fov_x, fov_y) in meters; need to pass in the camera height to get correct fov
    left_position = [center_position[0]+bin2tiff.STEREO_OFFSET, center_position[1], center_position[2]]
    right_position = [center_position[0]-bin2tiff.STEREO_OFFSET, center_position[1], center_position[2]]
    left_gps_bounds = bin2tiff.get_bounding_box(left_position, fov) # (lat_max, lat_min, lng_max, lng_min) in decimal degrees
    right_gps_bounds = bin2tiff.get_bounding_box(right_position, fov)

    print("Creating demosaicked images")
    left_out = os.path.join(temp_out_dir, img_left[:-4] + '.jpg')
    left_image = bin2tiff.process_image(left_shape, img_left, left_out)
    right_out = os.path.join(temp_out_dir, img_right[:-4] + '.jpg')
    right_image = bin2tiff.process_image(right_shape, img_right, right_out)
    print("Uploading output JPGs to dataset")
    extractors.upload_file_to_dataset(left_out, parameters)
    extractors.upload_file_to_dataset(right_out, parameters)

    print("Creating geoTIFF images")
    left_tiff_out = os.path.join(temp_out_dir, img_left[:-4] + '.tif')
    bin2tiff.create_geotiff('left', left_image, left_gps_bounds, left_tiff_out)
    right_tiff_out = os.path.join(temp_out_dir, img_right[:-4] + '.tif')
    bin2tiff.create_geotiff('right', right_image, right_gps_bounds, right_tiff_out)
    print("Uploading output geoTIFFs to dataset")
    extractors.upload_file_to_dataset(left_tiff_out, parameters)
    extractors.upload_file_to_dataset(right_tiff_out, parameters)

    # Tell Clowder this is completed so subsequent file updates don't daisy-chain
    metadata = {
        "@context": {
            "@vocab": "https://clowder.ncsa.illinois.edu/clowder/assets/docs/api/index.html#!/files/uploadToDataset"
        },
        "dataset_id": parameters["datasetId"],
        "content": {"status": "COMPLETED"},
        "agent": {
            "@type": "cat:extractor",
            "extractor_id": parameters['host'] + "/api/extractors/" + extractorName
        }
    }
    extractors.upload_dataset_metadata_jsonld(mdata=metadata, parameters=parameters)