layoutlm_resume_data.py

# coding=utf-8

import json
import os

from PIL import Image

import datasets

def load_image(image_path):
    image = Image.open(image_path).convert("RGB")
    w, h = image.size
    return image, (w, h)

def normalize_bbox(bbox, size):
    return [
        int(1000 * bbox[0] / size[0]),
        int(1000 * bbox[1] / size[1]),
        int(1000 * bbox[2] / size[0]),
        int(1000 * bbox[3] / size[1]),
    ]

logger = datasets.logging.get_logger(__name__)




class ResumeDataConfig(datasets.BuilderConfig):
    """BuilderConfig for Resume NER"""

    def __init__(self, **kwargs):
        """BuilderConfig for FUNSD.
        Args:
          **kwargs: keyword arguments forwarded to super.
        """
        super(ResumeDataConfig, self).__init__(**kwargs)


class ResumeData(datasets.GeneratorBasedBuilder):

    BUILDER_CONFIGS = [
        ResumeDataConfig(name="funsd",
                         version=datasets.Version("1.0.0"),
                         description="Resume Dataset"),
    ]

    def _info(self):
        return datasets.DatasetInfo(
            description="",
            features=datasets.Features(
                {
                    "id": datasets.Value("string"),
                    "tokens": datasets.Sequence(datasets.Value("string")),
                    "bboxes": datasets.Sequence(datasets.Sequence(datasets.Value("int64"))),
                    "ner_tags": datasets.Sequence(
                        datasets.features.ClassLabel(
                            names=["O",
                                   "B-ADDRESS",
                                    "B-EMAIL",
                                    "B-NAME",
                                    "B-PHONE",
                                    "B-SECTIONHEADER",
                                    "E-ADDRESS",
                                    "E-EMAIL",
                                    "E-NAME",
                                    "E-PHONE",
                                    "E-SECTIONHEADER",
                                    "I-ADDRESS",
                                    "I-EMAIL",
                                    "I-NAME",
                                    "I-PHONE",
                                    "I-SECTIONHEADER",
                                    "S-ADDRESS",
                                    "S-EMAIL",
                                    "S-NAME",
                                    "S-PHONE",
                                    "S-SECTIONHEADER"
                                   ]
                        )
                    ),
                    "image": datasets.features.Image(),
                }
            ),
            supervised_keys=None,
            homepage="",
            citation="",
        )

    def _split_generators(self, dl_manager):
        """Returns SplitGenerators."""
        downloaded_file = dl_manager.download_and_extract("https://huggingface.co/datasets/Kunling/layoutlm_resume_data/resolve/main/person_resume_funsd_format_v5.zip")
        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN, gen_kwargs={"filepath": f"{downloaded_file}/dataset/training/"}
            ),
            datasets.SplitGenerator(
                name=datasets.Split.TEST, gen_kwargs={"filepath": f"{downloaded_file}/dataset/testing/"}
            ),
        ]

    def get_line_bbox(self, bboxs):
        x = [bboxs[i][j] for i in range(len(bboxs)) for j in range(0, len(bboxs[i]), 2)]
        y = [bboxs[i][j] for i in range(len(bboxs)) for j in range(1, len(bboxs[i]), 2)]

        x0, y0, x1, y1 = min(x), min(y), max(x), max(y)

        assert x1 >= x0 and y1 >= y0
        bbox = [[x0, y0, x1, y1] for _ in range(len(bboxs))]
        return bbox

    def _generate_examples(self, filepath):
        logger.info("⏳ Generating examples from = %s", filepath)
        ann_dir = os.path.join(filepath, "annotations")
        img_dir = os.path.join(filepath, "images")
        for guid, file in enumerate(sorted(os.listdir(ann_dir))):
            tokens = []
            bboxes = []
            ner_tags = []

            file_path = os.path.join(ann_dir, file)
            with open(file_path, "r", encoding="utf8") as f:
                data = json.load(f)
            image_path = os.path.join(img_dir, file)
            image_path = image_path.replace("json", "jpeg")
            image, size = load_image(image_path)
            for item in data["form"]:
                cur_line_bboxes = []
                words, label = item["words"], item["label"]
                words = [w for w in words if w["text"].strip() != ""]
                if len(words) == 0:
                    continue
                if label.lower() == "other":
                    for w in words:
                        tokens.append(w["text"])
                        ner_tags.append("O")
                        cur_line_bboxes.append(normalize_bbox(w["box"], size))
                else:
                    tokens.append(words[0]["text"])
                    ner_tags.append("B-" + label.upper())
                    cur_line_bboxes.append(normalize_bbox(words[0]["box"], size))
                    for w in words[1:]:
                        tokens.append(w["text"])
                        ner_tags.append("I-" + label.upper())
                        cur_line_bboxes.append(normalize_bbox(w["box"], size))
                cur_line_bboxes = self.get_line_bbox(cur_line_bboxes)
                bboxes.extend(cur_line_bboxes)
            yield guid, {"id": str(guid), "tokens": tokens, "bboxes": bboxes, "ner_tags": ner_tags,
                         "image": image}