Spaces:
Running
Running
File size: 10,405 Bytes
4433b8c d5ea0f1 4433b8c d5ea0f1 4433b8c d5ea0f1 4433b8c d5ea0f1 4433b8c d5ea0f1 4433b8c d5ea0f1 4433b8c b0bdc40 4433b8c d5ea0f1 4433b8c d5ea0f1 4433b8c d5ea0f1 4433b8c d5ea0f1 4433b8c b0bdc40 4433b8c d5ea0f1 4433b8c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 |
import json
import os
from pathlib import Path
import shutil
import warnings
from PIL import Image
from dawsonia import io
from dawsonia import digitize
from dawsonia.ml import ml
import gradio as gr
import numpy as np
from numpy.typing import NDArray
import pandas as pd
import pooch
from .visualizer_functions import Page, TableCell
# Max number of images a user can upload at once
MAX_IMAGES = int(os.environ.get("MAX_IMAGES", 5))
# Setup the cache directory to point to the directory where the example images
# are located. The images must lay in the cache directory because otherwise they
# have to be reuploaded when drag-and-dropped to the input image widget.
GRADIO_CACHE = os.getenv("GRADIO_CACHE_DIR", ".gradio_cache")
DATA_CACHE = os.path.join(GRADIO_CACHE, "data")
EXAMPLES_DIRECTORY = os.path.join(os.getcwd(), "examples")
# Example books
PIPELINES: dict[str, dict[str, str]] = {
"bjuröklubb": dict(
url="https://git.smhi.se/ai-for-obs/data/-/raw/688c04f13e8e946962792fe4b4e0ded98800b154/raw_zarr/BJUR%C3%96KLUBB/DAGBOK_Bjur%C3%B6klubb_Station_Jan-Dec_1928.zarr.zip",
known_hash="sha256:6d87b7f79836ae6373cfab11260fe28787d93fe16199fefede6697ccd750f71a",
),
"härnösand": dict(
url="https://git.smhi.se/ai-for-obs/data/-/raw/688c04f13e8e946962792fe4b4e0ded98800b154/raw_zarr/H%C3%84RN%C3%96SAND/DAGBOK_H%C3%A4rn%C3%B6sand_Station_1934.zarr.zip",
known_hash="sha256:a58fdb6521214d0bd569c9325ce78d696738de28ce6ec869cde0d46616b697f2",
),
}
def run_dawsonia(
table_fmt_config_override,
first_page,
last_page,
prob_thresh,
book: io.Book,
book_path,
gallery,
progress=gr.Progress(),
):
if book is None:
raise ValueError("You need to select / upload the pages to digitize")
progress(0, desc="Dawsonia: starting")
model_path = Path("data/models/dawsonia/2024-07-02")
output_path = Path("output")
output_path.mkdir(exist_ok=True)
print("Dawsonia: digitizing", book)
table_fmt = book.table_format
final_output_path_book = output_path / book.station_name
output_path_book = Path(book_path).parent / "output"
output_path_book.mkdir(exist_ok=True, parents=True)
(output_path_book / "probablities").mkdir(exist_ok=True)
init_data: list[dict[str, NDArray]] = [
{
key: np.empty(len(table_fmt.rows), dtype="O")
for key in table_fmt.columns[table_idx]
}
for table_idx in table_fmt.preproc.idx_tables_size_verify
]
collection = []
images = []
with warnings.catch_warnings():
warnings.simplefilter("ignore", FutureWarning)
for page_number in range(first_page, last_page):
output_path_page = output_path_book / str(page_number)
gr.Info(f"Digitizing {page_number = }")
if (
not (output_path_book / str(page_number))
.with_suffix(".parquet")
.exists()
):
digitize.digitize_page_and_write_output(
book,
init_data,
page_number=page_number,
date_str=f"0000-page-{page_number}",
model_path=model_path,
model_predict=ml.model_predict,
prob_thresh=prob_thresh,
output_path_page=output_path_page,
output_text_fmt=False,
debug=False,
)
_synctree(output_path_book, final_output_path_book)
progress_value = (page_number - first_page) / max(1, last_page - first_page)
# if final_output_path_book.exists():
# shutil.rmtree(final_output_path_book)
# shutil.copytree(output_path_book, final_output_path_book)
for page_number, im_from_gallery in zip(range(first_page, last_page), gallery):
if results := read_page(
final_output_path_book,
str(page_number),
prob_thresh,
progress,
1.0,
table_fmt.preproc.idx_tables_size_verify,
): # , im_from_gallery[0])
page, im = results
collection.append(page)
images.append(im)
yield collection, gr.skip()
else:
gr.Info(f"No tables detected in {page_number = }")
gr.Info("Pages were succesfully digitized ✨")
# yield collection, images
yield collection, gr.skip()
def _synctree(source_dir, dest_dir):
source_dir = Path(source_dir)
dest_dir = Path(dest_dir)
if not dest_dir.exists():
dest_dir.mkdir(parents=True)
for root, _, files in os.walk(source_dir):
root = Path(root)
relative_root = root.relative_to(source_dir)
# Create subdirectories in the destination directory
dest_subdir_path = dest_dir / relative_root
if not dest_subdir_path.exists():
dest_subdir_path.mkdir(parents=True, exist_ok=True)
for file_ in files:
source_file_path = root / file_
dest_file_path = dest_subdir_path / file_
# Copy only if the file does not already exist or is newer
if (
not dest_file_path.exists()
or (source_file_path.stat().st_mtime - dest_file_path.stat().st_mtime) > 0
):
shutil.copy2(source_file_path, dest_file_path)
def read_page(
output_path_book: Path,
prefix: str,
prob_thresh: float,
progress,
progress_value,
idx_tables_size_verify: list[int],
im_path_from_gallery: str = "",
):
stats = digitize.Statistics.from_json(
(output_path_book / "statistics" / prefix).with_suffix(".json")
)
print(stats)
progress(progress_value, desc=f"Dawsonia: {stats!s:.50}")
if stats.tables_detected > 0:
values_df = pd.read_parquet((output_path_book / prefix).with_suffix(".parquet"))
prob_df = pd.read_parquet(
(output_path_book / "probablities" / prefix).with_suffix(".parquet")
)
table_meta = json.loads(
(output_path_book / "table_meta" / prefix).with_suffix(".json").read_text()
)
with Image.open(
image_path := (output_path_book / "pages" / prefix).with_suffix(".webp")
) as im:
width = im.width
height = im.height
values_array = values_df.values.flatten()
prob_array = prob_df.values.flatten()
# FIXME: hardcoded to get upto 2 tables. Use idx_tables_size_verify and reconstruct bbox_array
try:
bbox_array = np.hstack(table_meta["table_positions"][:2]).reshape(-1, 4)
except ValueError:
bbox_array = np.reshape(table_meta["table_positions"][0], (-1, 4))
cells = [
make_cell(value, bbox)
for value, prob, bbox in zip(values_array, prob_array, bbox_array)
if prob > prob_thresh
]
return Page(width, height, cells, im_path_from_gallery or str(image_path)), im
def make_cell(value: str, bbox: NDArray[np.int64]):
y, x, h, w = bbox
xmin, ymin = x - w // 2, y - h // 2
xmax, ymax = x + w // 2, y + h // 2
polygon = (xmin, ymin), (xmax, ymin), (xmax, ymax), (xmin, ymax), (xmin, ymin)
return TableCell(polygon, text_x=x - w // 4, text_y=y, text=value)
def all_example_images() -> list[str]:
"""
Get paths to all example images.
"""
examples = [
os.path.join(EXAMPLES_DIRECTORY, f"{pipeline}.png") for pipeline in PIPELINES
]
return examples
def get_selected_example_image(
first_page, last_page, event: gr.SelectData
) -> tuple[list[Image.Image], io.Book, str, str, str] | None:
"""
Get the name of the pipeline that corresponds to the selected image.
"""
orig_name = event.value["image"]["orig_name"]
# for name, details in PIPELINES.items():
orig_path = Path(orig_name)
name = orig_path.name
for suffix in orig_path.suffixes[::-1]:
name = name.removesuffix(suffix)
station_tf = Path("table_formats", name).with_suffix(".toml")
if (last_page - first_page) > MAX_IMAGES:
error = f"Maximum images you can digitize is set to: {MAX_IMAGES}"
gr.Warning(error)
raise ValueError(error)
if name in PIPELINES:
book_path = pooch.retrieve(**PIPELINES[name], path=DATA_CACHE)
first, last, book = io.read_book(book_path)
book._name = name
book.size_cell = [1.0, 1.0, 1.0, 1.0]
return (
[book.read_image(pg) for pg in range(first_page, last_page)],
book,
book_path,
station_tf.name,
station_tf.read_text(),
)
def move_uploaded_file(uploaded, table_fmt_filename):
current_directory = Path(uploaded).parent
# Define the target directory where you want to save the uploaded files
target_directory = current_directory / table_fmt_filename.removesuffix(".toml")
os.makedirs(target_directory, exist_ok=True)
# Move the uploaded file to the target directory
true_path = Path(target_directory / Path(uploaded).name)
# if true_path.exists():
# true_path.unlink()
shutil.copy2(uploaded, true_path)
print(f"Copy created", true_path)
return str(true_path)
def get_uploaded_image(
first_page: int, last_page: int, table_fmt_filename: str, filename: str
) -> tuple[list[NDArray], io.Book, str, str] | None:
orig_path = Path(filename)
name = orig_path.name
for suffix in orig_path.suffixes[::-1]:
name = name.removesuffix(suffix)
station_tf = Path("table_formats", table_fmt_filename)
if not station_tf.exists():
station_tf = Path("table_formats", "bjuröklubb.toml")
first, last, book = io.read_book(Path(filename))
book._name = name
book.size_cell = [1.0, 1.0, 1.0, 1.0]
return (
[book.read_page(pg) for pg in range(first_page, last_page)],
book,
filename,
station_tf.read_text(),
)
def overwrite_table_format_file(book: io.Book, book_path, table_fmt: str):
name = book.station_name
table_fmt_dir = Path("table_formats")
(table_fmt_dir / name).with_suffix(".toml").write_text(table_fmt)
book.table_format = io.read_specific_table_format(table_fmt_dir, Path(book_path))
gr.Info(f"Overwritten table format file for {name}")
return book
|