mapreader 1.6.3.post0.dev19__tar.gz → 1.7.0.post0.dev1__tar.gz

{mapreader-1.6.3.post0.dev19 → mapreader-1.7.0.post0.dev1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mapreader
-Version: 1.6.3.post0.dev19
+Version: 1.7.0.post0.dev1
 Summary: A computer vision pipeline for the semantic exploration of maps/images at scale
 Home-page: https://github.com/Living-with-machines/MapReader
 Download-URL: https://github.com/Living-with-machines/MapReader/archive/refs/heads/main.zip
@@ -62,6 +62,7 @@ Requires-Dist: opencv-python<5.0.0.0,>=4.8.1.78
 Requires-Dist: folium<1.0.0,>=0.12
 Requires-Dist: mapclassify<3.0.0,>=2.0.0
 Requires-Dist: xyzservices==2024.9.0
+Requires-Dist: lxml
 Provides-Extra: dev
 Requires-Dist: pytest<9.0.0; extra == "dev"
 Requires-Dist: pytest-cov<6.0.0,>=4.1.0; extra == "dev"

{mapreader-1.6.3.post0.dev19 → mapreader-1.7.0.post0.dev1}/mapreader/__init__.py

@@ -1,5 +1,10 @@
 # download
 from mapreader.download.sheet_downloader import SheetDownloader
+try:
+    from mapreader.download.iiif_downloader import IIIFDownloader
+except ImportError:
+    print("[WARNING] Piffle not installed, please install it from the `iiif_dataclasses` branch of https://github.com/rwood-97/piffle/tree/iiif_dataclasses")
+    pass
 from mapreader.download.downloader import Downloader
 from mapreader.download.downloader_utils import create_polygon_from_latlons, create_line_from_latlons
 
@@ -52,4 +57,4 @@ import mapreader
 
 def print_version():
     """Print the current version of mapreader."""
-    print(mapreader.__version__)
+    print(mapreader.__version__)

{mapreader-1.6.3.post0.dev19 → mapreader-1.7.0.post0.dev1}/mapreader/_version.py

@@ -8,11 +8,11 @@ import json
 
 version_json = '''
 {
- "date": "2025-03-24T08:13:15+0000",
+ "date": "2025-04-10T09:06:03+0100",
  "dirty": false,
  "error": null,
- "full-revisionid": "6a1154eaa44a30358fa4146cac49af242e1b51b0",
- "version": "1.6.3.post0.dev19"
+ "full-revisionid": "f7bfc0a12e8f33f0a79343b15bf2c1f8282ff294",
+ "version": "1.7.0.post0.dev1"
 }
 '''  # END VERSION_JSON
 

mapreader-1.7.0.post0.dev1/mapreader/download/iiif_downloader.py

@@ -0,0 +1,402 @@
+from __future__ import annotations
+
+import io
+import os
+import pathlib
+
+import geopandas as gpd
+import numpy as np
+import pandas as pd
+import rasterio
+import rasterio.mask
+import requests
+from bs4 import BeautifulSoup
+from piffle.iiif_dataclasses.presentation2 import IIIFPresentation2
+from piffle.iiif_dataclasses.presentation3 import (
+    GeoreferenceAnnotation3,
+    IIIFPresentation3,
+)
+from piffle.load_iiif import load_iiif_image, load_iiif_presentation
+from PIL import Image, ImageDraw
+from rasterio import transform
+from rasterio.plot import reshape_as_raster
+from shapely import Polygon
+from tqdm.auto import tqdm
+
+
+class IIIFDownloader:
+    def __init__(
+        self,
+        iiif: str
+        | IIIFPresentation3
+        | IIIFPresentation2
+        | list[str | IIIFPresentation3 | IIIFPresentation2],
+        iiif_versions: int | float | str | list[int | float | str] | None = None,
+        iiif_uris: str | list[str] | None = None,
+    ):
+        """A class to download images from IIIF presentation jsons.
+
+        Parameters
+        ----------
+        iiif : str | IIIFPresentation3 | IIIFPresentation2 | list[str  |  IIIFPresentation3  |  IIIFPresentation2]
+            The IIIF url or IIIF presentation API object(s) containing the maps to download
+        iiif_versions : int | float | str | list[int | float | str] | None
+            The IIIF version(s) of the IIIF object(s). Can only be None if `iiif` is/are already IIIF object(s). Default is None.
+        iiif_uris : str | list[str] | None
+            The URI(s) of the IIIF object(s), needed if the IIIF object is missing an 'id' field. Default is None.
+
+            If only one IIIF object is missing an `id`, create a list with None values for the other objects.
+            e.g. iiif_uri=[None, "https://example.com/iiif/manifest.json", None]
+        """
+        if not isinstance(iiif, list):
+            iiif = [iiif]
+        if iiif_versions is None:
+            iiif_versions = [None] * len(iiif)
+        if not isinstance(iiif_versions, list):
+            iiif_versions = [iiif_versions]
+        if len(iiif) != len(iiif_versions):
+            raise ValueError(
+                "[ERROR] Length of `iiif` must match length of `iiif_versions`."
+            )
+
+        iiif_objs = []
+        for iiif_obj, iiif_version in zip(iiif, iiif_versions):
+            if isinstance(iiif_obj, str):
+                iiif_obj = load_iiif_presentation(iiif_obj, iiif_version)
+            if isinstance(iiif_obj, (IIIFPresentation3, IIIFPresentation2)):
+                iiif_objs.append(iiif_obj)
+            else:
+                raise ValueError("[ERROR] Each `iiif` must be a string or IIIF object.")
+
+        if iiif_uris is None:
+            iiif_uris = [None] * len(iiif_objs)
+        if not isinstance(iiif_uris, list):
+            iiif_uris = [iiif_uris]
+        if len(iiif_uris) != len(iiif_objs):
+            raise ValueError(
+                "[ERROR] Length of `iiif_uris` must match length of `iiif`."
+            )
+
+        for iiif_obj, iiif_uri in zip(iiif_objs, iiif_uris):
+            if not hasattr(iiif_obj, "id"):
+                if iiif_uri is not None:
+                    iiif_obj.id = iiif_uri
+                else:
+                    raise ValueError(
+                        "[ERROR] One of your IIIF objects is missing an 'id' field so we cannot identify it's URL. Please pass the `iiif_uris` argument."
+                    )
+
+        self.iiif = iiif_objs
+
+    def save_georeferenced_maps(
+        self,
+        path_save: str | pathlib.Path = "maps",
+    ):
+        """Save georefereced maps from a IIIF presentation 3 json as tiffs.
+
+        Parameters
+        ----------
+        path_save : str | pathlib.Path
+            Path to save the images
+
+        Notes
+        -----
+        Calls the `save_georeferenced_map` method for each IIIF object.
+        """
+        for iiif in self.iiif:
+            self.save_georeferenced_map(iiif, path_save=path_save)
+
+    def save_georeferenced_map(
+        self,
+        iiif: str | IIIFPresentation3,
+        path_save: str | pathlib.Path = "maps",
+        iiif_uri: str = None,
+    ):
+        """Save a single georefereced map from a IIIF presentation 3 json as a tiff.
+
+        Parameters
+        ----------
+        iiif : str | dict | IIIFPresentation3
+            the IIIF url or IIIF presentation API object containing the map(s)
+        path_save : str | pathlib.Path
+            Path to save the images
+        iiif_uri : str
+            The URI of the IIIF object, needed if the IIIF object is missing an 'id' field.
+
+        Notes
+        -----
+        Only first order polynomial transformations are currently supported, see the [allmaps documentation](https://allmaps.org/) for more information on the transformation options.
+
+        Raises
+        ------
+        ValueError
+            Only first order polynomial transformations are currently supported.
+        """
+        if isinstance(iiif, str):
+            iiif_obj = load_iiif_presentation(iiif, 3)
+        elif isinstance(iiif, IIIFPresentation3):
+            iiif_obj = iiif
+        else:
+            raise ValueError("[ERROR] `iiif` must be a string or IIIF3 object.")
+
+        if not hasattr(iiif_obj, "id"):
+            if iiif_uri is not None:
+                iiif_obj.id = iiif_uri
+            else:
+                raise ValueError(
+                    "[ERROR] IIIF object is missing an 'id' field so we cannot identify its URI. Please manually pass the `iiif_uri` argument."
+                )
+        iiif_uri = iiif_obj.id
+
+        metadata = pd.DataFrame(columns=["name", "id", "iiif_uri"])
+        for annot in tqdm(iiif_obj.collect_annotations()):
+            if not isinstance(annot, GeoreferenceAnnotation3):
+                print(
+                    f"[WARNING] Skipping annotation '{annot.id}' as it is not a georeference annotation."
+                )
+                continue
+
+            # Check we are working with first order polynomial transformation (i.e. affine)
+            if (annot.body["transformation"]["type"] != "polynomial") or (
+                annot.body["transformation"]["options"]["order"] != 1
+            ):
+                raise ValueError(
+                    "[ERROR] Only first order polynomial transformations are currently supported"
+                )
+
+            # Get filename
+            # host, prefix, identifier
+            id = annot.resource["service"]["id"] if annot.id is None else annot.id
+            fname = ".".join(id.removeprefix("https://").split("/")[2:])
+
+            if not os.path.exists(path_save):
+                os.makedirs(path_save, exist_ok=True)
+            if os.path.exists(f"{path_save}/{fname}.tif"):
+                print(f"[INFO] '{fname}' already exists. Skipping download.")
+                continue
+
+            # Download image
+            image = self.download_image(3, annot)
+
+            # Get GCPs
+            features = gpd.GeoDataFrame.from_features(
+                annot.body["features"], crs="WGS84"
+            )
+            resource_points = features[
+                "resourceCoords"
+            ].to_list()  # in image coordinates
+            gcp_points = features["geometry"].to_list()  # in WGS84
+
+            # resource points are x, y (i.e col, row)
+            # gcp points are lon, lat (i.e x, y)
+
+            # Create affine transform
+            affine = transform.from_gcps(
+                [
+                    transform.GroundControlPoint(
+                        resource_point[1], resource_point[0], gcp_point.x, gcp_point.y
+                    )
+                    for resource_point, gcp_point in zip(resource_points, gcp_points)
+                ]
+            )
+
+            # Idenfity annotation bounds
+            svg = BeautifulSoup(annot.target["selector"]["value"], "lxml").find("svg")
+            points = svg.find("polygon")["points"]
+            points = np.array(
+                [
+                    [int(x), int(y)]
+                    for x, y in [point.split(",") for point in points.split()]
+                ]
+            )  # points are x, y (i.e col, row)
+
+            # Apply affine transform to annotation bounds
+            t = transform.AffineTransformer(affine)
+            x1, y1 = t.xy(*points[1][::-1])  # input as row, col (i.e y, x)
+            x0, y0 = t.xy(*points[0][::-1])
+            x2, y2 = t.xy(*points[2][::-1])
+            x3, y3 = t.xy(*points[3][::-1])
+
+            # Save tiff
+            height, width, channels = image.height, image.width, len(image.getbands())
+
+            with rasterio.open(
+                f"{path_save}/{fname}.tif",
+                "w",
+                driver="GTiff",
+                height=height,
+                width=width,
+                count=channels,
+                transform=affine,
+                dtype="uint8",
+                nodata=0,
+                crs="WGS84",
+            ) as dst:
+                array = reshape_as_raster(
+                    image
+                )  # (row, col, bands) to (bands, row, col)
+                dst.write(array)
+
+            # Mask tiff to annotation bounds
+            tiff_image = rasterio.open(f"{path_save}/{fname}.tif")
+
+            polygon = Polygon([(x0, y0), (x1, y1), (x2, y2), (x3, y3)])
+            masked_image, out_transform = rasterio.mask.mask(
+                tiff_image, [polygon], crop=False
+            )
+            out_meta = tiff_image.meta.copy()
+
+            out_meta.update(
+                {
+                    "driver": "GTiff",
+                    "height": masked_image.shape[1],
+                    "width": masked_image.shape[2],
+                    "transform": out_transform,
+                }
+            )
+
+            with rasterio.open(
+                f"{path_save}/{fname}_masked.tif", "w", **out_meta
+            ) as dst:
+                dst.write(masked_image)
+
+            metadata.loc[len(metadata)] = [f"{fname}_masked.tif", fname, iiif_obj.id]
+            metadata.to_csv(f"{path_save}/metadata.csv")
+
+    def save_maps(
+        self,
+        path_save: str | pathlib.Path = "maps",
+    ):
+        """Save maps from a IIIF presentation json as pngs.
+
+        Parameters
+        ----------
+        path_save : str | pathlib.Path
+            Path to save the images
+
+        Notes
+        -----
+        Calls the `save_map` method for each IIIF object.
+        """
+        for iiif in self.iiif:
+            self.save_map(iiif, path_save=path_save)
+
+    def save_map(
+        self,
+        iiif: str | IIIFPresentation3 | IIIFPresentation2,
+        iiif_version: int | float | str = 3,
+        path_save: str | pathlib.Path = "maps",
+        iiif_uri: str = None,
+    ):
+        """Save a single map from a IIIF presentation json as a png.
+
+        Parameters
+        ----------
+        iiif : str | dict | IIIFPresentation3 | IIIFPresentation2
+            the IIIF url or IIIF presentation API object containing the map(s)
+        iiif_version : int | float | str
+            The IIIF version, ignored if iiif is already a IIIF object
+        path_save : str | pathlib.Path
+            Path to save the images
+        iiif_uri : str
+            The URI of the IIIF object, needed if the IIIF object is missing an 'id' field.
+        """
+        if isinstance(iiif, str):
+            iiif_obj = load_iiif_presentation(iiif, iiif_version)
+        elif isinstance(iiif, (IIIFPresentation3, IIIFPresentation2)):
+            iiif_obj = iiif
+        else:
+            raise ValueError("`iiif` must be a string or IIIF object.")
+
+        if not hasattr(iiif_obj, "id"):
+            if iiif_uri is not None:
+                iiif_obj.id = iiif_uri
+            else:
+                raise ValueError(
+                    "[ERROR] IIIF object is missing an 'id' field so we cannot identify its URL. Please manually pass the `iiif_url` argument."
+                )
+        iiif_uri = iiif_obj.id
+
+        metadata = pd.DataFrame(columns=["filename", "iiif_uri"])
+        for annot in tqdm(iiif_obj.collect_annotations()):
+            # Get filename
+            # host, prefix, identifier
+            id = annot.resource["service"]["id"] if annot.id is None else annot.id
+            fname = ".".join(id.removeprefix("https://").split("/")[2:])
+
+            if not os.path.exists(path_save):
+                os.makedirs(path_save, exist_ok=True)
+            if os.path.exists(f"{path_save}/{fname}.png"):
+                print(f"[INFO] '{fname}' already exists. Skipping download.")
+                continue
+
+            metadata.loc[len(metadata)] = [fname, iiif_obj.id]
+
+            # Download image
+            image = self.download_image(iiif_version, annot)
+            image.save(f"{path_save}/{fname}.png")
+
+            if isinstance(iiif_obj, IIIFPresentation3):
+                if "selector" in annot.target.keys():
+                    # Idenfity annotation bounds
+                    svg = BeautifulSoup(annot.target["selector"]["value"], "xml").find(
+                        "svg"
+                    )
+                    points = svg.find("polygon")["points"]
+                    points = np.array(
+                        [
+                            [int(x), int(y)]
+                            for x, y in [point.split(",") for point in points.split()]
+                        ]
+                    )  # points are x, y (i.e col, row)
+
+                    # -- Mask image to annotation bounds
+                    # Convert to numpy array
+                    image_array = np.array(image)
+
+                    # Create mask
+                    polygon = Polygon(points)
+                    mask_image = Image.new(
+                        "L", (image_array.shape[1], image_array.shape[0]), 0
+                    )
+                    ImageDraw.Draw(mask_image).polygon(
+                        polygon.exterior.coords, outline=1, fill=1
+                    )
+                    mask = np.array(mask_image)
+
+                    # Apply mask
+                    image_array = image_array * mask[:, :, None]
+
+                    masked_image = Image.fromarray(image_array)
+                    masked_image.save(f"{path_save}/{fname}_masked.png")
+
+            metadata.to_csv(f"{path_save}/metadata.csv")
+
+    def download_image(iiif, iiif_version, annot):
+        image_url = annot.get_image_url()
+        image_manifest = load_iiif_image(
+            image_url, iiif_version
+        )  # TODO: method to get image version
+
+        width = image_manifest.width
+        height = image_manifest.height
+        tile_width = image_manifest.tiles[0]["width"]
+        tile_height = image_manifest.tiles[0].get("height", tile_width)
+
+        image = Image.new("RGB", (width, height))
+        for x in np.arange(0, ((width // tile_width) + 1) * tile_width, tile_width):
+            for y in np.arange(
+                0, ((height // tile_height) + 1) * tile_height, tile_height
+            ):
+                if x + tile_width > width:
+                    x = width - tile_width
+                if y + tile_height > height:
+                    y = height - tile_height
+                response = requests.get(
+                    f"{image_url}/{x},{y},512,512/full/0/default.jpg"
+                )
+                response.raise_for_status()
+
+                tile = Image.open(io.BytesIO(response.content))
+                image.paste(tile, (x, y))
+        return image

{mapreader-1.6.3.post0.dev19 → mapreader-1.7.0.post0.dev1}/mapreader.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mapreader
-Version: 1.6.3.post0.dev19
+Version: 1.7.0.post0.dev1
 Summary: A computer vision pipeline for the semantic exploration of maps/images at scale
 Home-page: https://github.com/Living-with-machines/MapReader
 Download-URL: https://github.com/Living-with-machines/MapReader/archive/refs/heads/main.zip
@@ -62,6 +62,7 @@ Requires-Dist: opencv-python<5.0.0.0,>=4.8.1.78
 Requires-Dist: folium<1.0.0,>=0.12
 Requires-Dist: mapclassify<3.0.0,>=2.0.0
 Requires-Dist: xyzservices==2024.9.0
+Requires-Dist: lxml
 Provides-Extra: dev
 Requires-Dist: pytest<9.0.0; extra == "dev"
 Requires-Dist: pytest-cov<6.0.0,>=4.1.0; extra == "dev"

{mapreader-1.6.3.post0.dev19 → mapreader-1.7.0.post0.dev1}/mapreader.egg-info/SOURCES.txt

@@ -24,6 +24,7 @@ mapreader/download/__init__.py
 mapreader/download/data_structures.py
 mapreader/download/downloader.py
 mapreader/download/downloader_utils.py
+mapreader/download/iiif_downloader.py
 mapreader/download/sheet_downloader.py
 mapreader/download/tile_loading.py
 mapreader/download/tile_merging.py

{mapreader-1.6.3.post0.dev19 → mapreader-1.7.0.post0.dev1}/mapreader.egg-info/requires.txt

@@ -33,6 +33,7 @@ opencv-python<5.0.0.0,>=4.8.1.78
 folium<1.0.0,>=0.12
 mapclassify<3.0.0,>=2.0.0
 xyzservices==2024.9.0
+lxml
 
 [dev]
 pytest<9.0.0

{mapreader-1.6.3.post0.dev19 → mapreader-1.7.0.post0.dev1}/setup.py

@@ -63,6 +63,8 @@ setuptools.setup(
         "folium>=0.12,<1.0.0",
         "mapclassify>=2.0.0,<3.0.0",
         "xyzservices==2024.9.0",
+#         "piffle @ git+https://github.com/rwood-97/piffle.git@iiif_dataclasses",
+        "lxml",
     ],
     extras_require={
         "dev": [

{mapreader-1.6.3.post0.dev19 → mapreader-1.7.0.post0.dev1}/tests/test_import.py

@@ -12,4 +12,5 @@ def test_import():
         PatchContextDataset,
         Downloader,
         SheetDownloader,
-    )
+        # IIIFDownloader,
+    )