make initial commit

app.py

@@ -0,0 +1,303 @@
+import json
+
+import streamlit as st
+from datasets import load_dataset
+from streamlit_folium import folium_static
+
+from catalogue import make_choro_map, region_tree
+
+##################
+## streamlit
+##################
+st.set_page_config(
+    page_title="BigScience Language Resource Catalogue Input Form",
+    page_icon="https://avatars.githubusercontent.com/u/82455566",
+    layout="wide",
+    initial_sidebar_state="auto",
+)
+
+query_params = st.experimental_get_query_params()
+
+
+def main():
+    if "save_state" not in st.session_state:
+        st.session_state.save_state = {}
+
+    viz_page()
+
+
+##################
+## SECTION: Explore the current catalogue
+##################
+
+app_categories = {
+    "entry_types": {
+        "primary": "Primary source",
+        "processed": "Processed language dataset",
+        "organization": "Language organization or advocate",
+    },
+    "language_lists": json.load(
+        open("resources/language_lists.json", encoding="utf-8")
+    ),
+    "programming_languages": [
+        x
+        for x in json.load(
+            open("resources/programming_languages.json", encoding="utf-8")
+        )["itemListElement"]
+    ],
+    "languages_bcp47": [
+        x
+        for x in json.load(open("resources/bcp47.json", encoding="utf-8"))["subtags"]
+        if x["type"] == "language"
+    ],
+    "custodian_types": [
+        "A private individual",
+        "A commercial entity",
+        "A library, museum, or archival institute",
+        "A university or research institution",
+        "A nonprofit/NGO (other)",
+        "A government organization",
+    ],
+    "pii_categories": json.load(
+        open("resources/pii_categories.json", encoding="utf-8")
+    ),
+    "licenses": json.load(open("resources/licenses.json", encoding="utf-8")),
+    "primary_taxonomy": json.load(
+        open("resources/primary_source_taxonomy.json", encoding="utf-8")
+    ),
+    "file_formats": json.load(open("resources/file_formats.json", encoding="utf-8")),
+}
+
+
+def filter_entry(entry, filter_dct):
+    res = True
+    for k, v in entry.items():
+        if k in filter_dct:
+            if isinstance(v, dict):
+                res = res and filter_entry(v, filter_dct[k])
+            elif isinstance(v, list):
+                res = res and (
+                    len(filter_dct[k]) == 0 or any([e in filter_dct[k] for e in v])
+                )
+            else:
+                res = res and (len(filter_dct[k]) == 0 or v in filter_dct[k])
+    return res
+
+
+def filter_catalogue_visualization(catalogue, options):
+    st.markdown("### Select entries to visualize")
+    st.markdown(
+        "##### Select entries by category, language, type of custodian or media"
+    )
+    st.markdown(
+        "You can select specific parts of the catalogue to visualize in this window."
+        + " Leave a field empty to select all values, or select specific options to only select entries that have one of the chosen values."
+    )
+    filter_by_options = [
+        "resource type",
+        "language names",
+        "custodian type",
+        "available for download",
+        "license type",
+        "source type",
+        "media type",
+    ]
+    filter_by = st.multiselect(
+        key="viz_filter_by",
+        label="You can filter the catalogue to only visualize entries that have certain properties, such as:",
+        options=filter_by_options,
+    )
+    filter_dict = {}
+    if "resource type" in filter_by:
+        filter_dict["type"] = st.multiselect(
+            key="viz_filter_type",
+            label="I want to only see entries that are of the following category:",
+            options=options["entry_types"],
+            format_func=lambda x: options["entry_types"][x],
+        )
+    if "language names" in filter_by:
+        filter_dict["languages"] = {}
+        filter_dict["languages"]["language_names"] = st.multiselect(
+            key="viz_filter_languages_language_names",
+            label="I want to only see entries that have one of the following languages:",
+            options=list(options["language_lists"]["language_groups"].keys())
+            + options["language_lists"]["niger_congo_languages"]
+            + options["language_lists"]["indic_languages"],
+        )
+    if "custodian type" in filter_by:
+        filter_dict["custodian"] = {}
+        filter_dict["custodian"]["type"] = st.multiselect(
+            key="viz_filter_custodian_type",
+            label="I want to only see entries that corresponds to organizations or to data that id owned/managed by organizations of the following types:",
+            options=options["custodian_types"],
+        )
+    if "available for download" in filter_by:
+        filter_dict["availability"] = filter_dict.get("availability", {})
+        filter_dict["availability"]["procurement"] = {}
+        download_options = [
+            "No - but the current owners/custodians have contact information for data queries",
+            "No - we would need to spontaneously reach out to the current owners/custodians",
+            "Yes - it has a direct download link or links",
+            "Yes - after signing a user agreement",
+        ]
+        filter_dict["availability"]["procurement"]["for_download"] = st.multiselect(
+            key="viz_availability_procurement_for_download",
+            label="Select based on whether the data can be obtained online:",
+            options=download_options,
+        )
+    if "license type" in filter_by:
+        filter_dict["availability"] = filter_dict.get("availability", {})
+        filter_dict["availability"]["licensing"] = {}
+        filter_dict["availability"]["licensing"]["license_properties"] = st.multiselect(
+            key="viz_availability_licensing_license_properties",
+            label="Select primary entries that have the following license types",
+            options=[
+                "public domain",
+                "multiple licenses",
+                "copyright - all rights reserved",
+                "open license",
+                "research use",
+                "non-commercial use",
+                "do not distribute",
+            ],
+        )
+        primary_license_options = [
+            "Unclear / I don't know",
+            "Yes - the source material has an open license that allows re-use",
+            "Yes - the dataset has the same license as the source material",
+            "Yes - the dataset curators have obtained consent from the source material owners",
+            "No - the license of the source material actually prohibits re-use in this manner",
+        ]
+        filter_dict["processed_from_primary"] = filter_dict.get(
+            "processed_from_primary", {}
+        )
+        filter_dict["processed_from_primary"]["primary_license"] = st.multiselect(
+            key="viz_processed_from_primary_primary_license",
+            label="For datasets, selected based on: Is the license or commercial status of the source material compatible with the license of the dataset?",
+            options=primary_license_options,
+        )
+    if "source type" in filter_by:
+        filter_dict["source_category"] = {}
+        filter_dict["source_category"]["category_type"] = st.multiselect(
+            key="viz_source_category_category_type",
+            label="Select primary sources that correspond to:",
+            options=["collection", "website"],
+        )
+        filter_dict["source_category"]["category_web"] = st.multiselect(
+            key="viz_source_category_category_web",
+            label="Select web-based primary sources that contain:",
+            options=options["primary_taxonomy"]["website"],
+        )
+        filter_dict["source_category"]["category_media"] = st.multiselect(
+            key="viz_source_category_category_media",
+            label="Select primary sources that are collections of:",
+            options=options["primary_taxonomy"]["collection"],
+        )
+        filter_dict["processed_from_primary"] = filter_dict.get(
+            "processed_from_primary", {}
+        )
+        filter_dict["processed_from_primary"]["primary_types"] = st.multiselect(
+            key="viz_processed_from_primary_primary_types",
+            label="Select processed datasets whose primary sources contain:",
+            options=[f"web | {w}" for w in options["primary_taxonomy"]["website"]]
+            + options["primary_taxonomy"]["collection"],
+        )
+    if "media type" in filter_by:
+        filter_dict["media"] = {}
+        filter_dict["media"]["category"] = st.multiselect(
+            key="viz_media_category",
+            label="Select language data resources that contain:",
+            options=["text", "audiovisual", "image"],
+            help="Media data provided with transcription should go into **text**, then select the *transcribed* option. PDFs that have pre-extracted text information should go into **text**, PDFs that need OCR should go into **images**, select the latter if you're unsure",
+        )
+    filtered_catalogue = [
+        entry
+        for entry in catalogue
+        if filter_entry(entry, filter_dict) and not (entry["uid"] == "")
+    ]
+    st.markdown(
+        f"##### Your query matched **{len(filtered_catalogue)}** entries in the current catalogue."
+    )
+    return filtered_catalogue
+
+
+def viz_page():
+    st.title("🌸 - BigScience Catalog of Language Resources")
+    st.markdown("---\n")
+    catalogue = load_dataset("bigscience/collaborative_catalog")["train"]
+    with st.sidebar:
+        filtered_catalogue = filter_catalogue_visualization(catalogue, app_categories)
+        entry_location_type = st.radio(
+            label="I want to visualize",
+            options=[
+                "Where the organizations or data custodians are located",
+                "Where the language data creators are located",
+            ],
+            key="viz_show_location_type",
+        )
+        show_by_org = (
+            entry_location_type
+            == "Where the organizations or data custodians are located"
+        )
+    with st.expander("Map of entries", expanded=True):
+        filtered_counts = {}
+        for entry in filtered_catalogue:
+            locations = (
+                [entry["custodian"]["location"]]
+                if show_by_org
+                else entry["languages"]["language_locations"]
+            )
+            # be as specific as possible
+            locations = [
+                loc
+                for loc in locations
+                if not any([l in region_tree.get(loc, []) for l in locations])
+            ]
+            for loc in locations:
+                filtered_counts[loc] = filtered_counts.get(loc, 0) + 1
+        world_map = make_choro_map(filtered_counts)
+        folium_static(world_map, width=1150, height=600)
+    with st.expander("View selected resources", expanded=False):
+        st.write("You can further select locations to select entries from here:")
+        filter_region_choices = sorted(
+            set(
+                [
+                    loc
+                    for entry in filtered_catalogue
+                    for loc in (
+                        [entry["custodian"]["location"]]
+                        if show_by_org
+                        else entry["languages"]["language_locations"]
+                    )
+                ]
+            )
+        )
+        filter_locs = st.multiselect(
+            "View entries from the following locations:",
+            options=filter_region_choices,
+            key="viz_select_location",
+        )
+        filter_loc_dict = (
+            {"custodian": {"location": filter_locs}}
+            if show_by_org
+            else {"languages": {"language_locations": filter_locs}}
+        )
+        filtered_catalogue_by_loc = [
+            entry
+            for entry in filtered_catalogue
+            if filter_entry(entry, filter_loc_dict)
+        ]
+        view_entry = st.selectbox(
+            label="Select an entry to see more detail:",
+            options=filtered_catalogue_by_loc,
+            format_func=lambda entry: f"{entry['uid']} | {entry['description']['name']} -- {entry['description']['description']}",
+            key="viz_select_entry",
+        )
+        st.markdown(
+            f"##### *Type:* {view_entry['type']} *UID:* {view_entry['uid']} - *Name:* {view_entry['description']['name']}\n\n{view_entry['description']['description']}"
+        )
+        st.write(view_entry)
+
+
+if __name__ == "__main__":
+    main()

catalogue/__init__.py

@@ -0,0 +1 @@
+from .geography import countries, make_choro_map, region_tree

catalogue/geography.py

@@ -0,0 +1,151 @@
+import json
+
+import folium
+import pandas as pd
+from folium import Marker
+from folium.plugins import MarkerCluster
+from jinja2 import Template
+
+regions, countries, region_tree = json.load(
+    open("resources/country_regions.json", encoding="utf-8")
+)
+country_centers = json.load(
+    open("resources/country_center_coordinates.json", encoding="utf-8")
+)
+country_mappings = json.load(open("resources/country_mappings.json", encoding="utf-8"))
+
+WORLD_GEO_URL = "https://raw.githubusercontent.com/python-visualization/folium/master/examples/data/world-countries.json"
+
+ICON_CREATE_FUNCTIOM = """
+    function(cluster) {
+        var markers = cluster.getAllChildMarkers();
+        var sum = 0;
+        for (var i = 0; i < markers.length; i++) {
+            sum += markers[i].options.props.resources;
+        }
+
+        return L.divIcon({
+             html: '<b>' + sum + '</b>',
+             className: 'marker-cluster marker-cluster-small',
+             iconSize: new L.Point(20, 20)
+        });
+    }
+"""
+
+
+class MarkerWithProps(Marker):
+    _template = Template(
+        """
+        {% macro script(this, kwargs) %}
+        var {{this.get_name()}} = L.marker(
+            [{{this.location[0]}}, {{this.location[1]}}],
+            {
+                icon: new L.Icon.Default(),
+                {%- if this.draggable %}
+                draggable: true,
+                autoPan: true,
+                {%- endif %}
+                {%- if this.props %}
+                props : {{ this.props }}
+                {%- endif %}
+                }
+            )
+            .addTo({{this._parent.get_name()}});
+        {% endmacro %}
+        """
+    )
+
+    def __init__(
+        self, location, popup=None, tooltip=None, icon=None, draggable=False, props=None
+    ):
+        super(MarkerWithProps, self).__init__(
+            location=location,
+            popup=popup,
+            tooltip=tooltip,
+            icon=icon,
+            draggable=draggable,
+        )
+        self.props = json.loads(json.dumps(props))
+
+
+def get_region_center(region_name):
+    latitudes = []
+    longitudes = []
+    for name in region_tree[region_name]:
+        if name in region_tree:
+            region_latitudes, region_longitudes = get_region_center(name)
+            latitudes += region_latitudes
+            longitudes += region_longitudes
+        elif name in country_centers or name in country_mappings["to_center"]:
+            country_center = country_centers[
+                country_mappings["to_center"].get(name, name)
+            ]
+            latitudes += [float(country_center["latitude"])]
+            longitudes += [float(country_center["longitude"])]
+    return latitudes, longitudes
+
+
+def get_region_countries(region_name):
+    countries = []
+    for name in region_tree[region_name]:
+        if name in region_tree:
+            countries += get_region_countries(name)
+        else:
+            countries += [name]
+    return countries
+
+
+def make_choro_map(resource_counts, marker_thres=0):
+    world_map = folium.Map(tiles="cartodbpositron", location=[0, 0], zoom_start=1.5)
+    marker_cluster = MarkerCluster(icon_create_function=ICON_CREATE_FUNCTIOM)
+    marker_cluster.add_to(world_map)
+    for name, count in resource_counts.items():
+        if name in country_centers or name in country_mappings["to_center"]:
+            country_center = country_centers[
+                country_mappings["to_center"].get(name, name)
+            ]
+            MarkerWithProps(
+                location=[country_center["latitude"], country_center["longitude"]],
+                popup=f"{'Region' if name in region_tree else 'Country'} : {name}<br> \n Resources : {count} <br>",
+                props={"name": name, "resources": count},
+            ).add_to(marker_cluster)
+        # put a pin at the center of the region
+        elif name in region_tree:
+            latitudes, longitudes = get_region_center(name)
+            if len(latitudes) > 0:
+                lat = sum(latitudes) / len(latitudes)
+                lon = sum(longitudes) / len(longitudes)
+                MarkerWithProps(
+                    location=[lat, lon],
+                    popup=f"{'Region' if name in region_tree else 'Country'} : {name}<br> \n Resources : {count} <br>",
+                    props={"name": name, "resources": count},
+                ).add_to(marker_cluster)
+    # for choropleth, add counts to all countries in a region
+    choropleth_counts = {}
+    for loc_name in list(resource_counts.keys()):
+        if loc_name in region_tree:
+            for country_name in get_region_countries(loc_name):
+                choropleth_counts[country_name] = (
+                    choropleth_counts.get(country_name, 0) + resource_counts[loc_name]
+                )
+        else:
+            choropleth_counts[loc_name] = (
+                choropleth_counts.get(loc_name, 0) + resource_counts[loc_name]
+            )
+    df_resource_counts = pd.DataFrame(
+        [
+            (country_mappings["to_outline"].get(n, n), c)
+            for n, c in choropleth_counts.items()
+        ],
+        columns=["Name", "Resources"],
+    )
+    folium.Choropleth(
+        geo_data=WORLD_GEO_URL,
+        name="resource map",
+        data=df_resource_counts,
+        columns=["Name", "Resources"],
+        key_on="feature.properties.name",
+        fill_color="PuRd",
+        nan_fill_color="white",
+    ).add_to(world_map)
+    return world_map

resources/.gitignore

@@ -0,0 +1,2 @@
+*/*.jsonl
+

resources/country_center_coordinates.json

@@ -0,0 +1,1237 @@
+{
+  "Andorra": {
+    "latitude": "42.546245",
+    "longitude": "1.601554",
+    "code": "AD"
+  },
+  "United Arab Emirates": {
+    "latitude": "23.424076",
+    "longitude": "53.847818",
+    "code": "AE"
+  },
+  "Afghanistan": {
+    "latitude": "33.93911",
+    "longitude": "67.709953",
+    "code": "AF"
+  },
+  "Antigua and Barbuda": {
+    "latitude": "17.060816",
+    "longitude": "-61.796428",
+    "code": "AG"
+  },
+  "Anguilla": {
+    "latitude": "18.220554",
+    "longitude": "-63.068615",
+    "code": "AI"
+  },
+  "Albania": {
+    "latitude": "41.153332",
+    "longitude": "20.168331",
+    "code": "AL"
+  },
+  "Armenia": {
+    "latitude": "40.069099",
+    "longitude": "45.038189",
+    "code": "AM"
+  },
+  "Netherlands Antilles": {
+    "latitude": "12.226079",
+    "longitude": "-69.060087",
+    "code": "AN"
+  },
+  "Angola": {
+    "latitude": "-11.202692",
+    "longitude": "17.873887",
+    "code": "AO"
+  },
+  "Antarctica": {
+    "latitude": "-75.250973",
+    "longitude": "-0.071389",
+    "code": "AQ"
+  },
+  "Argentina": {
+    "latitude": "-38.416097",
+    "longitude": "-63.616672",
+    "code": "AR"
+  },
+  "American Samoa": {
+    "latitude": "-14.270972",
+    "longitude": "-170.132217",
+    "code": "AS"
+  },
+  "Austria": {
+    "latitude": "47.516231",
+    "longitude": "14.550072",
+    "code": "AT"
+  },
+  "Australia": {
+    "latitude": "-25.274398",
+    "longitude": "133.775136",
+    "code": "AU"
+  },
+  "Aruba": {
+    "latitude": "12.52111",
+    "longitude": "-69.968338",
+    "code": "AW"
+  },
+  "Azerbaijan": {
+    "latitude": "40.143105",
+    "longitude": "47.576927",
+    "code": "AZ"
+  },
+  "Bosnia and Herzegovina": {
+    "latitude": "43.915886",
+    "longitude": "17.679076",
+    "code": "BA"
+  },
+  "Barbados": {
+    "latitude": "13.193887",
+    "longitude": "-59.543198",
+    "code": "BB"
+  },
+  "Bangladesh": {
+    "latitude": "23.684994",
+    "longitude": "90.356331",
+    "code": "BD"
+  },
+  "Belgium": {
+    "latitude": "50.503887",
+    "longitude": "4.469936",
+    "code": "BE"
+  },
+  "Burkina Faso": {
+    "latitude": "12.238333",
+    "longitude": "-1.561593",
+    "code": "BF"
+  },
+  "Bulgaria": {
+    "latitude": "42.733883",
+    "longitude": "25.48583",
+    "code": "BG"
+  },
+  "Bahrain": {
+    "latitude": "25.930414",
+    "longitude": "50.637772",
+    "code": "BH"
+  },
+  "Burundi": {
+    "latitude": "-3.373056",
+    "longitude": "29.918886",
+    "code": "BI"
+  },
+  "Benin": {
+    "latitude": "9.30769",
+    "longitude": "2.315834",
+    "code": "BJ"
+  },
+  "Bermuda": {
+    "latitude": "32.321384",
+    "longitude": "-64.75737",
+    "code": "BM"
+  },
+  "Brunei": {
+    "latitude": "4.535277",
+    "longitude": "114.727669",
+    "code": "BN"
+  },
+  "Bolivia": {
+    "latitude": "-16.290154",
+    "longitude": "-63.588653",
+    "code": "BO"
+  },
+  "Brazil": {
+    "latitude": "-14.235004",
+    "longitude": "-51.92528",
+    "code": "BR"
+  },
+  "Bahamas": {
+    "latitude": "25.03428",
+    "longitude": "-77.39628",
+    "code": "BS"
+  },
+  "Bhutan": {
+    "latitude": "27.514162",
+    "longitude": "90.433601",
+    "code": "BT"
+  },
+  "Bouvet Island": {
+    "latitude": "-54.423199",
+    "longitude": "3.413194",
+    "code": "BV"
+  },
+  "Botswana": {
+    "latitude": "-22.328474",
+    "longitude": "24.684866",
+    "code": "BW"
+  },
+  "Belarus": {
+    "latitude": "53.709807",
+    "longitude": "27.953389",
+    "code": "BY"
+  },
+  "Belize": {
+    "latitude": "17.189877",
+    "longitude": "-88.49765",
+    "code": "BZ"
+  },
+  "Canada": {
+    "latitude": "56.130366",
+    "longitude": "-106.346771",
+    "code": "CA"
+  },
+  "Cocos [Keeling] Islands": {
+    "latitude": "-12.164165",
+    "longitude": "96.870956",
+    "code": "CC"
+  },
+  "Congo [DRC]": {
+    "latitude": "-4.038333",
+    "longitude": "21.758664",
+    "code": "CD"
+  },
+  "Central African Republic": {
+    "latitude": "6.611111",
+    "longitude": "20.939444",
+    "code": "CF"
+  },
+  "Congo [Republic]": {
+    "latitude": "-0.228021",
+    "longitude": "15.827659",
+    "code": "CG"
+  },
+  "Switzerland": {
+    "latitude": "46.818188",
+    "longitude": "8.227512",
+    "code": "CH"
+  },
+  "C\u00f4te d'Ivoire": {
+    "latitude": "7.539989",
+    "longitude": "-5.54708",
+    "code": "CI"
+  },
+  "Cook Islands": {
+    "latitude": "-21.236736",
+    "longitude": "-159.777671",
+    "code": "CK"
+  },
+  "Chile": {
+    "latitude": "-35.675147",
+    "longitude": "-71.542969",
+    "code": "CL"
+  },
+  "Cameroon": {
+    "latitude": "7.369722",
+    "longitude": "12.354722",
+    "code": "CM"
+  },
+  "China": {
+    "latitude": "35.86166",
+    "longitude": "104.195397",
+    "code": "CN"
+  },
+  "Colombia": {
+    "latitude": "4.570868",
+    "longitude": "-74.297333",
+    "code": "CO"
+  },
+  "Costa Rica": {
+    "latitude": "9.748917",
+    "longitude": "-83.753428",
+    "code": "CR"
+  },
+  "Cuba": {
+    "latitude": "21.521757",
+    "longitude": "-77.781167",
+    "code": "CU"
+  },
+  "Cape Verde": {
+    "latitude": "16.002082",
+    "longitude": "-24.013197",
+    "code": "CV"
+  },
+  "Christmas Island": {
+    "latitude": "-10.447525",
+    "longitude": "105.690449",
+    "code": "CX"
+  },
+  "Cyprus": {
+    "latitude": "35.126413",
+    "longitude": "33.429859",
+    "code": "CY"
+  },
+  "Czech Republic": {
+    "latitude": "49.817492",
+    "longitude": "15.472962",
+    "code": "CZ"
+  },
+  "Germany": {
+    "latitude": "51.165691",
+    "longitude": "10.451526",
+    "code": "DE"
+  },
+  "Djibouti": {
+    "latitude": "11.825138",
+    "longitude": "42.590275",
+    "code": "DJ"
+  },
+  "Denmark": {
+    "latitude": "56.26392",
+    "longitude": "9.501785",
+    "code": "DK"
+  },
+  "Dominica": {
+    "latitude": "15.414999",
+    "longitude": "-61.370976",
+    "code": "DM"
+  },
+  "Dominican Republic": {
+    "latitude": "18.735693",
+    "longitude": "-70.162651",
+    "code": "DO"
+  },
+  "Algeria": {
+    "latitude": "28.033886",
+    "longitude": "1.659626",
+    "code": "DZ"
+  },
+  "Ecuador": {
+    "latitude": "-1.831239",
+    "longitude": "-78.183406",
+    "code": "EC"
+  },
+  "Estonia": {
+    "latitude": "58.595272",
+    "longitude": "25.013607",
+    "code": "EE"
+  },
+  "Egypt": {
+    "latitude": "26.820553",
+    "longitude": "30.802498",
+    "code": "EG"
+  },
+  "Western Sahara": {
+    "latitude": "24.215527",
+    "longitude": "-12.885834",
+    "code": "EH"
+  },
+  "Eritrea": {
+    "latitude": "15.179384",
+    "longitude": "39.782334",
+    "code": "ER"
+  },
+  "Spain": {
+    "latitude": "40.463667",
+    "longitude": "-3.74922",
+    "code": "ES"
+  },
+  "Ethiopia": {
+    "latitude": "9.145",
+    "longitude": "40.489673",
+    "code": "ET"
+  },
+  "Finland": {
+    "latitude": "61.92411",
+    "longitude": "25.748151",
+    "code": "FI"
+  },
+  "Fiji": {
+    "latitude": "-16.578193",
+    "longitude": "179.414413",
+    "code": "FJ"
+  },
+  "Falkland Islands [Islas Malvinas]": {
+    "latitude": "-51.796253",
+    "longitude": "-59.523613",
+    "code": "FK"
+  },
+  "Micronesia": {
+    "latitude": "7.425554",
+    "longitude": "150.550812",
+    "code": "FM"
+  },
+  "Faroe Islands": {
+    "latitude": "61.892635",
+    "longitude": "-6.911806",
+    "code": "FO"
+  },
+  "France": {
+    "latitude": "46.227638",
+    "longitude": "2.213749",
+    "code": "FR"
+  },
+  "Gabon": {
+    "latitude": "-0.803689",
+    "longitude": "11.609444",
+    "code": "GA"
+  },
+  "United Kingdom": {
+    "latitude": "55.378051",
+    "longitude": "-3.435973",
+    "code": "GB"
+  },
+  "Grenada": {
+    "latitude": "12.262776",
+    "longitude": "-61.604171",
+    "code": "GD"
+  },
+  "Georgia": {
+    "latitude": "42.315407",
+    "longitude": "43.356892",
+    "code": "GE"
+  },
+  "French Guiana": {
+    "latitude": "3.933889",
+    "longitude": "-53.125782",
+    "code": "GF"
+  },
+  "Guernsey": {
+    "latitude": "49.465691",
+    "longitude": "-2.585278",
+    "code": "GG"
+  },
+  "Ghana": {
+    "latitude": "7.946527",
+    "longitude": "-1.023194",
+    "code": "GH"
+  },
+  "Gibraltar": {
+    "latitude": "36.137741",
+    "longitude": "-5.345374",
+    "code": "GI"
+  },
+  "Greenland": {
+    "latitude": "71.706936",
+    "longitude": "-42.604303",
+    "code": "GL"
+  },
+  "Gambia": {
+    "latitude": "13.443182",
+    "longitude": "-15.310139",
+    "code": "GM"
+  },
+  "Guinea": {
+    "latitude": "9.945587",
+    "longitude": "-9.696645",
+    "code": "GN"
+  },
+  "Guadeloupe": {
+    "latitude": "16.995971",
+    "longitude": "-62.067641",
+    "code": "GP"
+  },
+  "Equatorial Guinea": {
+    "latitude": "1.650801",
+    "longitude": "10.267895",
+    "code": "GQ"
+  },
+  "Greece": {
+    "latitude": "39.074208",
+    "longitude": "21.824312",
+    "code": "GR"
+  },
+  "South Georgia and the South Sandwich Islands": {
+    "latitude": "-54.429579",
+    "longitude": "-36.587909",
+    "code": "GS"
+  },
+  "Guatemala": {
+    "latitude": "15.783471",
+    "longitude": "-90.230759",
+    "code": "GT"
+  },
+  "Guam": {
+    "latitude": "13.444304",
+    "longitude": "144.793731",
+    "code": "GU"
+  },
+  "Guinea-Bissau": {
+    "latitude": "11.803749",
+    "longitude": "-15.180413",
+    "code": "GW"
+  },
+  "Guyana": {
+    "latitude": "4.860416",
+    "longitude": "-58.93018",
+    "code": "GY"
+  },
+  "Gaza Strip": {
+    "latitude": "31.354676",
+    "longitude": "34.308825",
+    "code": "GZ"
+  },
+  "Hong Kong": {
+    "latitude": "22.396428",
+    "longitude": "114.109497",
+    "code": "HK"
+  },
+  "Heard Island and McDonald Islands": {
+    "latitude": "-53.08181",
+    "longitude": "73.504158",
+    "code": "HM"
+  },
+  "Honduras": {
+    "latitude": "15.199999",
+    "longitude": "-86.241905",
+    "code": "HN"
+  },
+  "Croatia": {
+    "latitude": "45.1",
+    "longitude": "15.2",
+    "code": "HR"
+  },
+  "Haiti": {
+    "latitude": "18.971187",
+    "longitude": "-72.285215",
+    "code": "HT"
+  },
+  "Hungary": {
+    "latitude": "47.162494",
+    "longitude": "19.503304",
+    "code": "HU"
+  },
+  "Indonesia": {
+    "latitude": "-0.789275",
+    "longitude": "113.921327",
+    "code": "ID"
+  },
+  "Ireland": {
+    "latitude": "53.41291",
+    "longitude": "-8.24389",
+    "code": "IE"
+  },
+  "Israel": {
+    "latitude": "31.046051",
+    "longitude": "34.851612",
+    "code": "IL"
+  },
+  "Isle of Man": {
+    "latitude": "54.236107",
+    "longitude": "-4.548056",
+    "code": "IM"
+  },
+  "India": {
+    "latitude": "20.593684",
+    "longitude": "78.96288",
+    "code": "IN"
+  },
+  "British Indian Ocean Territory": {
+    "latitude": "-6.343194",
+    "longitude": "71.876519",
+    "code": "IO"
+  },
+  "Iraq": {
+    "latitude": "33.223191",
+    "longitude": "43.679291",
+    "code": "IQ"
+  },
+  "Iran": {
+    "latitude": "32.427908",
+    "longitude": "53.688046",
+    "code": "IR"
+  },
+  "Iceland": {
+    "latitude": "64.963051",
+    "longitude": "-19.020835",
+    "code": "IS"
+  },
+  "Italy": {
+    "latitude": "41.87194",
+    "longitude": "12.56738",
+    "code": "IT"
+  },
+  "Jersey": {
+    "latitude": "49.214439",
+    "longitude": "-2.13125",
+    "code": "JE"
+  },
+  "Jamaica": {
+    "latitude": "18.109581",
+    "longitude": "-77.297508",
+    "code": "JM"
+  },
+  "Jordan": {
+    "latitude": "30.585164",
+    "longitude": "36.238414",
+    "code": "JO"
+  },
+  "Japan": {
+    "latitude": "36.204824",
+    "longitude": "138.252924",
+    "code": "JP"
+  },
+  "Kenya": {
+    "latitude": "-0.023559",
+    "longitude": "37.906193",
+    "code": "KE"
+  },
+  "Kyrgyzstan": {
+    "latitude": "41.20438",
+    "longitude": "74.766098",
+    "code": "KG"
+  },
+  "Cambodia": {
+    "latitude": "12.565679",
+    "longitude": "104.990963",
+    "code": "KH"
+  },
+  "Kiribati": {
+    "latitude": "-3.370417",
+    "longitude": "-168.734039",
+    "code": "KI"
+  },
+  "Comoros": {
+    "latitude": "-11.875001",
+    "longitude": "43.872219",
+    "code": "KM"
+  },
+  "Saint Kitts and Nevis": {
+    "latitude": "17.357822",
+    "longitude": "-62.782998",
+    "code": "KN"
+  },
+  "North Korea": {
+    "latitude": "40.339852",
+    "longitude": "127.510093",
+    "code": "KP"
+  },
+  "South Korea": {
+    "latitude": "35.907757",
+    "longitude": "127.766922",
+    "code": "KR"
+  },
+  "Kuwait": {
+    "latitude": "29.31166",
+    "longitude": "47.481766",
+    "code": "KW"
+  },
+  "Cayman Islands": {
+    "latitude": "19.513469",
+    "longitude": "-80.566956",
+    "code": "KY"
+  },
+  "Kazakhstan": {
+    "latitude": "48.019573",
+    "longitude": "66.923684",
+    "code": "KZ"
+  },
+  "Laos": {
+    "latitude": "19.85627",
+    "longitude": "102.495496",
+    "code": "LA"
+  },
+  "Lebanon": {
+    "latitude": "33.854721",
+    "longitude": "35.862285",
+    "code": "LB"
+  },
+  "Saint Lucia": {
+    "latitude": "13.909444",
+    "longitude": "-60.978893",
+    "code": "LC"
+  },
+  "Liechtenstein": {
+    "latitude": "47.166",
+    "longitude": "9.555373",
+    "code": "LI"
+  },
+  "Sri Lanka": {
+    "latitude": "7.873054",
+    "longitude": "80.771797",
+    "code": "LK"
+  },
+  "Liberia": {
+    "latitude": "6.428055",
+    "longitude": "-9.429499",
+    "code": "LR"
+  },
+  "Lesotho": {
+    "latitude": "-29.609988",
+    "longitude": "28.233608",
+    "code": "LS"
+  },
+  "Lithuania": {
+    "latitude": "55.169438",
+    "longitude": "23.881275",
+    "code": "LT"
+  },
+  "Luxembourg": {
+    "latitude": "49.815273",
+    "longitude": "6.129583",
+    "code": "LU"
+  },
+  "Latvia": {
+    "latitude": "56.879635",
+    "longitude": "24.603189",
+    "code": "LV"
+  },
+  "Libya": {
+    "latitude": "26.3351",
+    "longitude": "17.228331",
+    "code": "LY"
+  },
+  "Morocco": {
+    "latitude": "31.791702",
+    "longitude": "-7.09262",
+    "code": "MA"
+  },
+  "Monaco": {
+    "latitude": "43.750298",
+    "longitude": "7.412841",
+    "code": "MC"
+  },
+  "Moldova": {
+    "latitude": "47.411631",
+    "longitude": "28.369885",
+    "code": "MD"
+  },
+  "Montenegro": {
+    "latitude": "42.708678",
+    "longitude": "19.37439",
+    "code": "ME"
+  },
+  "Madagascar": {
+    "latitude": "-18.766947",
+    "longitude": "46.869107",
+    "code": "MG"
+  },
+  "Marshall Islands": {
+    "latitude": "7.131474",
+    "longitude": "171.184478",
+    "code": "MH"
+  },
+  "Macedonia [FYROM]": {
+    "latitude": "41.608635",
+    "longitude": "21.745275",
+    "code": "MK"
+  },
+  "Mali": {
+    "latitude": "17.570692",
+    "longitude": "-3.996166",
+    "code": "ML"
+  },
+  "Myanmar [Burma]": {
+    "latitude": "21.913965",
+    "longitude": "95.956223",
+    "code": "MM"
+  },
+  "Mongolia": {
+    "latitude": "46.862496",
+    "longitude": "103.846656",
+    "code": "MN"
+  },
+  "Macau": {
+    "latitude": "22.198745",
+    "longitude": "113.543873",
+    "code": "MO"
+  },
+  "Northern Mariana Islands": {
+    "latitude": "17.33083",
+    "longitude": "145.38469",
+    "code": "MP"
+  },
+  "Martinique": {
+    "latitude": "14.641528",
+    "longitude": "-61.024174",
+    "code": "MQ"
+  },
+  "Mauritania": {
+    "latitude": "21.00789",
+    "longitude": "-10.940835",
+    "code": "MR"
+  },
+  "Montserrat": {
+    "latitude": "16.742498",
+    "longitude": "-62.187366",
+    "code": "MS"
+  },
+  "Malta": {
+    "latitude": "35.937496",
+    "longitude": "14.375416",
+    "code": "MT"
+  },
+  "Mauritius": {
+    "latitude": "-20.348404",
+    "longitude": "57.552152",
+    "code": "MU"
+  },
+  "Maldives": {
+    "latitude": "3.202778",
+    "longitude": "73.22068",
+    "code": "MV"
+  },
+  "Malawi": {
+    "latitude": "-13.254308",
+    "longitude": "34.301525",
+    "code": "MW"
+  },
+  "Mexico": {
+    "latitude": "23.634501",
+    "longitude": "-102.552784",
+    "code": "MX"
+  },
+  "Malaysia": {
+    "latitude": "4.210484",
+    "longitude": "101.975766",
+    "code": "MY"
+  },
+  "Mozambique": {
+    "latitude": "-18.665695",
+    "longitude": "35.529562",
+    "code": "MZ"
+  },
+  "Namibia": {
+    "latitude": "-22.95764",
+    "longitude": "18.49041",
+    "code": "NA"
+  },
+  "New Caledonia": {
+    "latitude": "-20.904305",
+    "longitude": "165.618042",
+    "code": "NC"
+  },
+  "Niger": {
+    "latitude": "17.607789",
+    "longitude": "8.081666",
+    "code": "NE"
+  },
+  "Norfolk Island": {
+    "latitude": "-29.040835",
+    "longitude": "167.954712",
+    "code": "NF"
+  },
+  "Nigeria": {
+    "latitude": "9.081999",
+    "longitude": "8.675277",
+    "code": "NG"
+  },
+  "Nicaragua": {
+    "latitude": "12.865416",
+    "longitude": "-85.207229",
+    "code": "NI"
+  },
+  "Netherlands": {
+    "latitude": "52.132633",
+    "longitude": "5.291266",
+    "code": "NL"
+  },
+  "Norway": {
+    "latitude": "60.472024",
+    "longitude": "8.468946",
+    "code": "NO"
+  },
+  "Nepal": {
+    "latitude": "28.394857",
+    "longitude": "84.124008",
+    "code": "NP"
+  },
+  "Nauru": {
+    "latitude": "-0.522778",
+    "longitude": "166.931503",
+    "code": "NR"
+  },
+  "Niue": {
+    "latitude": "-19.054445",
+    "longitude": "-169.867233",
+    "code": "NU"
+  },
+  "New Zealand": {
+    "latitude": "-40.900557",
+    "longitude": "174.885971",
+    "code": "NZ"
+  },
+  "Oman": {
+    "latitude": "21.512583",
+    "longitude": "55.923255",
+    "code": "OM"
+  },
+  "Panama": {
+    "latitude": "8.537981",
+    "longitude": "-80.782127",
+    "code": "PA"
+  },
+  "Peru": {
+    "latitude": "-9.189967",
+    "longitude": "-75.015152",
+    "code": "PE"
+  },
+  "French Polynesia": {
+    "latitude": "-17.679742",
+    "longitude": "-149.406843",
+    "code": "PF"
+  },
+  "Papua New Guinea": {
+    "latitude": "-6.314993",
+    "longitude": "143.95555",
+    "code": "PG"
+  },
+  "Philippines": {
+    "latitude": "12.879721",
+    "longitude": "121.774017",
+    "code": "PH"
+  },
+  "Pakistan": {
+    "latitude": "30.375321",
+    "longitude": "69.345116",
+    "code": "PK"
+  },
+  "Poland": {
+    "latitude": "51.919438",
+    "longitude": "19.145136",
+    "code": "PL"
+  },
+  "Saint Pierre and Miquelon": {
+    "latitude": "46.941936",
+    "longitude": "-56.27111",
+    "code": "PM"
+  },
+  "Pitcairn Islands": {
+    "latitude": "-24.703615",
+    "longitude": "-127.439308",
+    "code": "PN"
+  },
+  "Puerto Rico": {
+    "latitude": "18.220833",
+    "longitude": "-66.590149",
+    "code": "PR"
+  },
+  "Palestinian Territories": {
+    "latitude": "31.952162",
+    "longitude": "35.233154",
+    "code": "PS"
+  },
+  "Portugal": {
+    "latitude": "39.399872",
+    "longitude": "-8.224454",
+    "code": "PT"
+  },
+  "Palau": {
+    "latitude": "7.51498",
+    "longitude": "134.58252",
+    "code": "PW"
+  },
+  "Paraguay": {
+    "latitude": "-23.442503",
+    "longitude": "-58.443832",
+    "code": "PY"
+  },
+  "Qatar": {
+    "latitude": "25.354826",
+    "longitude": "51.183884",
+    "code": "QA"
+  },
+  "R\u00e9union": {
+    "latitude": "-21.115141",
+    "longitude": "55.536384",
+    "code": "RE"
+  },
+  "Romania": {
+    "latitude": "45.943161",
+    "longitude": "24.96676",
+    "code": "RO"
+  },
+  "Serbia": {
+    "latitude": "44.016521",
+    "longitude": "21.005859",
+    "code": "RS"
+  },
+  "Russia": {
+    "latitude": "61.52401",
+    "longitude": "105.318756",
+    "code": "RU"
+  },
+  "Rwanda": {
+    "latitude": "-1.940278",
+    "longitude": "29.873888",
+    "code": "RW"
+  },
+  "Saudi Arabia": {
+    "latitude": "23.885942",
+    "longitude": "45.079162",
+    "code": "SA"
+  },
+  "Solomon Islands": {
+    "latitude": "-9.64571",
+    "longitude": "160.156194",
+    "code": "SB"
+  },
+  "Seychelles": {
+    "latitude": "-4.679574",
+    "longitude": "55.491977",
+    "code": "SC"
+  },
+  "Sudan": {
+    "latitude": "12.862807",
+    "longitude": "30.217636",
+    "code": "SD"
+  },
+  "Sweden": {
+    "latitude": "60.128161",
+    "longitude": "18.643501",
+    "code": "SE"
+  },
+  "Singapore": {
+    "latitude": "1.352083",
+    "longitude": "103.819836",
+    "code": "SG"
+  },
+  "Saint Helena": {
+    "latitude": "-24.143474",
+    "longitude": "-10.030696",
+    "code": "SH"
+  },
+  "Slovenia": {
+    "latitude": "46.151241",
+    "longitude": "14.995463",
+    "code": "SI"
+  },
+  "Svalbard and Jan Mayen": {
+    "latitude": "77.553604",
+    "longitude": "23.670272",
+    "code": "SJ"
+  },
+  "Slovakia": {
+    "latitude": "48.669026",
+    "longitude": "19.699024",
+    "code": "SK"
+  },
+  "Sierra Leone": {
+    "latitude": "8.460555",
+    "longitude": "-11.779889",
+    "code": "SL"
+  },
+  "San Marino": {
+    "latitude": "43.94236",
+    "longitude": "12.457777",
+    "code": "SM"
+  },
+  "Senegal": {
+    "latitude": "14.497401",
+    "longitude": "-14.452362",
+    "code": "SN"
+  },
+  "Somalia": {
+    "latitude": "5.152149",
+    "longitude": "46.199616",
+    "code": "SO"
+  },
+  "Suriname": {
+    "latitude": "3.919305",
+    "longitude": "-56.027783",
+    "code": "SR"
+  },
+  "S\u00e3o Tom\u00e9 and Pr\u00edncipe": {
+    "latitude": "0.18636",
+    "longitude": "6.613081",
+    "code": "ST"
+  },
+  "El Salvador": {
+    "latitude": "13.794185",
+    "longitude": "-88.89653",
+    "code": "SV"
+  },
+  "Syria": {
+    "latitude": "34.802075",
+    "longitude": "38.996815",
+    "code": "SY"
+  },
+  "Swaziland": {
+    "latitude": "-26.522503",
+    "longitude": "31.465866",
+    "code": "SZ"
+  },
+  "Turks and Caicos Islands": {
+    "latitude": "21.694025",
+    "longitude": "-71.797928",
+    "code": "TC"
+  },
+  "Chad": {
+    "latitude": "15.454166",
+    "longitude": "18.732207",
+    "code": "TD"
+  },
+  "French Southern Territories": {
+    "latitude": "-49.280366",
+    "longitude": "69.348557",
+    "code": "TF"
+  },
+  "Togo": {
+    "latitude": "8.619543",
+    "longitude": "0.824782",
+    "code": "TG"
+  },
+  "Thailand": {
+    "latitude": "15.870032",
+    "longitude": "100.992541",
+    "code": "TH"
+  },
+  "Tajikistan": {
+    "latitude": "38.861034",
+    "longitude": "71.276093",
+    "code": "TJ"
+  },
+  "Tokelau": {
+    "latitude": "-8.967363",
+    "longitude": "-171.855881",
+    "code": "TK"
+  },
+  "Timor-Leste": {
+    "latitude": "-8.874217",
+    "longitude": "125.727539",
+    "code": "TL"
+  },
+  "Turkmenistan": {
+    "latitude": "38.969719",
+    "longitude": "59.556278",
+    "code": "TM"
+  },
+  "Tunisia": {
+    "latitude": "33.886917",
+    "longitude": "9.537499",
+    "code": "TN"
+  },
+  "Tonga": {
+    "latitude": "-21.178986",
+    "longitude": "-175.198242",
+    "code": "TO"
+  },
+  "Turkey": {
+    "latitude": "38.963745",
+    "longitude": "35.243322",
+    "code": "TR"
+  },
+  "Trinidad and Tobago": {
+    "latitude": "10.691803",
+    "longitude": "-61.222503",
+    "code": "TT"
+  },
+  "Tuvalu": {
+    "latitude": "-7.109535",
+    "longitude": "177.64933",
+    "code": "TV"
+  },
+  "Taiwan": {
+    "latitude": "23.69781",
+    "longitude": "120.960515",
+    "code": "TW"
+  },
+  "Tanzania": {
+    "latitude": "-6.369028",
+    "longitude": "34.888822",
+    "code": "TZ"
+  },
+  "Ukraine": {
+    "latitude": "48.379433",
+    "longitude": "31.16558",
+    "code": "UA"
+  },
+  "Uganda": {
+    "latitude": "1.373333",
+    "longitude": "32.290275",
+    "code": "UG"
+  },
+  "U.S. Minor Outlying Islands": {
+    "latitude": "",
+    "longitude": "",
+    "code": "UM"
+  },
+  "United States": {
+    "latitude": "37.09024",
+    "longitude": "-95.712891",
+    "code": "US"
+  },
+  "Uruguay": {
+    "latitude": "-32.522779",
+    "longitude": "-55.765835",
+    "code": "UY"
+  },
+  "Uzbekistan": {
+    "latitude": "41.377491",
+    "longitude": "64.585262",
+    "code": "UZ"
+  },
+  "Vatican City": {
+    "latitude": "41.902916",
+    "longitude": "12.453389",
+    "code": "VA"
+  },
+  "Saint Vincent and the Grenadines": {
+    "latitude": "12.984305",
+    "longitude": "-61.287228",
+    "code": "VC"
+  },
+  "Venezuela": {
+    "latitude": "6.42375",
+    "longitude": "-66.58973",
+    "code": "VE"
+  },
+  "British Virgin Islands": {
+    "latitude": "18.420695",
+    "longitude": "-64.639968",
+    "code": "VG"
+  },
+  "U.S. Virgin Islands": {
+    "latitude": "18.335765",
+    "longitude": "-64.896335",
+    "code": "VI"
+  },
+  "Vietnam": {
+    "latitude": "14.058324",
+    "longitude": "108.277199",
+    "code": "VN"
+  },
+  "Vanuatu": {
+    "latitude": "-15.376706",
+    "longitude": "166.959158",
+    "code": "VU"
+  },
+  "Wallis and Futuna": {
+    "latitude": "-13.768752",
+    "longitude": "-177.156097",
+    "code": "WF"
+  },
+  "Samoa": {
+    "latitude": "-13.759029",
+    "longitude": "-172.104629",
+    "code": "WS"
+  },
+  "Kosovo": {
+    "latitude": "42.602636",
+    "longitude": "20.902977",
+    "code": "XK"
+  },
+  "Yemen": {
+    "latitude": "15.552727",
+    "longitude": "48.516388",
+    "code": "YE"
+  },
+  "Mayotte": {
+    "latitude": "-12.8275",
+    "longitude": "45.166244",
+    "code": "YT"
+  },
+  "South Africa": {
+    "latitude": "-30.559482",
+    "longitude": "22.937506",
+    "code": "ZA"
+  },
+  "Zambia": {
+    "latitude": "-13.133897",
+    "longitude": "27.849332",
+    "code": "ZM"
+  },
+  "Zimbabwe": {
+    "latitude": "-19.015438",
+    "longitude": "29.154857",
+    "code": "ZW"
+  },
+  "Northern Cyprus": {
+    "latitude": "35.11",
+    "longitude": "33.22",
+    "code": "TRC"
+  },
+  "South Sudan": {
+    "latitude": "04.51",
+    "longitude": "31.36",
+    "code": ""
+  }
+}

resources/country_mappings.json

@@ -0,0 +1,116 @@
+{
+  "to_outline": {
+    "Democratic Republic of the Congo": "Democratic Republic of the Congo",
+    "Congo [DRC]": "Democratic Republic of the Congo",
+    "Congo": "Democratic Republic of the Congo",
+    "Republic of the Congo": "Republic of the Congo",
+    "Congo [Republic]": "Republic of the Congo",
+    "East Timor": "East Timor",
+    "Timor-Leste": "East Timor",
+    "Falkland Islands": "Falkland Islands",
+    "Falkland Islands [Islas Malvinas]": "Falkland Islands",
+    "Gambia": "Gambia",
+    "The Gambia": "Gambia",
+    "Guinea Bissau": "Guinea Bissau",
+    "Guinea-Bissau": "Guinea Bissau",
+    "Ivory Coast": "Ivory Coast",
+    "C\u00f4te d'Ivoire": "Ivory Coast",
+    "Cote d'Ivoire": "Ivory Coast",
+    "Macedonia": "Macedonia",
+    "Macedonia [FYROM]": "Macedonia",
+    "North Macedonia": "Macedonia",
+    "Myanmar": "Myanmar",
+    "Myanmar [Burma]": "Myanmar",
+    "Republic of Serbia": "Republic of Serbia",
+    "Serbia": "Republic of Serbia",
+    "Somaliland": "Somaliland",
+    "Somalia": "Somalia",
+    "Swaziland": "Swaziland",
+    "Eswatini": "Swaziland",
+    "The Bahamas": "The Bahamas",
+    "Bahamas": "The Bahamas",
+    "United Republic of Tanzania": "United Republic of Tanzania",
+    "Tanzania": "United Republic of Tanzania",
+    "United States of America": "United States of America",
+    "United States": "United States of America",
+    "West Bank": "West Bank",
+    "Palestinian Territories": "West Bank",
+    "Palestine": "West Bank"
+  },
+  "to_center": {
+    "Democratic Republic of the Congo": "Congo [DRC]",
+    "Congo [DRC]": "Congo [DRC]",
+    "Congo": "Congo [DRC]",
+    "Republic of the Congo": "Congo [Republic]",
+    "Congo [Republic]": "Congo [Republic]",
+    "East Timor": "Timor-Leste",
+    "Timor-Leste": "Timor-Leste",
+    "Falkland Islands": "Falkland Islands [Islas Malvinas]",
+    "Falkland Islands [Islas Malvinas]": "Falkland Islands [Islas Malvinas]",
+    "Gambia": "Gambia",
+    "The Gambia": "Gambia",
+    "Guinea Bissau": "Guinea-Bissau",
+    "Guinea-Bissau": "Guinea-Bissau",
+    "Ivory Coast": "C\u00f4te d'Ivoire",
+    "C\u00f4te d'Ivoire": "C\u00f4te d'Ivoire",
+    "Cote d'Ivoire": "C\u00f4te d'Ivoire",
+    "Macedonia": "Macedonia [FYROM]",
+    "Macedonia [FYROM]": "Macedonia [FYROM]",
+    "North Macedonia": "Macedonia [FYROM]",
+    "Myanmar": "Myanmar [Burma]",
+    "Myanmar [Burma]": "Myanmar [Burma]",
+    "Republic of Serbia": "Serbia",
+    "Serbia": "Serbia",
+    "Somaliland": "Somalia",
+    "Somalia": "Somalia",
+    "Swaziland": "Swaziland",
+    "Eswatini": "Swaziland",
+    "The Bahamas": "Bahamas",
+    "Bahamas": "Bahamas",
+    "United Republic of Tanzania": "Tanzania",
+    "Tanzania": "Tanzania",
+    "United States of America": "United States",
+    "United States": "United States",
+    "West Bank": "Palestinian Territories",
+    "Palestinian Territories": "Palestinian Territories",
+    "Palestine": "Palestinian Territories"
+  },
+  "to_region": {
+    "Democratic Republic of the Congo": "Congo",
+    "Congo [DRC]": "Congo",
+    "Congo": "Congo",
+    "Republic of the Congo": "Republic of the Congo",
+    "Congo [Republic]": "Republic of the Congo",
+    "East Timor": "Timor-Leste",
+    "Timor-Leste": "Timor-Leste",
+    "Falkland Islands": "Falkland Islands",
+    "Falkland Islands [Islas Malvinas]": "Falkland Islands",
+    "Gambia": "The Gambia",
+    "The Gambia": "The Gambia",
+    "Guinea Bissau": "Guinea-Bissau",
+    "Guinea-Bissau": "Guinea-Bissau",
+    "Ivory Coast": "Cote d'Ivoire",
+    "C\u00f4te d'Ivoire": "Cote d'Ivoire",
+    "Cote d'Ivoire": "Cote d'Ivoire",
+    "Macedonia": "North Macedonia",
+    "Macedonia [FYROM]": "North Macedonia",
+    "North Macedonia": "North Macedonia",
+    "Myanmar": "Myanmar",
+    "Myanmar [Burma]": "Myanmar",
+    "Republic of Serbia": "Serbia",
+    "Serbia": "Serbia",
+    "Somaliland": "Somalia",
+    "Somalia": "Somalia",
+    "Swaziland": "Eswatini",
+    "Eswatini": "Eswatini",
+    "The Bahamas": "The Bahamas",
+    "Bahamas": "The Bahamas",
+    "United Republic of Tanzania": "Tanzania",
+    "Tanzania": "Tanzania",
+    "United States of America": "United States of America",
+    "United States": "United States of America",
+    "West Bank": "Palestine",
+    "Palestinian Territories": "Palestine",
+    "Palestine": "Palestine"
+  }
+}

resources/country_regions.json

@@ -0,0 +1,652 @@
+[
+  [
+    "Africa",
+    "Americas",
+    "Asia",
+    "Europe",
+    "Oceania"
+  ],
+  [
+    "Afghanistan",
+    "Albania",
+    "Algeria",
+    "American Samoa",
+    "Andorra",
+    "Angola",
+    "Anguilla",
+    "Antigua and Barbuda",
+    "Argentina",
+    "Armenia",
+    "Aruba",
+    "Australia",
+    "Austria",
+    "Azerbaijan",
+    "Azores",
+    "Bahrain",
+    "Bangladesh",
+    "Barbados",
+    "Belarus",
+    "Belgium",
+    "Belize",
+    "Benin",
+    "Bermuda",
+    "Bhutan",
+    "Bolivia",
+    "Bonaire, Saint Eustatius and Saba",
+    "Bosnia and Herzegovina",
+    "Botswana",
+    "Bouvet Island",
+    "Brazil",
+    "British Indian Ocean Territory",
+    "British Virgin Islands",
+    "Brunei",
+    "Bulgaria",
+    "Burkina Faso",
+    "Burundi",
+    "Cambodia",
+    "Cameroon",
+    "Canada",
+    "Cape Verde",
+    "Cayman Islands",
+    "Central African Republic",
+    "Chad",
+    "Chile",
+    "Christmas Island",
+    "Cocos (Keeling) Islands",
+    "Colombia",
+    "Comoros",
+    "Congo",
+    "Cook Islands",
+    "Costa Rica",
+    "Cote d'Ivoire",
+    "Croatia",
+    "Cuba",
+    "Cura\u00e7ao",
+    "Cyprus",
+    "Czech Republic",
+    "Denmark",
+    "Djibouti",
+    "Dominica",
+    "Dominican Republic",
+    "Ecuador",
+    "Egypt",
+    "El Salvador",
+    "England",
+    "Equatorial Guinea",
+    "Eritrea",
+    "Estonia",
+    "Eswatini",
+    "Ethiopia",
+    "Falkland Islands",
+    "Faroe Islands",
+    "Federated States of Micronesia",
+    "Fiji",
+    "Finland",
+    "France",
+    "French Guiana",
+    "French Polynesia",
+    "French Southern Territories",
+    "Gabon",
+    "Georgia",
+    "Germany",
+    "Ghana",
+    "Gibraltar",
+    "Greece",
+    "Greenland",
+    "Grenada",
+    "Guadeloupe",
+    "Guam",
+    "Guatemala",
+    "Guernsey",
+    "Guinea",
+    "Guinea-Bissau",
+    "Guyana",
+    "Haiti",
+    "Heard Island and McDonald Islands",
+    "Holy See",
+    "Honduras",
+    "Hong Kong",
+    "Hungary",
+    "Iceland",
+    "India",
+    "Indonesia",
+    "Iran",
+    "Iraq",
+    "Ireland",
+    "Isle of Man",
+    "Israel",
+    "Jamaica",
+    "Japan",
+    "Jersey",
+    "Jordan",
+    "Kazakhstan",
+    "Kenya",
+    "Kiribati",
+    "Kosovo",
+    "Kuwait",
+    "Kyrgyzstan",
+    "Laos",
+    "Latvia",
+    "Lebanon",
+    "Lesotho",
+    "Liberia",
+    "Libya",
+    "Liechtenstein",
+    "Lithuania",
+    "Luxembourg",
+    "Macau",
+    "Madagascar",
+    "Madeira",
+    "Malawi",
+    "Malaysia",
+    "Maldives",
+    "Mali",
+    "Malta",
+    "Marshall Islands",
+    "Martinique",
+    "Mauritania",
+    "Mauritius",
+    "Mayotte",
+    "Mexico",
+    "Moldova",
+    "Monaco",
+    "Mongolia",
+    "Montenegro",
+    "Montserrat",
+    "Morocco",
+    "Mozambique",
+    "Myanmar",
+    "Namibia",
+    "Nauru",
+    "Nepal",
+    "Netherlands",
+    "New Caledonia",
+    "New Zealand",
+    "Nicaragua",
+    "Niger",
+    "Nigeria",
+    "Niue",
+    "Norfolk Island",
+    "North Korea",
+    "North Macedonia",
+    "Northern Ireland",
+    "Northern Mariana Islands",
+    "Norway",
+    "Oman",
+    "Pakistan",
+    "Palau",
+    "Palestine",
+    "Panama",
+    "Papua New Guinea",
+    "Paraguay",
+    "Peru",
+    "Philippines",
+    "Pitcairn Islands",
+    "Poland",
+    "Puerto Rico",
+    "Qatar",
+    "Republic of the Congo",
+    "Romania",
+    "Russia",
+    "Rwanda",
+    "R\u00e9union",
+    "Saint Barth\u00e9lemy",
+    "Saint Helena, Ascension and Tristan da Cunha",
+    "Saint Kitts and Nevis",
+    "Saint Lucia",
+    "Saint Martin",
+    "Saint Pierre and Miquelon",
+    "Saint Vincent and the Grenadines",
+    "Samoa",
+    "San Marino",
+    "Sark",
+    "Saudi Arabia",
+    "Scotland",
+    "Senegal",
+    "Seychelles",
+    "Sierra Leone",
+    "Singapore",
+    "Sint Maarten",
+    "Slovakia",
+    "Slovenia",
+    "Solomon Islands",
+    "Somalia",
+    "Somaliland",
+    "South Africa",
+    "South Georgia and the South Sandwich Islands",
+    "South Korea",
+    "South Sudan",
+    "Spain",
+    "Sri Lanka",
+    "Sudan",
+    "Suriname",
+    "Svalbard and Jan Mayen Islands",
+    "Sweden",
+    "Switzerland",
+    "Syria",
+    "S\u00e3o Tom\u00e9 and Pr\u00edncipe",
+    "Taiwan",
+    "Tajikistan",
+    "Tanzania",
+    "Thailand",
+    "The Bahamas",
+    "The Gambia",
+    "Timor-Leste",
+    "Togo",
+    "Tokelau",
+    "Tonga",
+    "Trinidad and Tobago",
+    "Tunisia",
+    "Turkey",
+    "Turkmenistan",
+    "Turks and Caicos Islands",
+    "Tuvalu",
+    "Uganda",
+    "Ukraine",
+    "United Arab Emirates",
+    "United Kingdom",
+    "United States Minor Outlying Islands",
+    "United States Virgin Islands",
+    "United States of America",
+    "Uruguay",
+    "Uzbekistan",
+    "Vanuatu",
+    "Vatican City",
+    "Venezuela",
+    "Vietnam",
+    "Wales",
+    "Wallis and Futuna Islands",
+    "Western Sahara",
+    "Yemen",
+    "Zambia",
+    "Zimbabwe",
+    "\u00c5land Islands",
+    "China",
+    "Italy",
+    "Northern Cyprus",
+    "Portugal",
+    "Serbia"
+  ],
+  {
+    "World-Wide": [
+      "Africa",
+      "Americas",
+      "Asia",
+      "Europe",
+      "Oceania"
+    ],
+    "Africa": [
+      "Northern Africa",
+      "Sub-Saharan Africa"
+    ],
+    "Northern Africa": [
+      "Algeria",
+      "Egypt",
+      "Libya",
+      "Morocco",
+      "Sudan",
+      "Tunisia",
+      "Western Sahara"
+    ],
+    "Sub-Saharan Africa": [
+      "Eastern Africa",
+      "Middle Africa",
+      "Southern Africa",
+      "Western Africa"
+    ],
+    "Eastern Africa": [
+      "British Indian Ocean Territory",
+      "Burundi",
+      "Comoros",
+      "Djibouti",
+      "Eritrea",
+      "Ethiopia",
+      "French Southern Territories",
+      "Kenya",
+      "Madagascar",
+      "Malawi",
+      "Mauritius",
+      "Mayotte",
+      "Mozambique",
+      "R\u00e9union",
+      "Rwanda",
+      "Seychelles",
+      "Somalia",
+      "Somaliland",
+      "South Sudan",
+      "Uganda",
+      "Zambia",
+      "Zimbabwe",
+      "Tanzania"
+    ],
+    "Middle Africa": [
+      "Angola",
+      "Cameroon",
+      "Central African Republic",
+      "Chad",
+      "Congo",
+      "Equatorial Guinea",
+      "Gabon",
+      "S\u00e3o Tom\u00e9 and Pr\u00edncipe",
+      "Republic of the Congo"
+    ],
+    "Southern Africa": [
+      "Botswana",
+      "Eswatini",
+      "Lesotho",
+      "Namibia",
+      "South Africa"
+    ],
+    "Western Africa": [
+      "Benin",
+      "Burkina Faso",
+      "Ghana",
+      "Guinea",
+      "Guinea-Bissau",
+      "Liberia",
+      "Mali",
+      "Mauritania",
+      "Niger",
+      "Nigeria",
+      "Senegal",
+      "Sierra Leone",
+      "Togo",
+      "Cote d'Ivoire",
+      "Cape Verde",
+      "Saint Helena, Ascension and Tristan da Cunha",
+      "The Gambia"
+    ],
+    "Americas": [
+      "Latin America and the Caribbean",
+      "Northern America"
+    ],
+    "Latin America and the Caribbean": [
+      "Caribbean",
+      "Central America",
+      "South America"
+    ],
+    "Caribbean": [
+      "Anguilla",
+      "Antigua and Barbuda",
+      "Aruba",
+      "Barbados",
+      "British Virgin Islands",
+      "Cayman Islands",
+      "Cuba",
+      "Cura\u00e7ao",
+      "Dominica",
+      "Dominican Republic",
+      "Grenada",
+      "Guadeloupe",
+      "Haiti",
+      "Jamaica",
+      "Martinique",
+      "Montserrat",
+      "Puerto Rico",
+      "Saint Barth\u00e9lemy",
+      "Saint Kitts and Nevis",
+      "Saint Lucia",
+      "Saint Vincent and the Grenadines",
+      "Trinidad and Tobago",
+      "Turks and Caicos Islands",
+      "United States Virgin Islands",
+      "The Bahamas",
+      "Sint Maarten",
+      "Saint Martin",
+      "Bonaire, Saint Eustatius and Saba"
+    ],
+    "Central America": [
+      "Belize",
+      "Costa Rica",
+      "El Salvador",
+      "Guatemala",
+      "Honduras",
+      "Mexico",
+      "Nicaragua",
+      "Panama"
+    ],
+    "South America": [
+      "Argentina",
+      "Bouvet Island",
+      "Brazil",
+      "Chile",
+      "Colombia",
+      "Ecuador",
+      "French Guiana",
+      "Guyana",
+      "Paraguay",
+      "Peru",
+      "South Georgia and the South Sandwich Islands",
+      "Suriname",
+      "Uruguay",
+      "Bolivia",
+      "Venezuela",
+      "Falkland Islands"
+    ],
+    "Northern America": [
+      "Bermuda",
+      "Canada",
+      "Greenland",
+      "Saint Pierre and Miquelon",
+      "United States of America"
+    ],
+    "Asia": [
+      "Central Asia",
+      "Eastern Asia",
+      "South-eastern Asia",
+      "Southern Asia",
+      "Western Asia"
+    ],
+    "Central Asia": [
+      "Kazakhstan",
+      "Kyrgyzstan",
+      "Tajikistan",
+      "Turkmenistan",
+      "Uzbekistan"
+    ],
+    "Eastern Asia": [
+      "Japan",
+      "Mongolia",
+      "North Korea",
+      "South Korea",
+      "Hong Kong",
+      "Macau",
+      "Taiwan",
+      "China"
+    ],
+    "South-eastern Asia": [
+      "Cambodia",
+      "Indonesia",
+      "Malaysia",
+      "Myanmar",
+      "Philippines",
+      "Singapore",
+      "Thailand",
+      "Timor-Leste",
+      "Vietnam",
+      "Laos",
+      "Brunei"
+    ],
+    "Southern Asia": [
+      "Afghanistan",
+      "Bangladesh",
+      "Bhutan",
+      "India",
+      "Maldives",
+      "Nepal",
+      "Pakistan",
+      "Sri Lanka",
+      "Iran"
+    ],
+    "Western Asia": [
+      "Armenia",
+      "Azerbaijan",
+      "Bahrain",
+      "Cyprus",
+      "Georgia",
+      "Iraq",
+      "Israel",
+      "Jordan",
+      "Kuwait",
+      "Lebanon",
+      "Oman",
+      "Qatar",
+      "Saudi Arabia",
+      "Turkey",
+      "United Arab Emirates",
+      "Yemen",
+      "Syria",
+      "Palestine"
+    ],
+    "Europe": [
+      "Eastern Europe",
+      "Northern Europe",
+      "Southern Europe",
+      "Western Europe"
+    ],
+    "Eastern Europe": [
+      "Belarus",
+      "Bulgaria",
+      "Hungary",
+      "Poland",
+      "Romania",
+      "Slovakia",
+      "Ukraine",
+      "Czech Republic",
+      "Russia",
+      "Moldova",
+      "Serbia"
+    ],
+    "Northern Europe": [
+      "\u00c5land Islands",
+      "Channel Islands",
+      "Denmark",
+      "Estonia",
+      "Faroe Islands",
+      "Finland",
+      "Iceland",
+      "Ireland",
+      "Isle of Man",
+      "Latvia",
+      "Lithuania",
+      "Norway",
+      "Svalbard and Jan Mayen Islands",
+      "Sweden",
+      "United Kingdom",
+      "Scotland",
+      "England",
+      "Northern Ireland",
+      "Wales"
+    ],
+    "Channel Islands": [
+      "Guernsey",
+      "Jersey",
+      "Sark"
+    ],
+    "Southern Europe": [
+      "Albania",
+      "Andorra",
+      "Bosnia and Herzegovina",
+      "Croatia",
+      "Gibraltar",
+      "Greece",
+      "Holy See",
+      "Malta",
+      "Montenegro",
+      "North Macedonia",
+      "San Marino",
+      "Slovenia",
+      "Spain",
+      "Azores",
+      "Vatican City",
+      "Madeira",
+      "Kosovo",
+      "Italy",
+      "Northern Cyprus",
+      "Portugal"
+    ],
+    "Western Europe": [
+      "Austria",
+      "Belgium",
+      "France",
+      "Germany",
+      "Liechtenstein",
+      "Luxembourg",
+      "Monaco",
+      "Netherlands",
+      "Switzerland"
+    ],
+    "Oceania": [
+      "Australia and New Zealand",
+      "Pacific Islands"
+    ],
+    "Australia and New Zealand": [
+      "Australia",
+      "Christmas Island",
+      "Cocos (Keeling) Islands",
+      "Heard Island and McDonald Islands",
+      "New Zealand",
+      "Norfolk Island"
+    ],
+    "Melanesia": [
+      "Fiji",
+      "New Caledonia",
+      "Papua New Guinea",
+      "Solomon Islands",
+      "Vanuatu"
+    ],
+    "Micronesia": [
+      "Guam",
+      "Kiribati",
+      "Marshall Islands",
+      "Nauru",
+      "Northern Mariana Islands",
+      "Palau",
+      "United States Minor Outlying Islands",
+      "Federated States of Micronesia"
+    ],
+    "Polynesia": [
+      "American Samoa",
+      "Cook Islands",
+      "French Polynesia",
+      "Niue",
+      "Samoa",
+      "Tokelau",
+      "Tonga",
+      "Tuvalu",
+      "Wallis and Futuna Islands",
+      "Pitcairn Islands"
+    ],
+    "Pacific Islands": [
+      "Melanesia",
+      "Micronesia",
+      "Polynesia"
+    ],
+    "Middle East and North Africa": [
+      "Algeria",
+      "Bahrain",
+      "Djibouti",
+      "Egypt",
+      "Iraq",
+      "Israel",
+      "Jordan",
+      "Kuwait",
+      "Lebanon",
+      "Libya",
+      "Malta",
+      "Mauritania",
+      "Morocco",
+      "Oman",
+      "Palestine",
+      "Qatar",
+      "Saudi Arabia",
+      "Somalia",
+      "Sudan",
+      "Syria",
+      "Tunisia",
+      "United Arab Emirates",
+      "Western Sahara",
+      "Yemen"
+    ]
+  }
+]

resources/file_formats.json

@@ -0,0 +1,164 @@
+{
+    "Text": {
+        ".CSV": "Comma-Separated Values File",
+        ".DOC": "Microsoft Word Document (Legacy)",
+        ".DOCX": "Microsoft Word Document",
+        ".ODT": "OpenDocument Text Document",
+        ".PAGES": "Apple Pages Document",
+        ".PDF": "Portable Document Format File",
+        ".RTF": "Rich Text Format File",
+        ".TEX": "LaTeX Source Document",
+        ".TXT": "Plain Text File",
+        ".WPD": "WordPerfect Document",
+        ".WPS": "Microsoft Works Document"
+    },
+    "Data": {
+        ".CSV" : "Comma-Separated Values File",
+        ".DAT" : "Data File",
+        ".GED" : "GEDCOM Genealogy Data File",
+        ".JSON": "JavaScript Object Notation File",
+        ".KEY" : "Apple Keynote Presentation",
+        ".KEYCHAIN" : "Mac OS X Keychain File",
+        ".NUMBERS" : "Apple Numbers Spreadsheet",
+        ".ODS": "OpenDocument Spreadsheet File",
+        ".ODP": "OpenDocument Presentation File",
+        ".PPT" : "Microsoft PowerPoint Presentation (Legacy)",
+        ".PPTX" : "Microsoft PowerPoint Presentation",
+        ".SDF" : "Standard Data File",
+        ".TAR" : "Consolidated Unix File Archive",
+        ".TOML" : "TOML (TOML Language) File",
+        ".VCF" : "vCard File",
+        ".XLR" : "Works Spreadsheet",
+        ".XLS": "Microsoft Excel Spreadsheet (Legacy)",
+        ".XLSX": "Microsoft Excel Spreadsheet",
+        ".XML" : "XML File",
+        ".YAML" : "YAML File"
+    },
+    "Database": {
+        ".ACCDB" : "Microsoft Access Database",
+        ".DB": "Database File",
+        ".DBF": "dBASE File",
+        ".FDB": "Firebird Database File",
+        ".MDB": "Microsoft Access Database File",
+        ".SQL": "SQL Database File",
+        ".SQLITE": "SQLite Database File",
+        ".TDB": "SQLite Database File",
+        ".VSDB": "Visual Studio Database File",
+        ".XML": "XML Database File"
+    },
+    "Web": {
+        ".ASP" : "Active Server Page",
+        ".ASPX" : "Active Server Page Extended File",
+        ".CER" : "Internet Security Certificate",
+        ".CFM" : "ColdFusion Markup File",
+        ".CRDOWNLOAD" : "Chrome Partially Downloaded File",
+        ".CSR" : "Certificate Signing Request File",
+        ".CSS" : "Cascading Style Sheet",
+        ".DCR" : "Shockwave Media File",
+        ".HTM" : "Hypertext Markup Language File",
+        ".HTML" : "Hypertext Markup Language File",
+        ".JS" : "JavaScript File",
+        ".JSP" : "Java Server Page",
+        ".PHP" : "PHP Source Code File",
+        ".RSS" : "Rich Site Summary",
+        ".XHTML" : "Extensible Hypertext Markup Language File"
+    },
+    "Audio": {
+        ".AIF" : "Audio Interchange File",
+        ".AIFF" : "Audio Interchange File",
+        ".IFF" : "Audio Interchange File",
+        ".AU" : "Sun Audio File",
+        ".CAF" : "Core Audio File",
+        ".CDA" : "Compact Disc Audio File",
+        ".FLAC" : "Free Lossless Audio Codec File",
+        ".M4A" : "MPEG-4 Audio File",
+        ".M4B" : "MPEG-4 Audio Book File",
+        ".M4P" : "MPEG-4 Protected Audio File",
+        ".M4R" : "MPEG-4 Ringtone File",
+        ".M4V" : "MPEG-4 Video File",
+        ".MID" : "MIDI File",
+        ".MKA" : "Matroska Audio File",
+        ".MP3" : "MPEG-1 Audio Layer 3 File",
+        ".MP4" : "MPEG-4 Audio File",
+        ".MPC" : "Musepack Audio File",
+        ".OGG" : "Ogg Vorbis File",
+        ".RA" : "Real Audio File",
+        ".RAM" : "Real Media Audio File",
+        ".WAV" : "Waveform Audio File",
+        ".WMA" : "Windows Media Audio File"
+    },
+    "Video": {
+        ".3G2" : "3GPP2 Multimedia File",
+        ".3GP" : "3GPP Multimedia File",
+        ".ASF" : "Advanced Systems Format File",
+        ".AVI" : "Audio Video Interleave File",
+        ".FLV" : "Flash Video File",
+        ".M4V" : "iTunes Video File",
+        ".MOV" : "Apple QuickTime Movie",
+        ".MP4" : "MPEG-4 Video File",
+        ".MPG" : "MPEG Video File",
+        ".RM" : "RealMedia File",
+        ".SRT" : "SubRip Subtitle File",
+        ".SWF" : "Shockwave Flash Movie",
+        ".VOB" : "DVD Video Object File",
+        ".WMV" : "Windows Media Video"
+    },
+    "Image": {
+        ".AI" : "Adobe Illustrator File",
+        ".BMP" : "Bitmap Image",
+        ".CUR" : "Windows Cursor File",
+        ".DCX" : "ZSoft IBM PC Multi-page Paintbrush File",
+        ".DDS" : "DirectDraw Surface",
+        ".EPS" : "Encapsulated PostScript File",
+        ".GIF" : "Graphics Interchange Format File",
+        ".HEIC" : "High Efficiency Image File Format",
+        ".ICO" : "Windows Icon File",
+        ".JP2" : "JPEG 2000 File",
+        ".JPEG2" : "JPEG 2000 File",
+        ".JPG" : "JPEG Image",
+        ".JPEG" : "JPEG Image",
+        ".PBM" : "Portable Bitmap Image",
+        ".PCT" : "Macintosh Picture File",
+        ".PCX" : "PC Paintbrush Exchange File",
+        ".PGM" : "Portable Graymap Image",
+        ".PNG" : "Portable Network Graphics File",
+        ".PPM" : "Portable Pixmap Image",
+        ".PSD" : "Adobe Photoshop File",
+        ".PSP" : "Paint Shop Pro Image",
+        ".SVG" : "Scalable Vector Graphics File",
+        ".TGA" : "Truevision Targa File",
+        ".THM" : "Microsoft Thumbnail File",
+        ".TIF" : "Tagged Image File",
+        ".TIFF" : "Tagged Image File",
+        ".XBM" : "X11 Bitmap Image",
+        ".XPM" : "X11 Pixmap Image",
+        ".YUB" : "YUV Encoded Image File"
+    },
+    "Compressed" : {
+        ".7Z" : "7-Zip Compressed File",
+        ".ACE" : "WinZip Compressed File",
+        ".ARC" : "ARJ Compressed File",
+        ".ARJ" : "ARJ Compressed File",
+        ".BZ" : "BZip Compressed File",
+        ".BZ2" : "BZip2 Compressed File",
+        ".CAB" : "Microsoft Cabinet File",
+        ".CRB" : "Comic Book RAR Archive",
+        ".DMG" : "Apple Disk Image",
+        ".DEB" : "Debian Software Package",
+        ".GZ" : "GZip Compressed File",
+        ".GZIP" : "GZip Compressed File",
+        ".LZH" : "LZH Compressed File",
+        ".LZIP" : "LZip Compressed File",
+        ".LZMA" : "LZMA Compressed File",
+        ".LZOP" : "LZOP Compressed File",
+        ".PKG" : "Mac OS X Installer Package",
+        ".RAR" : "RAR Compressed File",
+        ".RPM" : "Red Hat Package Manager File",
+        ".TAR.GZ" : "Compressed Tarball File",
+        ".TGZ" : "Tar Compressed File",
+        ".Z" : "Unix Compressed File",
+        ".ZIP" : "ZIP Compressed File",
+        ".ZIPX" : "ZIPX Compressed File",
+        ".ZOO" : "ZOO Compressed File"
+    }
+}

resources/language_lists.json

@@ -0,0 +1,86 @@
+{
+  "language_groups": {
+    "Niger-Congo": "African languages of the Niger-Congo family, incl. Bantu languages",
+    "Arabic": "Arabic",
+    "Basque": "Basque",
+    "Catalan": "Catalan",
+    "Chinese": "Chinese",
+    "English": "English",
+    "French": "French",
+    "Indic": "Indic languages, incl. Bengali, Hindi, Urdu...",
+    "Indonesian": "Indonesian",
+    "Portuguese": "Portuguese",
+    "Spanish": "Spanish",
+    "Vietnamese": "Vietnamese",
+    "Programming Language": "Programming Language"
+  },
+  "niger_congo_languages": [
+    "Akan",
+    "Bambara",
+    "Chi Chewa",
+    "ChiShona",
+    "ChiTumbuka",
+    "Fon",
+    "Igbo",
+    "isiZulu",
+    "Kinyarwanda",
+    "Kikongo",
+    "Kikuyu",
+    "Kirundi",
+    "Lingala",
+    "Luganda",
+    "Northern Sotho",
+    "Sesotho",
+    "Setswana",
+    "Swahili",
+    "Twi",
+    "Wolof",
+    "Xhosa",
+    "Xitsonga",
+    "Yoruba"
+  ],
+  "indic_languages": [
+    "Assamese",
+    "Bengali",
+    "Gujarati",
+    "Hindi",
+    "Kannada",
+    "Malayalam",
+    "Marathi",
+    "Odia",
+    "Punjabi",
+    "Telugu",
+    "Tamil",
+    "Urdu"
+  ],
+  "arabic": {
+    "ar-CLS": "(Arabic (Classic))",
+    "ar-MSA": "(Arabic (Modern Standard Arabic))",
+    "ar-LEV": "(Arabic (Levant))",
+    "ar-NOR": "(Arabic (North Africa))",
+    "ar-GLF": "(Arabic (Gulf))",
+    "ar-AE": "(Arabic (United Arab Emirates))",
+    "ar-BH": "(Arabic (Bahrain))",
+    "ar-DJ": "(Arabic (Djibouti))",
+    "ar-DZ": "(Arabic (Algeria))",
+    "ar-EG": "(Arabic (Egypt))",
+    "ar-IQ": "(Arabic (Iraq))",
+    "ar-JO": "(Arabic (Jordan))",
+    "ar-KM": "(Arabic (Comoros))",
+    "ar-KW": "(Arabic (Kuwait))",
+    "ar-LB": "(Arabic (Lebanon))",
+    "ar-LY": "(Arabic (Libya))",
+    "ar-MA": "(Arabic (Morocco))",
+    "ar-MR": "(Arabic (Mauritania))",
+    "ar-OM": "(Arabic (Oman))",
+    "ar-PS": "(Arabic (Palestinian Territories))",
+    "ar-QA": "(Arabic (Qatar))",
+    "ar-SA": "(Arabic (Saudi Arabia))",
+    "ar-SD": "(Arabic (Sudan))",
+    "ar-SO": "(Arabic (Somalia))",
+    "ar-SS": "(Arabic (South Sudan))",
+    "ar-SY": "(Arabic (Syria))",
+    "ar-TN": "(Arabic (Tunisia))",
+    "ar-YE": "(Arabic (Yemen))"
+  }
+}

resources/licenses.json

@@ -0,0 +1,452 @@
+[
+    "other: Other license",
+    "unknown: License information unavailable",
+    "0bsd: BSD Zero Clause License",
+    "aal: Attribution Assurance License",
+    "abstyles: Abstyles License",
+    "adobe-2006: Adobe Systems Incorporated Source Code License Agreement",
+    "adobe-glyph: Adobe Glyph List License",
+    "adsl: Amazon Digital Services License",
+    "afl-1.1: Academic Free License v1.1",
+    "afl-1.2: Academic Free License v1.2",
+    "afl-2.0: Academic Free License v2.0",
+    "afl-2.1: Academic Free License v2.1",
+    "afl-3.0: Academic Free License v3.0",
+    "afmparse: Afmparse License",
+    "agpl-1.0: Affero General Public License v1.0",
+    "agpl-1.0-only: Affero General Public License v1.0 only",
+    "agpl-1.0-or-later: Affero General Public License v1.0 or later",
+    "agpl-3.0: GNU Affero General Public License v3.0",
+    "agpl-3.0-only: GNU Affero General Public License v3.0 only",
+    "agpl-3.0-or-later: GNU Affero General Public License v3.0 or later",
+    "aladdin: Aladdin Free Public License",
+    "amdplpa: AMD's plpa_map.c License",
+    "aml: Apple MIT License",
+    "ampas: Academy of Motion Picture Arts and Sciences BSD",
+    "antlr-pd: ANTLR Software Rights Notice",
+    "antlr-pd-fallback: ANTLR Software Rights Notice with license fallback",
+    "apache-1.0: Apache License 1.0",
+    "apache-1.1: Apache License 1.1",
+    "apache-2.0: Apache License 2.0",
+    "apafml: Adobe Postscript AFM License",
+    "apl-1.0: Adaptive Public License 1.0",
+    "apsl-1.0: Apple Public Source License 1.0",
+    "apsl-1.1: Apple Public Source License 1.1",
+    "apsl-1.2: Apple Public Source License 1.2",
+    "apsl-2.0: Apple Public Source License 2.0",
+    "artistic-1.0: Artistic License 1.0",
+    "artistic-1.0-cl8: Artistic License 1.0 w/clause 8",
+    "artistic-1.0-perl: Artistic License 1.0 (Perl)",
+    "artistic-2.0: Artistic License 2.0",
+    "bahyph: Bahyph License",
+    "barr: Barr License",
+    "beerware: Beerware License",
+    "bittorrent-1.0: BitTorrent Open Source License v1.0",
+    "bittorrent-1.1: BitTorrent Open Source License v1.1",
+    "blessing: SQLite Blessing",
+    "blueoak-1.0.0: Blue Oak Model License 1.0.0",
+    "borceux: Borceux license",
+    "bsd-1-clause: BSD 1-Clause License",
+    "bsd-2-clause: BSD 2-Clause \"Simplified\" License",
+    "bsd-2-clause-freebsd: BSD 2-Clause FreeBSD License",
+    "bsd-2-clause-netbsd: BSD 2-Clause NetBSD License",
+    "bsd-2-clause-patent: BSD-2-Clause Plus Patent License",
+    "bsd-2-clause-views: BSD 2-Clause with views sentence",
+    "bsd-3-clause: BSD 3-Clause \"New\" or \"Revised\" License",
+    "bsd-3-clause-attribution: BSD with attribution",
+    "bsd-3-clause-clear: BSD 3-Clause Clear License",
+    "bsd-3-clause-lbnl: Lawrence Berkeley National Labs BSD variant license",
+    "bsd-3-clause-no-nuclear-license: BSD 3-Clause No Nuclear License",
+    "bsd-3-clause-no-nuclear-license-2014: BSD 3-Clause No Nuclear License 2014",
+    "bsd-3-clause-no-nuclear-warranty: BSD 3-Clause No Nuclear Warranty",
+    "bsd-3-clause-open-mpi: BSD 3-Clause Open MPI variant",
+    "bsd-4-clause: BSD 4-Clause \"Original\" or \"Old\" License",
+    "bsd-4-clause-uc: BSD-4-Clause (University of California-Specific)",
+    "bsd-protection: BSD Protection License",
+    "bsd-source-code: BSD Source Code Attribution",
+    "bsl-1.0: Boost Software License 1.0",
+    "busl-1.1: Business Source License 1.1",
+    "bzip2-1.0.5: bzip2 and libbzip2 License v1.0.5",
+    "bzip2-1.0.6: bzip2 and libbzip2 License v1.0.6",
+    "cal-1.0: Cryptographic Autonomy License 1.0",
+    "cal-1.0-combined-work-exception: Cryptographic Autonomy License 1.0 (Combined Work Exception)",
+    "caldera: Caldera License",
+    "catosl-1.1: Computer Associates Trusted Open Source License 1.1",
+    "cc-by-1.0: Creative Commons Attribution 1.0 Generic",
+    "cc-by-2.0: Creative Commons Attribution 2.0 Generic",
+    "cc-by-2.5: Creative Commons Attribution 2.5 Generic",
+    "cc-by-3.0: Creative Commons Attribution 3.0 Unported",
+    "cc-by-3.0-at: Creative Commons Attribution 3.0 Austria",
+    "cc-by-3.0-us: Creative Commons Attribution 3.0 United States",
+    "cc-by-4.0: Creative Commons Attribution 4.0 International",
+    "cc-by-nc-1.0: Creative Commons Attribution Non Commercial 1.0 Generic",
+    "cc-by-nc-2.0: Creative Commons Attribution Non Commercial 2.0 Generic",
+    "cc-by-nc-2.5: Creative Commons Attribution Non Commercial 2.5 Generic",
+    "cc-by-nc-3.0: Creative Commons Attribution Non Commercial 3.0 Unported",
+    "cc-by-nc-4.0: Creative Commons Attribution Non Commercial 4.0 International",
+    "cc-by-nc-nd-1.0: Creative Commons Attribution Non Commercial No Derivatives 1.0 Generic",
+    "cc-by-nc-nd-2.0: Creative Commons Attribution Non Commercial No Derivatives 2.0 Generic",
+    "cc-by-nc-nd-2.5: Creative Commons Attribution Non Commercial No Derivatives 2.5 Generic",
+    "cc-by-nc-nd-3.0: Creative Commons Attribution Non Commercial No Derivatives 3.0 Unported",
+    "cc-by-nc-nd-3.0-igo: Creative Commons Attribution Non Commercial No Derivatives 3.0 IGO",
+    "cc-by-nc-nd-4.0: Creative Commons Attribution Non Commercial No Derivatives 4.0 International",
+    "cc-by-nc-sa-1.0: Creative Commons Attribution Non Commercial Share Alike 1.0 Generic",
+    "cc-by-nc-sa-2.0: Creative Commons Attribution Non Commercial Share Alike 2.0 Generic",
+    "cc-by-nc-sa-2.5: Creative Commons Attribution Non Commercial Share Alike 2.5 Generic",
+    "cc-by-nc-sa-3.0: Creative Commons Attribution Non Commercial Share Alike 3.0 Unported",
+    "cc-by-nc-sa-4.0: Creative Commons Attribution Non Commercial Share Alike 4.0 International",
+    "cc-by-nd-1.0: Creative Commons Attribution No Derivatives 1.0 Generic",
+    "cc-by-nd-2.0: Creative Commons Attribution No Derivatives 2.0 Generic",
+    "cc-by-nd-2.5: Creative Commons Attribution No Derivatives 2.5 Generic",
+    "cc-by-nd-3.0: Creative Commons Attribution No Derivatives 3.0 Unported",
+    "cc-by-nd-4.0: Creative Commons Attribution No Derivatives 4.0 International",
+    "cc-by-sa-1.0: Creative Commons Attribution Share Alike 1.0 Generic",
+    "cc-by-sa-2.0: Creative Commons Attribution Share Alike 2.0 Generic",
+    "cc-by-sa-2.0-uk: Creative Commons Attribution Share Alike 2.0 England and Wales",
+    "cc-by-sa-2.5: Creative Commons Attribution Share Alike 2.5 Generic",
+    "cc-by-sa-3.0: Creative Commons Attribution Share Alike 3.0 Unported",
+    "cc-by-sa-3.0-at: Creative Commons Attribution-Share Alike 3.0 Austria",
+    "cc-by-sa-4.0: Creative Commons Attribution Share Alike 4.0 International",
+    "cc-pddc: Creative Commons Public Domain Dedication and Certification",
+    "cc0-1.0: Creative Commons Zero v1.0 Universal",
+    "cddl-1.0: Common Development and Distribution License 1.0",
+    "cddl-1.1: Common Development and Distribution License 1.1",
+    "cdla-permissive-1.0: Community Data License Agreement Permissive 1.0",
+    "cdla-sharing-1.0: Community Data License Agreement Sharing 1.0",
+    "cecill-1.0: CeCILL Free Software License Agreement v1.0",
+    "cecill-1.1: CeCILL Free Software License Agreement v1.1",
+    "cecill-2.0: CeCILL Free Software License Agreement v2.0",
+    "cecill-2.1: CeCILL Free Software License Agreement v2.1",
+    "cecill-b: CeCILL-B Free Software License Agreement",
+    "cecill-c: CeCILL-C Free Software License Agreement",
+    "cern-ohl-1.1: CERN Open Hardware Licence v1.1",
+    "cern-ohl-1.2: CERN Open Hardware Licence v1.2",
+    "cern-ohl-p-2.0: CERN Open Hardware Licence Version 2 - Permissive",
+    "cern-ohl-s-2.0: CERN Open Hardware Licence Version 2 - Strongly Reciprocal",
+    "cern-ohl-w-2.0: CERN Open Hardware Licence Version 2 - Weakly Reciprocal",
+    "clartistic: Clarified Artistic License",
+    "cnri-jython: CNRI Jython License",
+    "cnri-python: CNRI Python License",
+    "cnri-python-gpl-compatible: CNRI Python Open Source GPL Compatible License Agreement",
+    "condor-1.1: Condor Public License v1.1",
+    "copyleft-next-0.3.0: copyleft-next 0.3.0",
+    "copyleft-next-0.3.1: copyleft-next 0.3.1",
+    "cpal-1.0: Common Public Attribution License 1.0",
+    "cpl-1.0: Common Public License 1.0",
+    "cpol-1.02: Code Project Open License 1.02",
+    "crossword: Crossword License",
+    "crystalstacker: CrystalStacker License",
+    "cua-opl-1.0: CUA Office Public License v1.0",
+    "cube: Cube License",
+    "curl: curl License",
+    "d-fsl-1.0: Deutsche Freie Software Lizenz",
+    "diffmark: diffmark license",
+    "doc: DOC License",
+    "dotseqn: Dotseqn License",
+    "dsdp: DSDP License",
+    "dvipdfm: dvipdfm License",
+    "ecl-1.0: Educational Community License v1.0",
+    "ecl-2.0: Educational Community License v2.0",
+    "ecos-2.0: eCos license version 2.0",
+    "efl-1.0: Eiffel Forum License v1.0",
+    "efl-2.0: Eiffel Forum License v2.0",
+    "egenix: eGenix.com Public License 1.1.0",
+    "entessa: Entessa Public License v1.0",
+    "epics: EPICS Open License",
+    "epl-1.0: Eclipse Public License 1.0",
+    "epl-2.0: Eclipse Public License 2.0",
+    "erlpl-1.1: Erlang Public License v1.1",
+    "etalab-2.0: Etalab Open License 2.0",
+    "eudatagrid: EU DataGrid Software License",
+    "eupl-1.0: European Union Public License 1.0",
+    "eupl-1.1: European Union Public License 1.1",
+    "eupl-1.2: European Union Public License 1.2",
+    "eurosym: Eurosym License",
+    "fair: Fair License",
+    "frameworx-1.0: Frameworx Open License 1.0",
+    "freeimage: FreeImage Public License v1.0",
+    "fsfap: FSF All Permissive License",
+    "fsful: FSF Unlimited License",
+    "fsfullr: FSF Unlimited License (with License Retention)",
+    "ftl: Freetype Project License",
+    "gfdl-1.1: GNU Free Documentation License v1.1",
+    "gfdl-1.1-invariants-only: GNU Free Documentation License v1.1 only - invariants",
+    "gfdl-1.1-invariants-or-later: GNU Free Documentation License v1.1 or later - invariants",
+    "gfdl-1.1-no-invariants-only: GNU Free Documentation License v1.1 only - no invariants",
+    "gfdl-1.1-no-invariants-or-later: GNU Free Documentation License v1.1 or later - no invariants",
+    "gfdl-1.1-only: GNU Free Documentation License v1.1 only",
+    "gfdl-1.1-or-later: GNU Free Documentation License v1.1 or later",
+    "gfdl-1.2: GNU Free Documentation License v1.2",
+    "gfdl-1.2-invariants-only: GNU Free Documentation License v1.2 only - invariants",
+    "gfdl-1.2-invariants-or-later: GNU Free Documentation License v1.2 or later - invariants",
+    "gfdl-1.2-no-invariants-only: GNU Free Documentation License v1.2 only - no invariants",
+    "gfdl-1.2-no-invariants-or-later: GNU Free Documentation License v1.2 or later - no invariants",
+    "gfdl-1.2-only: GNU Free Documentation License v1.2 only",
+    "gfdl-1.2-or-later: GNU Free Documentation License v1.2 or later",
+    "gfdl-1.3: GNU Free Documentation License v1.3",
+    "gfdl-1.3-invariants-only: GNU Free Documentation License v1.3 only - invariants",
+    "gfdl-1.3-invariants-or-later: GNU Free Documentation License v1.3 or later - invariants",
+    "gfdl-1.3-no-invariants-only: GNU Free Documentation License v1.3 only - no invariants",
+    "gfdl-1.3-no-invariants-or-later: GNU Free Documentation License v1.3 or later - no invariants",
+    "gfdl-1.3-only: GNU Free Documentation License v1.3 only",
+    "gfdl-1.3-or-later: GNU Free Documentation License v1.3 or later",
+    "giftware: Giftware License",
+    "gl2ps: GL2PS License",
+    "glide: 3dfx Glide License",
+    "glulxe: Glulxe License",
+    "glwtpl: Good Luck With That Public License",
+    "gnuplot: gnuplot License",
+    "gpl-1.0: GNU General Public License v1.0 only",
+    "gpl-1.0+: GNU General Public License v1.0 or later",
+    "gpl-1.0-only: GNU General Public License v1.0 only",
+    "gpl-1.0-or-later: GNU General Public License v1.0 or later",
+    "gpl-2.0: GNU General Public License v2.0 only",
+    "gpl-2.0+: GNU General Public License v2.0 or later",
+    "gpl-2.0-only: GNU General Public License v2.0 only",
+    "gpl-2.0-or-later: GNU General Public License v2.0 or later",
+    "gpl-2.0-with-autoconf-exception: GNU General Public License v2.0 w/Autoconf exception",
+    "gpl-2.0-with-bison-exception: GNU General Public License v2.0 w/Bison exception",
+    "gpl-2.0-with-classpath-exception: GNU General Public License v2.0 w/Classpath exception",
+    "gpl-2.0-with-font-exception: GNU General Public License v2.0 w/Font exception",
+    "gpl-2.0-with-gcc-exception: GNU General Public License v2.0 w/GCC Runtime Library exception",
+    "gpl-3.0: GNU General Public License v3.0 only",
+    "gpl-3.0+: GNU General Public License v3.0 or later",
+    "gpl-3.0-only: GNU General Public License v3.0 only",
+    "gpl-3.0-or-later: GNU General Public License v3.0 or later",
+    "gpl-3.0-with-autoconf-exception: GNU General Public License v3.0 w/Autoconf exception",
+    "gpl-3.0-with-gcc-exception: GNU General Public License v3.0 w/GCC Runtime Library exception",
+    "gsoap-1.3b: gSOAP Public License v1.3b",
+    "haskellreport: Haskell Language Report License",
+    "hippocratic-2.1: Hippocratic License 2.1",
+    "hpnd: Historical Permission Notice and Disclaimer",
+    "hpnd-sell-variant: Historical Permission Notice and Disclaimer - sell variant",
+    "htmltidy: HTML Tidy License",
+    "ibm-pibs: IBM PowerPC Initialization and Boot Software",
+    "icu: ICU License",
+    "ijg: Independent JPEG Group License",
+    "imagemagick: ImageMagick License",
+    "imatix: iMatix Standard Function Library Agreement",
+    "imlib2: Imlib2 License",
+    "info-zip: Info-ZIP License",
+    "intel: Intel Open Source License",
+    "intel-acpi: Intel ACPI Software License Agreement",
+    "interbase-1.0: Interbase Public License v1.0",
+    "ipa: IPA Font License",
+    "ipl-1.0: IBM Public License v1.0",
+    "isc: ISC License",
+    "jasper-2.0: JasPer License",
+    "jpnic: Japan Network Information Center License",
+    "json: JSON License",
+    "lal-1.2: Licence Art Libre 1.2",
+    "lal-1.3: Licence Art Libre 1.3",
+    "latex2e: Latex2e License",
+    "leptonica: Leptonica License",
+    "lgpl-2.0: GNU Library General Public License v2 only",
+    "lgpl-2.0+: GNU Library General Public License v2 or later",
+    "lgpl-2.0-only: GNU Library General Public License v2 only",
+    "lgpl-2.0-or-later: GNU Library General Public License v2 or later",
+    "lgpl-2.1: GNU Lesser General Public License v2.1 only",
+    "lgpl-2.1+: GNU Library General Public License v2.1 or later",
+    "lgpl-2.1-only: GNU Lesser General Public License v2.1 only",
+    "lgpl-2.1-or-later: GNU Lesser General Public License v2.1 or later",
+    "lgpl-3.0: GNU Lesser General Public License v3.0 only",
+    "lgpl-3.0+: GNU Lesser General Public License v3.0 or later",
+    "lgpl-3.0-only: GNU Lesser General Public License v3.0 only",
+    "lgpl-3.0-or-later: GNU Lesser General Public License v3.0 or later",
+    "lgpllr: Lesser General Public License For Linguistic Resources",
+    "libpng: libpng License",
+    "libpng-2.0: PNG Reference Library version 2",
+    "libselinux-1.0: libselinux public domain notice",
+    "libtiff: libtiff License",
+    "liliq-p-1.1: Licence Libre du Qu\u00e9bec \u2013 Permissive version 1.1",
+    "liliq-r-1.1: Licence Libre du Qu\u00e9bec \u2013 R\u00e9ciprocit\u00e9 version 1.1",
+    "liliq-rplus-1.1: Licence Libre du Qu\u00e9bec \u2013 R\u00e9ciprocit\u00e9 forte version 1.1",
+    "linux-openib: Linux Kernel Variant of OpenIB.org license",
+    "lpl-1.0: Lucent Public License Version 1.0",
+    "lpl-1.02: Lucent Public License v1.02",
+    "lppl-1.0: LaTeX Project Public License v1.0",
+    "lppl-1.1: LaTeX Project Public License v1.1",
+    "lppl-1.2: LaTeX Project Public License v1.2",
+    "lppl-1.3a: LaTeX Project Public License v1.3a",
+    "lppl-1.3c: LaTeX Project Public License v1.3c",
+    "makeindex: MakeIndex License",
+    "miros: The MirOS Licence",
+    "mit: MIT License",
+    "mit-0: MIT No Attribution",
+    "mit-advertising: Enlightenment License (e16)",
+    "mit-cmu: CMU License",
+    "mit-enna: enna License",
+    "mit-feh: feh License",
+    "mit-open-group: MIT Open Group variant",
+    "mitnfa: MIT +no-false-attribs license",
+    "motosoto: Motosoto License",
+    "mpich2: mpich2 License",
+    "mpl-1.0: Mozilla Public License 1.0",
+    "mpl-1.1: Mozilla Public License 1.1",
+    "mpl-2.0: Mozilla Public License 2.0",
+    "mpl-2.0-no-copyleft-exception: Mozilla Public License 2.0 (no copyleft exception)",
+    "ms-pl: Microsoft Public License",
+    "ms-rl: Microsoft Reciprocal License",
+    "mtll: Matrix Template Library License",
+    "mulanpsl-1.0: Mulan Permissive Software License, Version 1",
+    "mulanpsl-2.0: Mulan Permissive Software License, Version 2",
+    "multics: Multics License",
+    "mup: Mup License",
+    "nasa-1.3: NASA Open Source Agreement 1.3",
+    "naumen: Naumen Public License",
+    "nbpl-1.0: Net Boolean Public License v1",
+    "ncgl-uk-2.0: Non-Commercial Government Licence",
+    "ncsa: University of Illinois/NCSA Open Source License",
+    "net-snmp: Net-SNMP License",
+    "netcdf: NetCDF license",
+    "newsletr: Newsletr License",
+    "ngpl: Nethack General Public License",
+    "nist-pd: NIST Public Domain Notice",
+    "nist-pd-fallback: NIST Public Domain Notice with license fallback",
+    "nlod-1.0: Norwegian Licence for Open Government Data",
+    "nlpl: No Limit Public License",
+    "nokia: Nokia Open Source License",
+    "nosl: Netizen Open Source License",
+    "noweb: Noweb License",
+    "npl-1.0: Netscape Public License v1.0",
+    "npl-1.1: Netscape Public License v1.1",
+    "nposl-3.0: Non-Profit Open Software License 3.0",
+    "nrl: NRL License",
+    "ntp: NTP License",
+    "ntp-0: NTP No Attribution",
+    "nunit: Nunit License",
+    "o-uda-1.0: Open Use of Data Agreement v1.0",
+    "occt-pl: Open CASCADE Technology Public License",
+    "oclc-2.0: OCLC Research Public License 2.0",
+    "odbl-1.0: ODC Open Database License v1.0",
+    "odc-by-1.0: Open Data Commons Attribution License v1.0",
+    "ofl-1.0: SIL Open Font License 1.0",
+    "ofl-1.0-no-rfn: SIL Open Font License 1.0 with no Reserved Font Name",
+    "ofl-1.0-rfn: SIL Open Font License 1.0 with Reserved Font Name",
+    "ofl-1.1: SIL Open Font License 1.1",
+    "ofl-1.1-no-rfn: SIL Open Font License 1.1 with no Reserved Font Name",
+    "ofl-1.1-rfn: SIL Open Font License 1.1 with Reserved Font Name",
+    "ogc-1.0: OGC Software License, Version 1.0",
+    "ogl-canada-2.0: Open Government Licence - Canada",
+    "ogl-uk-1.0: Open Government Licence v1.0",
+    "ogl-uk-2.0: Open Government Licence v2.0",
+    "ogl-uk-3.0: Open Government Licence v3.0",
+    "ogtsl: Open Group Test Suite License",
+    "oldap-1.1: Open LDAP Public License v1.1",
+    "oldap-1.2: Open LDAP Public License v1.2",
+    "oldap-1.3: Open LDAP Public License v1.3",
+    "oldap-1.4: Open LDAP Public License v1.4",
+    "oldap-2.0: Open LDAP Public License v2.0 (or possibly 2.0A and 2.0B)",
+    "oldap-2.0.1: Open LDAP Public License v2.0.1",
+    "oldap-2.1: Open LDAP Public License v2.1",
+    "oldap-2.2: Open LDAP Public License v2.2",
+    "oldap-2.2.1: Open LDAP Public License v2.2.1",
+    "oldap-2.2.2: Open LDAP Public License 2.2.2",
+    "oldap-2.3: Open LDAP Public License v2.3",
+    "oldap-2.4: Open LDAP Public License v2.4",
+    "oldap-2.5: Open LDAP Public License v2.5",
+    "oldap-2.6: Open LDAP Public License v2.6",
+    "oldap-2.7: Open LDAP Public License v2.7",
+    "oldap-2.8: Open LDAP Public License v2.8",
+    "oml: Open Market License",
+    "openssl: OpenSSL License",
+    "opl-1.0: Open Public License v1.0",
+    "oset-pl-2.1: OSET Public License version 2.1",
+    "osl-1.0: Open Software License 1.0",
+    "osl-1.1: Open Software License 1.1",
+    "osl-2.0: Open Software License 2.0",
+    "osl-2.1: Open Software License 2.1",
+    "osl-3.0: Open Software License 3.0",
+    "parity-6.0.0: The Parity Public License 6.0.0",
+    "parity-7.0.0: The Parity Public License 7.0.0",
+    "pddl-1.0: ODC Public Domain Dedication & License 1.0",
+    "php-3.0: PHP License v3.0",
+    "php-3.01: PHP License v3.01",
+    "plexus: Plexus Classworlds License",
+    "polyform-noncommercial-1.0.0: PolyForm Noncommercial License 1.0.0",
+    "polyform-small-business-1.0.0: PolyForm Small Business License 1.0.0",
+    "postgresql: PostgreSQL License",
+    "psf-2.0: Python Software Foundation License 2.0",
+    "psfrag: psfrag License",
+    "psutils: psutils License",
+    "python-2.0: Python License 2.0",
+    "qhull: Qhull License",
+    "qpl-1.0: Q Public License 1.0",
+    "rdisc: Rdisc License",
+    "rhecos-1.1: Red Hat eCos Public License v1.1",
+    "rpl-1.1: Reciprocal Public License 1.1",
+    "rpl-1.5: Reciprocal Public License 1.5",
+    "rpsl-1.0: RealNetworks Public Source License v1.0",
+    "rsa-md: RSA Message-Digest License",
+    "rscpl: Ricoh Source Code Public License",
+    "ruby: Ruby License",
+    "sax-pd: Sax Public Domain Notice",
+    "saxpath: Saxpath License",
+    "scea: SCEA Shared Source License",
+    "sendmail: Sendmail License",
+    "sendmail-8.23: Sendmail License 8.23",
+    "sgi-b-1.0: SGI Free Software License B v1.0",
+    "sgi-b-1.1: SGI Free Software License B v1.1",
+    "sgi-b-2.0: SGI Free Software License B v2.0",
+    "shl-0.5: Solderpad Hardware License v0.5",
+    "shl-0.51: Solderpad Hardware License, Version 0.51",
+    "simpl-2.0: Simple Public License 2.0",
+    "sissl: Sun Industry Standards Source License v1.1",
+    "sissl-1.2: Sun Industry Standards Source License v1.2",
+    "sleepycat: Sleepycat License",
+    "smlnj: Standard ML of New Jersey License",
+    "smppl: Secure Messaging Protocol Public License",
+    "snia: SNIA Public License 1.1",
+    "spencer-86: Spencer License 86",
+    "spencer-94: Spencer License 94",
+    "spencer-99: Spencer License 99",
+    "spl-1.0: Sun Public License v1.0",
+    "ssh-openssh: SSH OpenSSH license",
+    "ssh-short: SSH short notice",
+    "sspl-1.0: Server Side Public License, v 1",
+    "standardml-nj: Standard ML of New Jersey License",
+    "sugarcrm-1.1.3: SugarCRM Public License v1.1.3",
+    "swl: Scheme Widget Library (SWL) Software License Agreement",
+    "tapr-ohl-1.0: TAPR Open Hardware License v1.0",
+    "tcl: TCL/TK License",
+    "tcp-wrappers: TCP Wrappers License",
+    "tmate: TMate Open Source License",
+    "torque-1.1: TORQUE v2.5+ Software License v1.1",
+    "tosl: Trusster Open Source License",
+    "tu-berlin-1.0: Technische Universitaet Berlin License 1.0",
+    "tu-berlin-2.0: Technische Universitaet Berlin License 2.0",
+    "ucl-1.0: Upstream Compatibility License v1.0",
+    "unicode-dfs-2015: Unicode License Agreement - Data Files and Software (2015)",
+    "unicode-dfs-2016: Unicode License Agreement - Data Files and Software (2016)",
+    "unicode-tou: Unicode Terms of Use",
+    "unlicense: The Unlicense",
+    "upl-1.0: Universal Permissive License v1.0",
+    "vim: Vim License",
+    "vostrom: VOSTROM Public License for Open Source",
+    "vsl-1.0: Vovida Software License v1.0",
+    "w3c: W3C Software Notice and License (2002-12-31)",
+    "w3c-19980720: W3C Software Notice and License (1998-07-20)",
+    "w3c-20150513: W3C Software Notice and Document License (2015-05-13)",
+    "watcom-1.0: Sybase Open Watcom Public License 1.0",
+    "wsuipa: Wsuipa License",
+    "wtfpl: Do What The F*ck You Want To Public License",
+    "wxwindows: wxWindows Library License",
+    "x11: X11 License",
+    "xerox: Xerox License",
+    "xfree86-1.1: XFree86 License 1.1",
+    "xinetd: xinetd License",
+    "xnet: X.Net License",
+    "xpp: XPP License",
+    "xskat: XSkat License",
+    "ypl-1.0: Yahoo! Public License v1.0",
+    "ypl-1.1: Yahoo! Public License v1.1",
+    "zed: Zed License",
+    "zend-2.0: Zend License v2.0",
+    "zimbra-1.3: Zimbra Public License v1.3",
+    "zimbra-1.4: Zimbra Public License v1.4",
+    "zlib: zlib License",
+    "zlib-acknowledgement: zlib/libpng License with Acknowledgement",
+    "zpl-1.1: Zope Public License 1.1",
+    "zpl-2.0: Zope Public License 2.0",
+    "zpl-2.1: Zope Public License 2.1"
+]

resources/pii_categories.json

@@ -0,0 +1,34 @@
+{
+  "generic": [
+    "names",
+    "physical addresses",
+    "email addresses",
+    "website account name or handle",
+    "dates (birth, death, etc.)",
+    "full-face photographs and comparable images",
+    "URLs",
+    "biometric identifiers (fingerprints, voice, etc.)"
+  ],
+  "sensitive": [
+    "racial or ethnic origin",
+    "political opinions",
+    "religious or philosophical beliefs",
+    "trade-union membership",
+    "genetic data",
+    "health-related data",
+    "data concerning a person's sex life or sexual orientation"
+  ],
+  "numbers": [
+    "telephone numbers",
+    "fax numbers",
+    "vehicle identifiers and serial numbers",
+    "device identifiers and serial numbers",
+    "social security numbers",
+    "IP addresses",
+    "medical record numbers",
+    "health plan beneficiary numbers",
+    "account numbers",
+    "uniquely identifying numbers",
+    "certificate/license numbers"
+  ]
+}

resources/primary_source_taxonomy.json

@@ -0,0 +1,23 @@
+{
+  "website": [
+    "wiki",
+    "content repository, archive, or collection",
+    "forum",
+    "news or magazine website",
+    "official website",
+    "social media",
+    "blog",
+    "other"
+  ],
+  "collection": [
+    "books/book publisher",
+    "scientific articles/journal",
+    "news articles",
+    "radio",
+    "clips",
+    "movies and documentaries",
+    "podcasts",
+    "other"
+  ],
+  "other": []
+}