hdx-python-country 3.7.1__tar.gz → 3.7.3__tar.gz

{hdx_python_country-3.7.1 → hdx_python_country-3.7.3}/.config/pre-commit-config.yaml

@@ -1,14 +1,14 @@
 default_language_version:
-    python: python3.11
+    python: python3.12
 repos:
   - repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: v4.5.0
+    rev: v4.6.0
     hooks:
       - id: trailing-whitespace
       - id: end-of-file-fixer
       - id: check-ast
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.4.4
+    rev: v0.4.8
     hooks:
       # Run the linter.
       - id: ruff
@@ -17,7 +17,7 @@ repos:
       - id: ruff-format
         args: [--config, .config/ruff.toml]
   - repo: https://github.com/astral-sh/uv-pre-commit
-    rev: 0.1.45
+    rev: 0.2.10
     hooks:
       # Run the pip compile
       - id: pip-compile

{hdx_python_country-3.7.1 → hdx_python_country-3.7.3}/.github/workflows/publish.yaml

@@ -21,7 +21,8 @@ jobs:
     - name: Install dependencies
       run: |
         python -m pip install --upgrade pip
-        pip install --upgrade hatch
+    - name: Install Hatch
+      uses: pypa/hatch@install
     - name: Build with hatch
       run: |
         hatch build

{hdx_python_country-3.7.1 → hdx_python_country-3.7.3}/.github/workflows/run-python-tests.yaml

@@ -26,17 +26,15 @@ jobs:
     - name: Install dependencies
       run: |
         python -m pip install --upgrade pip
-        pip install --upgrade hatch
+    - name: Install Hatch
+      uses: pypa/hatch@install
     - name: Test with hatch/pytest
-      env:
-        WFP_KEY: ${{ secrets.WFP_KEY }}
-        WFP_SECRET: ${{ secrets.WFP_SECRET }}
       run: |
-        hatch run test:test
+        hatch test
     - name: Check styling
       if: always()
       run: |
-        hatch run lint:style
+        hatch fmt --check
     - name: Publish Unit Test Results
       uses: EnricoMi/publish-unit-test-result-action@v2
       if: always()

{hdx_python_country-3.7.1 → hdx_python_country-3.7.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: hdx-python-country
-Version: 3.7.1
+Version: 3.7.3
 Summary: HDX Python country code and exchange rate (fx) utilities
 Project-URL: Homepage, https://github.com/OCHA-DAP/hdx-python-country
 Author: Michael Rans
@@ -25,9 +25,9 @@ Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Requires-Python: >=3.8
-Requires-Dist: hdx-python-utilities>=3.6.8
+Requires-Dist: hdx-python-utilities>=3.7.0
 Requires-Dist: libhxl>=5.2.1
-Requires-Dist: pyphonetics
+Requires-Dist: rapidfuzz
 Provides-Extra: dev
 Requires-Dist: pre-commit; extra == 'dev'
 Provides-Extra: test

{hdx_python_country-3.7.1 → hdx_python_country-3.7.3}/documentation/main.md

@@ -113,10 +113,19 @@ The administration level mappings takes input configuration dictionary,
 
 *admin_config* can have the following optional keys:
 
-*countries_fuzzy_try* are countries (iso3 codes) for which to try fuzzy matching. Default is all countries.
-*admin_name_mappings* is a dictionary of mappings from name to pcode (for where fuzzy matching fails)
-*admin_name_replacements* is a dictionary of textual replacements to try when fuzzy matching
-*admin_fuzzy_dont* is a list of names for which fuzzy matching should not be tried
+*countries_fuzzy_try* are countries (iso3 codes) for which to try fuzzy
+matching. Default is all countries.
+*admin_name_mappings* is a dictionary of mappings from name to pcode. These can
+be global or they can be restricted by country or parent (if the AdminLevel
+object has been set up with parents). Keys take the form "MAPPING",
+"AFG|MAPPING" or "AF01|MAPPING".
+*admin_name_replacements* is a dictionary of textual replacements to try when
+fuzzy matching. It maps from string to string replacement. The replacements can
+be global or they can be restricted by country or parent (if the AdminLevel
+object has been set up with parents). Keys take the form "STRING_TO_REPLACE",
+"AFG|STRING_TO_REPLACE" or "AF01|STRING_TO_REPLACE".
+*admin_fuzzy_dont* is a list of names for which fuzzy matching should not be
+tried
 
 Once an AdminLevel object is constructed, one of three setup methods must be
 called: *setup_from_admin_info*, *setup_from_libhxl_dataset* or

{hdx_python_country-3.7.1 → hdx_python_country-3.7.3}/pyproject.toml

@@ -34,9 +34,9 @@ classifiers = [
 requires-python = ">=3.8"
 
 dependencies = [
-    "hdx-python-utilities>=3.6.8",
+    "hdx-python-utilities>=3.7.0",
     "libhxl>=5.2.1",
-    "pyphonetics",
+    "rapidfuzz",
 ]
 dynamic = ["version"]
 
@@ -79,25 +79,21 @@ version_scheme = "python-simplified-semver"
 
 # Tests
 
-[tool.hatch.envs.test]
-features = ["test"]#, "wfp"]
+[tool.hatch.envs.hatch-test]
+features = ["test"]
 
-[tool.hatch.envs.test.scripts]
-test = """
+[[tool.hatch.envs.hatch-test.matrix]]
+python = ["3.12"]
+
+[tool.hatch.envs.hatch-test.scripts]
+run = """
        pytest -c .config/pytest.ini --rootdir=. --junitxml=test-results.xml \
        --cov --cov-config=.config/coveragerc --no-cov-on-fail \
        --cov-report=lcov --cov-report=term-missing
        """
 
-[[tool.hatch.envs.test.matrix]]
-python = ["3.12"]
-
-[tool.hatch.envs.lint]
-detached = true
-dependencies = ["ruff"]
-
-[tool.hatch.envs.lint.scripts]
-style = [
-  "ruff check --config .config/ruff.toml --diff {args:.}",
-  "ruff format --config .config/ruff.toml --diff {args:.}",
-]
+[tool.hatch.envs.hatch-static-analysis.scripts]
+format-check = ["ruff format --config .config/ruff.toml --check --diff {args:.}",]
+format-fix = ["ruff format --config .config/ruff.toml {args:.}",]
+lint-check = ["ruff check --config .config/ruff.toml {args:.}",]
+lint-fix = ["ruff check --config .config/ruff.toml --fix {args:.}",]

{hdx_python_country-3.7.1 → hdx_python_country-3.7.3}/requirements.txt

@@ -8,7 +8,7 @@ attrs==23.2.0
     #   jsonlines
     #   jsonschema
     #   referencing
-certifi==2024.2.2
+certifi==2024.6.2
     # via requests
 cfgv==3.4.0
     # via pre-commit
@@ -18,17 +18,17 @@ charset-normalizer==3.3.2
     # via requests
 click==8.1.7
     # via typer
-coverage==7.5.1
+coverage==7.5.3
     # via pytest-cov
 distlib==0.3.8
     # via virtualenv
 et-xmlfile==1.1.0
     # via openpyxl
-filelock==3.14.0
+filelock==3.15.3
     # via virtualenv
 frictionless==5.17.0
     # via hdx-python-utilities
-hdx-python-utilities==3.6.8
+hdx-python-utilities==3.7.0
     # via hdx-python-country (pyproject.toml)
 humanize==4.9.0
     # via frictionless
@@ -36,7 +36,7 @@ identify==2.5.36
     # via pre-commit
 idna==3.7
     # via requests
-ijson==3.2.3
+ijson==3.3.0
     # via hdx-python-utilities
 iniconfig==2.0.0
     # via pytest
@@ -60,17 +60,17 @@ loguru==0.7.2
     # via hdx-python-utilities
 markdown-it-py==3.0.0
     # via rich
-marko==2.0.3
+marko==2.1.1
     # via frictionless
 markupsafe==2.1.5
     # via jinja2
 mdurl==0.1.2
     # via markdown-it-py
-nodeenv==1.8.0
+nodeenv==1.9.1
     # via pre-commit
-openpyxl==3.1.2
+openpyxl==3.1.4
     # via hdx-python-utilities
-packaging==24.0
+packaging==24.1
     # via pytest
 petl==1.7.15
     # via frictionless
@@ -84,21 +84,19 @@ ply==3.11
     #   libhxl
 pre-commit==3.7.1
     # via hdx-python-country (pyproject.toml)
-pydantic==2.7.1
+pydantic==2.7.4
     # via frictionless
-pydantic-core==2.18.2
+pydantic-core==2.18.4
     # via pydantic
 pygments==2.18.0
     # via rich
-pyphonetics==0.5.3
-    # via hdx-python-country (pyproject.toml)
-pytest==8.2.1
+pytest==8.2.2
     # via
     #   hdx-python-country (pyproject.toml)
     #   pytest-cov
 pytest-cov==5.0.0
     # via hdx-python-country (pyproject.toml)
-python-dateutil==2.8.2
+python-dateutil==2.9.0.post0
     # via
     #   frictionless
     #   hdx-python-utilities
@@ -112,13 +110,15 @@ pyyaml==6.0.1
     #   frictionless
     #   pre-commit
     #   tableschema-to-template
+rapidfuzz==3.9.3
+    # via hdx-python-country (pyproject.toml)
 ratelimit==2.2.1
     # via hdx-python-utilities
 referencing==0.35.1
     # via
     #   jsonschema
     #   jsonschema-specifications
-requests==2.32.2
+requests==2.32.3
     # via
     #   frictionless
     #   libhxl
@@ -137,8 +137,6 @@ ruamel-yaml==0.18.6
     # via hdx-python-utilities
 ruamel-yaml-clib==0.2.8
     # via ruamel-yaml
-setuptools==70.0.0
-    # via nodeenv
 shellingham==1.5.4
     # via typer
 simpleeval==0.9.13
@@ -149,7 +147,7 @@ six==1.16.0
     #   python-dateutil
 stringcase==1.2.0
     # via frictionless
-structlog==24.1.0
+structlog==24.2.0
     # via libhxl
 tableschema-to-template==0.0.13
     # via hdx-python-utilities
@@ -159,21 +157,19 @@ text-unidecode==1.3
     # via python-slugify
 typer==0.12.3
     # via frictionless
-typing-extensions==4.11.0
+typing-extensions==4.12.2
     # via
     #   frictionless
     #   pydantic
     #   pydantic-core
     #   typer
 unidecode==1.3.8
-    # via
-    #   libhxl
-    #   pyphonetics
-urllib3==2.2.1
+    # via libhxl
+urllib3==2.2.2
     # via
     #   libhxl
     #   requests
-validators==0.28.1
+validators==0.28.3
     # via frictionless
 virtualenv==20.26.2
     # via pre-commit

{hdx_python_country-3.7.1 → hdx_python_country-3.7.3}/src/hdx/location/_version.py

@@ -12,5 +12,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '3.7.1'
-__version_tuple__ = version_tuple = (3, 7, 1)
+__version__ = version = '3.7.3'
+__version_tuple__ = version_tuple = (3, 7, 3)

{hdx_python_country-3.7.1 → hdx_python_country-3.7.3}/src/hdx/location/adminlevel.py

@@ -5,7 +5,6 @@ from typing import Any, Dict, List, Optional, Tuple
 import hxl
 from hxl import InputOptions
 from hxl.input import HXLIOException
-from unidecode import unidecode
 
 from hdx.location.country import Country
 from hdx.location.names import clean_name
@@ -147,7 +146,7 @@ class AdminLevel:
         self.pcode_to_name[pcode] = adm_name
 
         name_to_pcode = self.name_to_pcode.get(countryiso3, {})
-        name_to_pcode[unidecode(adm_name).lower()] = pcode
+        name_to_pcode[clean_name(adm_name)] = pcode
         self.name_to_pcode[countryiso3] = name_to_pcode
         self.pcode_to_iso3[pcode] = countryiso3
         self.pcode_to_iso3[pcode] = countryiso3
@@ -157,7 +156,7 @@ class AdminLevel:
                 countryiso3, {}
             )
             name_to_pcode = name_parent_to_pcode.get(parent, {})
-            name_to_pcode[unidecode(adm_name).lower()] = pcode
+            name_to_pcode[clean_name(adm_name)] = pcode
             name_parent_to_pcode[parent] = name_to_pcode
             self.name_parent_to_pcode[countryiso3] = name_parent_to_pcode
             self.pcode_to_parent[pcode] = parent
@@ -306,7 +305,7 @@ class AdminLevel:
         """Get admin level for country
 
         Args:
-            countryiso3 (str): Iso3 country code
+            countryiso3 (str): ISO3 country code
 
         Returns:
             int: Admin level
@@ -320,7 +319,7 @@ class AdminLevel:
         """Get pcode length for country
 
         Args:
-            countryiso3 (str): Iso3 country code
+            countryiso3 (str): ISO3 country code
 
         Returns:
             Optional[int]: Country's pcode length or None
@@ -517,6 +516,41 @@ class AdminLevel:
             return pcode
         return None
 
+    def get_admin_name_replacements(
+        self, countryiso3: str, parent: Optional[str]
+    ) -> Dict[str, str]:
+        """Get relevant admin name replacements from admin name replacements
+        which is a dictionary of mappings from string to string replacement.
+        These can be global or they can be restricted by
+        country or parent (if the AdminLevel object has been set up with
+        parents). Keys take the form "STRING_TO_REPLACE",
+        "AFG|STRING_TO_REPLACE" or "AF01|STRING_TO_REPLACE".
+
+        Args:
+            countryiso3 (str): ISO3 country code
+            parent (Optional[str]): Parent admin code
+
+        Returns:
+            Dict[str, str]: Relevant admin name replacements
+        """
+        relevant_name_replacements = {}
+        for key, value in self.admin_name_replacements.items():
+            if "|" not in key:
+                if key not in relevant_name_replacements:
+                    relevant_name_replacements[key] = value
+                continue
+            prefix, name = key.split("|")
+            if parent:
+                if prefix == parent:
+                    if name not in relevant_name_replacements:
+                        relevant_name_replacements[name] = value
+                    continue
+            if prefix == countryiso3:
+                if name not in relevant_name_replacements:
+                    relevant_name_replacements[name] = value
+                continue
+        return relevant_name_replacements
+
     def fuzzy_pcode(
         self,
         countryiso3: str,
@@ -526,7 +560,7 @@ class AdminLevel:
         """Fuzzy match name to pcode
 
         Args:
-            countryiso3 (str): Iso3 country code
+            countryiso3 (str): ISO3 country code
             name (str): Name to match
             **kwargs:
             parent (Optional[str]): Parent admin code
@@ -543,8 +577,17 @@ class AdminLevel:
             if logname:
                 self.ignored.add((logname, countryiso3))
             return None
-        if self.use_parent and "parent" in kwargs:
-            parent = kwargs["parent"]
+        if self.use_parent:
+            parent = kwargs.get("parent")
+        else:
+            parent = None
+        if parent is None:
+            name_to_pcode = self.name_to_pcode.get(countryiso3)
+            if not name_to_pcode:
+                if logname:
+                    self.errors.add((logname, countryiso3))
+                return None
+        else:
             name_parent_to_pcode = self.name_parent_to_pcode.get(countryiso3)
             if not name_parent_to_pcode:
                 if logname:
@@ -555,15 +598,10 @@ class AdminLevel:
                 if logname:
                     self.errors.add((logname, countryiso3, parent))
                 return None
-        else:
-            name_to_pcode = self.name_to_pcode.get(countryiso3)
-            if not name_to_pcode:
-                if logname:
-                    self.errors.add((logname, countryiso3))
-                return None
         adm_name_lookup = clean_name(name)
         adm_name_lookup2 = multiple_replace(
-            adm_name_lookup, self.admin_name_replacements
+            adm_name_lookup,
+            self.get_admin_name_replacements(countryiso3, parent),
         )
         pcode = name_to_pcode.get(
             adm_name_lookup, name_to_pcode.get(adm_name_lookup2)
@@ -603,22 +641,25 @@ class AdminLevel:
                     break
         if not pcode:
             map_names = list(name_to_pcode.keys())
-            lower_mapnames = [x.lower() for x in map_names]
 
             def al_transform_1(name):
-                if name[:3] == "al ":
+                prefix = name[:3]
+                if prefix == "al ":
                     return f"ad {name[3:]}"
+                elif prefix == "ad ":
+                    return f"al {name[3:]}"
                 else:
                     return None
 
             def al_transform_2(name):
-                if name[:3] == "al ":
+                prefix = name[:3]
+                if prefix == "al " or prefix == "ad ":
                     return name[3:]
                 else:
                     return None
 
             matching_index = self.phonetics.match(
-                lower_mapnames,
+                map_names,
                 adm_name_lookup,
                 alternative_name=adm_name_lookup2,
                 transform_possible_names=[al_transform_1, al_transform_2],
@@ -643,19 +684,48 @@ class AdminLevel:
                 )
         return pcode
 
+    def get_name_mapped_pcode(
+        self, countryiso3: str, name: str, parent: Optional[str]
+    ) -> Optional[str]:
+        """Get pcode from admin name mappings which is a dictionary of mappings
+        from name to pcode. These can be global or they can be restricted by
+        country or parent (if the AdminLevel object has been set up with
+        parents). Keys take the form "MAPPING", "AFG|MAPPING" or
+        "AF01|MAPPING".
+
+        Args:
+            countryiso3 (str): ISO3 country code
+            name (str): Name to match
+            parent (Optional[str]): Parent admin code
+
+        Returns:
+            Optional[str]: P code match from admin name mappings or None if no match
+        """
+        if parent:
+            pcode = self.admin_name_mappings.get(f"{parent}|{name}")
+            if pcode is None:
+                pcode = self.admin_name_mappings.get(f"{countryiso3}|{name}")
+        else:
+            pcode = self.admin_name_mappings.get(f"{countryiso3}|{name}")
+        if pcode is None:
+            pcode = self.admin_name_mappings.get(name)
+        return pcode
+
     def get_pcode(
         self,
         countryiso3: str,
         name: str,
         fuzzy_match: bool = True,
+        fuzzy_length: int = 4,
         **kwargs: Any,
     ) -> Tuple[Optional[str], bool]:
         """Get pcode for a given name
 
         Args:
-            countryiso3 (str): Iso3 country code
+            countryiso3 (str): ISO3 country code
             name (str): Name to match
             fuzzy_match (bool): Whether to try fuzzy matching. Defaults to True.
+            fuzzy_length (int): Minimum length for fuzzy matching. Defaults to 4.
             **kwargs:
             parent (Optional[str]): Parent admin code
             logname (str): Log using this identifying name. Defaults to not logging.
@@ -663,9 +733,17 @@ class AdminLevel:
         Returns:
             Tuple[Optional[str], bool]: (Matched P code or None if no match, True if exact match or False if not)
         """
-        pcode = self.admin_name_mappings.get(name)
+        if self.use_parent:
+            parent = kwargs.get("parent")
+        else:
+            parent = None
+        pcode = self.get_name_mapped_pcode(countryiso3, name, parent)
         if pcode and self.pcode_to_iso3[pcode] == countryiso3:
-            return pcode, True
+            if parent:
+                if self.pcode_to_parent[pcode] == parent:
+                    return pcode, True
+            else:
+                return pcode, True
         if self.looks_like_pcode(name):
             pcode = name.upper()
             if pcode in self.pcodes:  # name is a p-code
@@ -677,8 +755,7 @@ class AdminLevel:
             )
             return pcode, True
         else:
-            if self.use_parent and "parent" in kwargs:
-                parent = kwargs["parent"]
+            if parent:
                 name_parent_to_pcode = self.name_parent_to_pcode.get(
                     countryiso3
                 )
@@ -694,7 +771,7 @@ class AdminLevel:
                     pcode = name_to_pcode.get(name.lower())
                     if pcode:
                         return pcode, True
-            if not fuzzy_match:
+            if not fuzzy_match or len(name) < fuzzy_length:
                 return None, True
             pcode = self.fuzzy_pcode(countryiso3, name, **kwargs)
             return pcode, False
@@ -756,3 +833,16 @@ class AdminLevel:
             logger.info(line)
             output.append(line)
         return output
+
+    def output_admin_name_replacements(self) -> List[str]:
+        """Output log of name replacements
+
+        Returns:
+            List[str]: List of name replacements
+        """
+        output = []
+        for name, replacement in self.admin_name_replacements.items():
+            line = f"{name}: {replacement}"
+            logger.info(line)
+            output.append(line)
+        return output

{hdx_python_country-3.7.1 → hdx_python_country-3.7.3}/src/hdx/location/names.py

@@ -3,7 +3,7 @@ import unicodedata
 
 from unidecode import unidecode
 
-non_ascii = "([^\x00-\x7f])+"
+non_ascii = r"([^\x00-\x7f])+"
 
 
 def clean_name(name: str) -> str:
@@ -25,5 +25,7 @@ def clean_name(name: str) -> str:
     )
     # Remove all non-ASCII characters
     clean_name = re.sub(non_ascii, " ", clean_name)
+    clean_name = clean_name.replace("'", "")
+    clean_name = re.sub(r"[\W_]", " ", clean_name)
     clean_name = unidecode(clean_name)
     return clean_name.strip().lower()

{hdx_python_country-3.7.1 → hdx_python_country-3.7.3}/src/hdx/location/phonetics.py

@@ -1,18 +1,18 @@
 from typing import Callable, Optional
 
-import pyphonetics
+from rapidfuzz import fuzz
 
 from hdx.utilities.typehint import ListTuple
 
 
-class Phonetics(pyphonetics.RefinedSoundex):
+class Phonetics:
     def match(
         self,
         possible_names: ListTuple,
         name: str,
         alternative_name: Optional[str] = None,
         transform_possible_names: ListTuple[Callable] = [],
-        threshold: int = 2,
+        threshold: float = 60,
     ) -> Optional[int]:
         """
         Match name to one of the given possible names. Returns None if no match
@@ -23,22 +23,22 @@ class Phonetics(pyphonetics.RefinedSoundex):
             name (str): Name to match
             alternative_name (str): Alternative name to match. Defaults to None.
             transform_possible_names (ListTuple[Callable]): Functions to transform possible names.
-            threshold: Match threshold. Defaults to 2.
+            threshold (float): Match threshold. Value is 0-100. Defaults to 60.
 
         Returns:
             Optional[int]: Index of matching name from possible names or None
         """
-        mindistance = None
+        max_similarity = 0
         matching_index = None
 
         transform_possible_names.insert(0, lambda x: x)
 
         def check_name(name, possible_name):
-            nonlocal mindistance, matching_index  # noqa: E999
+            nonlocal max_similarity, matching_index  # noqa: E999
 
-            distance = self.distance(name, possible_name)
-            if mindistance is None or distance < mindistance:
-                mindistance = distance
+            similarity = fuzz.token_sort_ratio(name, possible_name)
+            if similarity > max_similarity:
+                max_similarity = similarity
                 matching_index = i
 
         for i, possible_name in enumerate(possible_names):
@@ -51,6 +51,6 @@ class Phonetics(pyphonetics.RefinedSoundex):
                 check_name(name, transformed_possible_name)
                 if alternative_name:
                     check_name(alternative_name, transformed_possible_name)
-        if mindistance is None or mindistance > threshold:
+        if max_similarity < threshold:
             return None
         return matching_index

{hdx_python_country-3.7.1 → hdx_python_country-3.7.3}/tests/fixtures/adminlevel.yaml

@@ -434,6 +434,7 @@ admin_info:
   - {pcode: XYZ123456, name: Random, iso3: XYZ}
 
 countries_fuzzy_try:
+  - AFG
   - NER
   - NGA
   - UKR

hdx_python_country-3.7.3/tests/fixtures/adminlevelparent.yaml

@@ -0,0 +1,26 @@
+admin_info_with_parent:
+  - {pcode: AF0101, name: Kabul, iso3: AFG, parent: AF01}
+  - {pcode: AF0102, name: Paghman, iso3: AFG, parent: AF01}
+  - {pcode: AF0201, name: Kabul, iso3: AFG, parent: AF02}  # testing purposes
+  - {pcode: AF0301, name: Charikar, iso3: AFG, parent: AF03}
+  - {pcode: AF0401, name: Maydan Shahr, iso3: AFG, parent: AF04}
+  - {pcode: AF0501, name: Pul-e-Alam, iso3: AFG, parent: AF05}
+  - {pcode: AF0501, name: Pul-e-Alam, iso3: AFG, parent: AF05}
+  - {pcode: CD2013, name: Mbanza-Ngungu, iso3: COD, parent: CD20}
+  - {pcode: CD3102, name: Kenge, iso3: COD, parent: CD31}
+  - {pcode: MW305, name: Blantyre, iso3: MWI, parent: MW3}
+
+admin_name_mappings:
+  "MyMapping": "AF0301"
+  "AFG|MyMapping2": "AF0401"
+  "AF05|MyMapping3": "AF0501"
+
+admin_name_replacements:
+  "wx": "an"
+
+alt1_admin_name_replacements:
+  "COD|wx": "an"
+
+alt2_admin_name_replacements:
+  "CD20|wx": "an"
+  "CD31|wx": "en"

{hdx_python_country-3.7.1 → hdx_python_country-3.7.3}/tests/hdx/location/test_adminlevel.py

@@ -14,6 +14,10 @@ class TestAdminLevel:
     def config(self):
         return load_yaml(join("tests", "fixtures", "adminlevel.yaml"))
 
+    @pytest.fixture(scope="function")
+    def config_parent(self):
+        return load_yaml(join("tests", "fixtures", "adminlevelparent.yaml"))
+
     @pytest.fixture(scope="function")
     def url(self):
         return "https://raw.githubusercontent.com/OCHA-DAP/hdx-python-country/main/tests/fixtures/global_pcodes_adm_1_2.csv"
@@ -161,9 +165,27 @@ class TestAdminLevel:
         assert output[31] == "Juba Dhexe: Middle Juba (SO27)"
         assert output[61] == "CU Niamey: Niamey (NER008)"
 
+        output = adminone.output_admin_name_replacements()
+        assert output == [
+            " urban: ",
+            "sud: south",
+            "ouest: west",
+            "est: east",
+            "nord: north",
+            "': ",
+            "/:  ",
+            ".:  ",
+            " region: ",
+            " oblast: ",
+        ]
+
     def test_adminlevel_fuzzy(self, config):
         adminone = AdminLevel(config)
         adminone.setup_from_admin_info(config["admin_info"])
+        assert adminone.get_pcode("YEM", "Al_Dhale'a", logname="test") == (
+            "YE30",
+            False,
+        )
         assert adminone.get_pcode("YEM", "Al Dali", logname="test") == (
             "YE30",
             False,
@@ -174,20 +196,30 @@ class TestAdminLevel:
             "YE30",
             False,
         )
+        assert adminone.get_pcode("YEM", "Ad Jauf", logname="test") == (
+            "YE16",
+            False,
+        )
+        assert adminone.get_pcode("AFG", "Sar-E-Pul", logname="test") == (
+            "AF22",
+            False,
+        )
         assert adminone.get_pcode("SOM", "Bay", logname="test") == (
             "SO24",
             True,
         )
         output = adminone.output_matches()
         assert output == [
+            "test - YEM: Matching (fuzzy) Ad Jauf to Al Jawf on map",
             "test - YEM: Matching (fuzzy) Al Dali to Ad Dali on map",
             "test - YEM: Matching (fuzzy) Al Dhale'e / الضالع to Ad Dali on map",
+            "test - YEM: Matching (fuzzy) Al_Dhale'a to Ad Dali on map",
         ]
 
-    def test_adminlevel_parent(self, config):
-        admintwo = AdminLevel(config)
+    def test_adminlevel_parent(self, config_parent):
+        admintwo = AdminLevel(config_parent)
         admintwo.countries_fuzzy_try = None
-        admintwo.setup_from_admin_info(config["admin_info_with_parent"])
+        admintwo.setup_from_admin_info(config_parent["admin_info_with_parent"])
         assert admintwo.use_parent is True
         assert admintwo.pcode_to_parent["AF0101"] == "AF01"
         assert admintwo.get_pcode("AFG", "AF0101", logname="test") == (
@@ -223,6 +255,129 @@ class TestAdminLevel:
             "ABC", "Kabull", parent="AF02", logname="test"
         ) == (None, False)
 
+        output = admintwo.output_admin_name_mappings()
+        assert output == [
+            "MyMapping: Charikar (AF0301)",
+            "AFG|MyMapping2: Maydan Shahr (AF0401)",
+            "AF05|MyMapping3: Pul-e-Alam (AF0501)",
+        ]
+        assert admintwo.get_pcode("AFG", "MyMapping", logname="test") == (
+            "AF0301",
+            True,
+        )
+        assert admintwo.get_pcode(
+            "AFG", "MyMapping", parent="AF03", logname="test"
+        ) == ("AF0301", True)
+        assert admintwo.get_pcode(
+            "AFG", "MyMapping", parent="AF04", logname="test"
+        ) == (None, False)
+
+        assert admintwo.get_pcode("AFG", "MyMapping2", logname="test") == (
+            "AF0401",
+            True,
+        )
+        assert admintwo.get_pcode(
+            "AFG", "MyMapping2", parent="AF04", logname="test"
+        ) == ("AF0401", True)
+        assert admintwo.get_pcode(
+            "AFG", "MyMapping2", parent="AF05", logname="test"
+        ) == (None, False)
+
+        assert admintwo.get_pcode("AFG", "MyMapping3", logname="test") == (
+            None,
+            False,
+        )
+        assert admintwo.get_pcode(
+            "AFG", "MyMapping3", parent="AF05", logname="test"
+        ) == ("AF0501", True)
+        assert admintwo.get_pcode(
+            "AFG", "MyMapping3", parent="AF04", logname="test"
+        ) == (None, False)
+
+        output = admintwo.output_admin_name_replacements()
+        assert output == ["wx: an"]
+        assert admintwo.get_pcode(
+            "COD", "City of Mbwxza-Nanga", logname="test"
+        ) == ("CD2013", False)
+        assert admintwo.get_pcode(
+            "COD", "City of Mbwxza-Nanga", parent="CD20", logname="test"
+        ) == ("CD2013", False)
+        assert admintwo.get_pcode(
+            "COD", "City of Mbwxza-Nanga", parent="CD19", logname="test"
+        ) == (None, False)
+        assert admintwo.get_pcode(
+            "MWI", "City of Blwxtyre", logname="test"
+        ) == (
+            "MW305",
+            False,
+        )
+
+        admintwo.admin_name_replacements = config_parent[
+            "alt1_admin_name_replacements"
+        ]
+        output = admintwo.output_admin_name_replacements()
+        assert output == ["COD|wx: an"]
+        assert admintwo.get_pcode(
+            "COD", "City of Mbwxza-Nanga", logname="test"
+        ) == ("CD2013", False)
+        assert admintwo.get_pcode(
+            "COD", "City of Mbwxza-Nanga", parent="CD20", logname="test"
+        ) == ("CD2013", False)
+        assert admintwo.get_pcode(
+            "COD", "City of Mbwxza-Nanga", parent="CD19", logname="test"
+        ) == (None, False)
+        assert admintwo.get_pcode("COD", "Kenge City", logname="test") == (
+            "CD3102",
+            False,
+        )
+        assert admintwo.get_pcode("COD", "Kwxge City", logname="test") == (
+            None,
+            False,
+        )
+        assert admintwo.get_pcode(
+            "COD", "Kwxge City", parent="CD31", logname="test"
+        ) == (None, False)
+        assert admintwo.get_pcode(
+            "MWI", "City of Blwxtyre", logname="test"
+        ) == (
+            None,
+            False,
+        )
+        assert admintwo.get_pcode(
+            "MWI", "City of Blwxtyre", parent="MW3", logname="test"
+        ) == (None, False)
+
+        admintwo.admin_name_replacements = config_parent[
+            "alt2_admin_name_replacements"
+        ]
+        output = admintwo.output_admin_name_replacements()
+        assert output == ["CD20|wx: an", "CD31|wx: en"]
+        assert admintwo.get_pcode(
+            "COD", "City of Mbwxza-Nanga", logname="test"
+        ) == (None, False)
+        assert admintwo.get_pcode(
+            "COD", "City of Mbwxza-Nanga", parent="CD20", logname="test"
+        ) == ("CD2013", False)
+        assert admintwo.get_pcode(
+            "COD", "City of Mbwxza-Nanga", parent="CD19", logname="test"
+        ) == (None, False)
+        assert admintwo.get_pcode("COD", "Kwxge City", logname="test") == (
+            None,
+            False,
+        )
+        assert admintwo.get_pcode(
+            "COD", "Kwxge City", parent="CD31", logname="test"
+        ) == ("CD3102", False)
+        assert admintwo.get_pcode(
+            "MWI", "City of Blwxtyre", logname="test"
+        ) == (
+            None,
+            False,
+        )
+        assert admintwo.get_pcode(
+            "MWI", "City of Blwxtyre", parent="MW3", logname="test"
+        ) == (None, False)
+
     def test_adminlevel_with_url(self, config, url):
         adminone = AdminLevel(config)
         with pytest.raises(HXLIOException):
@@ -322,7 +477,7 @@ class TestAdminLevel:
         )
         assert adminone.get_pcode("YEM", "Ad Dali", logname="test") == (
             "YE30",
-            False,
+            True,
         )
         assert adminone.get_pcode("YEM", "Ad Dal", logname="test") == (
             "YE30",
@@ -365,7 +520,6 @@ class TestAdminLevel:
             "test - NGA: Matching (pcode length conversion) NG015 to Federal Capital Territory on map",
             "test - UKR: Matching (substring) Chernihiv Oblast to Chernihivska on map",
             "test - YEM: Matching (substring) Ad Dal to Ad Dali' on map",
-            "test - YEM: Matching (substring) Ad Dali to Ad Dali' on map",
             "test - YEM: Matching (pcode length conversion) YE30 to Ad Dali' on map",
         ]
         output = adminone.output_ignored()