Merge pull request #94 from stac-utils/in_memory

Allow for linting in-memory dictionaries

CHANGELOG.md

@@ -5,6 +5,10 @@ All notable changes to this project will be documented in this file.
 The format is (loosely) based on [Keep a Changelog](http://keepachangelog.com/) and this project adheres to [Semantic Versioning](http://semver.org/).
 
 ## Unreleased
+## Added
+- Ability to lint dictionaries https://github.com/stac-utils/stac-check/pull/94
+## Fixed
+- Fixed the check_catalog_file_name() method to only work on static catalogs https://github.com/stac-utils/stac-check/pull/94
 
 ## [v1.3.1] - 2022-10-05
 ## Changed

README.md

@@ -12,7 +12,7 @@ or for local development
 `$ pip install -e .` 
 
 ---
-### Usage
+### CLI Usage
 ```
 Usage: stac-check [OPTIONS] FILE
 
@@ -34,7 +34,18 @@ $ make build
 $ make shell
 ```
 ---
-### Examples
+### Lint Dictionary
+
+```
+from stac_linter.lint import Linter
+
+linter = Linter(dict)
+
+for k,v in linter.create_best_practices_dict().items():
+    print(k,":",v)
+```
+---
+### CLI Examples
 
 ``` stac-check https://raw.githubusercontent.com/stac-utils/pystac/main/tests/data-files/examples/0.9.0/collection-spec/examples/landsat-collection.json --recursive ```
 ```
@@ -208,4 +219,4 @@ Validation error message:
     'id' is a required property of the root of the STAC object
 
 This object has 5 links
-</pre>
+</pre>

stac_check/lint.py

@@ -6,15 +6,15 @@
 import os
 from dataclasses import dataclass
 import requests
-from typing import Optional
+from typing import Optional, Union
 from dotenv import load_dotenv
 import pkg_resources
 
 load_dotenv()
 
 @dataclass
 class Linter:
-    item: str
+    item: Union[str, dict] # url, file name, or dictionary
     config_file: Optional[str] = None
     assets: bool = False
     links: bool = False
@@ -38,7 +38,7 @@ def __post_init__(self):
         self.invalid_link_request = self.check_links_assets(10, "links", "request") if self.links else None
         self.schema = self.message["schema"] if "schema" in self.message else []
         self.object_id = self.data["id"] if "id" in self.data else ""
-        self.file_name = os.path.basename(self.item).split('.')[0]
+        self.file_name = self.get_asset_name(self.item)
         self.best_practices_msg = self.create_best_practices_msg()
 
     @staticmethod
@@ -57,24 +57,41 @@ def parse_config(config_file):
 
         return default_config
 
+    def get_asset_name(self, file):
+        if isinstance(file, str):
+            return os.path.basename(file).split('.')[0]
+        else:
+            return file["id"]
+
     def load_data(self, file):
-        if is_valid_url(file):
-            resp = requests.get(file)
-            data = resp.json()
+        if isinstance(file, str):
+            if is_valid_url(file):
+                resp = requests.get(file)
+                data = resp.json()
+            else:
+                with open(file) as json_file:
+                    data = json.load(json_file)
+            return data
         else:
-            with open(file) as json_file:
-                data = json.load(json_file)
-        return data
+            return file
 
     def validate_file(self, file):
-        stac = StacValidate(file, links=self.links, assets=self.assets)
-        stac.run()
+        if isinstance(file, str):
+            stac = StacValidate(file, links=self.links, assets=self.assets)
+            stac.run()
+        else:
+            stac = StacValidate()
+            stac.validate_dict(file)
         return stac.message[0]
 
     def recursive_validation(self, file):
         if self.recursive:
-            stac = StacValidate(file, recursive=True, max_depth=self.max_depth)
-            stac.run()
+            if isinstance(file, str):
+                stac = StacValidate(file, recursive=True, max_depth=self.max_depth)
+                stac.run()
+            else:
+                stac = StacValidate(recursive=True, max_depth=self.max_depth)
+                stac.validate_dict(file)
             return stac.message
 
     def set_update_message(self):
@@ -179,11 +196,13 @@ def check_item_id_file_name(self):
         else:
             return True
 
-    def check_catalog_id_file_name(self):
-        if self.asset_type == "CATALOG" and self.file_name != 'catalog':
-            return False 
-        elif self.asset_type == "COLLECTION" and self.file_name != 'collection':
-            return False
+    def check_catalog_file_name(self):
+        if isinstance(self.item, str) and ".json" in self.item:
+            if self.asset_type == "CATALOG" and 'catalog.json' not in self.item:
+                return False 
+            elif self.asset_type == "COLLECTION" and 'collection.json' not in self.item:
+                return False
+            return True
         else:
             return True
 
@@ -211,7 +230,7 @@ def create_best_practices_dict(self):
             best_practices_dict["check_item_id"] = [msg_1]
 
         # best practices - collection and catalog file names should be collection.json and catalog.json 
-        if not self.check_catalog_id_file_name() and config["catalog_id_file_name"] == True: 
+        if self.check_catalog_file_name() == False and config["catalog_id_file_name"] == True: 
             msg_1 = f"Object should be called '{self.asset_type.lower()}.json' not '{self.file_name}.json'"
             best_practices_dict["check_catalog_id"] = [msg_1]
 

test_dict.py

@@ -0,0 +1,141 @@
+# from stac_check.lint import Linter
+
+# file = {
+#     "stac_version": "1.0.0",
+#     "stac_extensions": [],
+#     "type": "Feature",
+#     "id": "20201211_223832_CS2",
+#     "bbox": [
+#         172.91173669923782,
+#         1.3438851951615003,
+#         172.95469614953714,
+#         1.3690476620161975
+#     ],
+#     "geometry": {
+#         "type": "Polygon",
+#         "coordinates": [
+#         [
+#             [
+#             172.91173669923782,
+#             1.3438851951615003
+#             ],
+#             [
+#             172.95469614953714,
+#             1.3438851951615003
+#             ],
+#             [
+#             172.95469614953714,
+#             1.3690476620161975
+#             ],
+#             [
+#             172.91173669923782,
+#             1.3690476620161975
+#             ],
+#             [
+#             172.91173669923782,
+#             1.3438851951615003
+#             ]
+#         ]
+#         ]
+#     },
+#     "properties": {
+#         "title": "Core Item",
+#         "description": "A sample STAC Item that includes examples of all common metadata",
+#         "datetime": None,
+#         "start_datetime": "2020-12-11T22:38:32.125Z",
+#         "end_datetime": "2020-12-11T22:38:32.327Z",
+#         "created": "2020-12-12T01:48:13.725Z",
+#         "updated": "2020-12-12T01:48:13.725Z",
+#         "platform": "cool_sat1",
+#         "instruments": [
+#         "cool_sensor_v1"
+#         ],
+#         "constellation": "ion",
+#         "mission": "collection 5624",
+#         "gsd": 0.512
+#     },
+#     "collection": "simple-collection",
+#     "links": [
+#         {
+#         "rel": "collection",
+#         "href": "./collection.json",
+#         "type": "application/json",
+#         "title": "Simple Example Collection"
+#         },
+#         {
+#         "rel": "root",
+#         "href": "./collection.json",
+#         "type": "application/json",
+#         "title": "Simple Example Collection"
+#         },
+#         {
+#         "rel": "parent",
+#         "href": "./collection.json",
+#         "type": "application/json",
+#         "title": "Simple Example Collection"
+#         },
+#         {
+#         "rel": "alternate",
+#         "type": "text/html",
+#         "href": "http://remotedata.io/catalog/20201211_223832_CS2/index.html",
+#         "title": "HTML version of this STAC Item"
+#         }
+#     ],
+#     "assets": {
+#         "analytic": {
+#         "href": "https://storage.googleapis.com/open-cogs/stac-examples/20201211_223832_CS2_analytic.tif",
+#         "type": "image/tiff; application=geotiff; profile=cloud-optimized",
+#         "title": "4-Band Analytic",
+#         "roles": [
+#             "data"
+#         ]
+#         },
+#         "thumbnail": {
+#         "href": "https://storage.googleapis.com/open-cogs/stac-examples/20201211_223832_CS2.jpg",
+#         "title": "Thumbnail",
+#         "type": "image/png",
+#         "roles": [
+#             "thumbnail"
+#         ]
+#         },
+#         "visual": {
+#         "href": "https://storage.googleapis.com/open-cogs/stac-examples/20201211_223832_CS2.tif",
+#         "type": "image/tiff; application=geotiff; profile=cloud-optimized",
+#         "title": "3-Band Visual",
+#         "roles": [
+#             "visual"
+#         ]
+#         },
+#         "udm": {
+#         "href": "https://storage.googleapis.com/open-cogs/stac-examples/20201211_223832_CS2_analytic_udm.tif",
+#         "title": "Unusable Data Mask",
+#         "type": "image/tiff; application=geotiff;"
+#         },
+#         "json-metadata": {
+#         "href": "http://remotedata.io/catalog/20201211_223832_CS2/extended-metadata.json",
+#         "title": "Extended Metadata",
+#         "type": "application/json",
+#         "roles": [
+#             "metadata"
+#         ]
+#         },
+#         "ephemeris": {
+#         "href": "http://cool-sat.com/catalog/20201211_223832_CS2/20201211_223832_CS2.EPH",
+#         "title": "Satellite Ephemeris Metadata"
+#         }
+#     }
+# }
+# linter = Linter(file, assets=True)
+
+# print(linter.valid_stac)
+
+# print(linter.error_type)
+
+# print(linter.error_msg)
+
+# for k,v in linter.create_best_practices_dict().items():
+#     print(k,":",v)
+
+# print(linter.schema)
+
+# print(linter.file_name)