v0.0.2

fiboa · Feb 16, 2024 · 3a2abc4 · 3a2abc4
1 parent 9e6f39f
commit 3a2abc4
Show file tree

Hide file tree

Showing 11 changed files with 484 additions and 386 deletions.
diff --git a/.editorconfig b/.editorconfig
@@ -0,0 +1,26 @@
+# EditorConfig is awesome: http://EditorConfig.org
+
+# top-most EditorConfig file
+root = true
+
+# Unix-style newlines with a newline ending every file
+[*]
+end_of_line = lf
+insert_final_newline = true
+trim_trailing_whitespace = true
+charset = utf-8
+
+# 4 space indentation
+[*.py]
+indent_style = space
+indent_size = 4
+
+# 2 space indentation
+[*.{json,yaml}]
+indent_style = space
+indent_size = 2
+
+[*.md]
+trim_trailing_whitespace = false
+indent_style = space
+indent_size = 2
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -27,9 +27,16 @@ and this project adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.
 
 - ...
 
+## [v0.0.2] - 2024-02-16
+
+- Basic validation for collection
+- Minimal validation for data
+- Fixed creation of GeoParquet files
+
 ## [v0.0.1] - 2024-02-16
 
 - First release
 
-[Unreleased]: <https://github.com/radiantearth/stac-spec/compare/v0.0.1...main>
+[Unreleased]: <https://github.com/radiantearth/stac-spec/compare/v0.0.2...main>
+[v0.0.2]: <https://github.com/radiantearth/stac-spec/compare/v0.0.1...v0.0.2>
 [v0.0.1]: <https://github.com/radiantearth/stac-spec/tree/v0.0.1>
diff --git a/fiboa_cli/__init__.py b/fiboa_cli/__init__.py
@@ -1,4 +1,5 @@
 import click
+import sys
 
 from .validate import validate as validate_
 from .create import create as create_
@@ -8,58 +9,98 @@
 @click.group()
 @click.version_option(version=__version__)
 def cli():
-  """
-  A simple CLI app.
-  """
-  pass
+    """
+    A simple CLI app.
+    """
+    pass
 
 @click.command()
 @click.argument('files', nargs=-1, callback=valid_files_for_cli)
-@click.option('--schema', '-s', type=click.Path(exists=True), help='fiboa Schema to validate against. Can be a local file or a URL. If not provided, uses the fiboa version to load the schema for the released version.')
-@click.option('--fiboa-version', '-f', type=click.STRING, help='The fiboa version to validate against. Default is the version given in the collection.', default=None)
-@click.option('--collection', '-c', type=click.Path(exists=True), help='Points to the STAC collection that defines the fiboa version and extensions.', default=None)
-@click.option('--verbose', '-v', type=click.BOOL, help='Run with verbose logging.', default=False)
-def validate(files, schema, fiboa_version, collection, verbose):
-  """
-  Validates a fiboa GeoParquet file.
-  """
-  log(f"fiboa CLI {__version__} - Validator\n", "success")
-  config = {
-    "schema": schema,
-    "fiboa_version": fiboa_version,
-    "collection": collection,
-    "verbose": verbose
-  }
-  for file in files:
-    log(f"Validating {file}", "info")
-    try:
-      result = validate_(file, config)
-      if result:
-        log(f"  => VALID\n", "success")
-      else:
-        log(f"  => INVALID\n", "error")
-    except Exception as e:
-      log(f"  => UNKNOWN: {e}\n", "error")
+@click.option(
+    '--schema', '-s',
+    type=click.Path(exists=True),
+    help='fiboa Schema to validate against. Can be a local file or a URL. If not provided, uses the fiboa version to load the schema for the released version.'
+)
+@click.option(
+    '--fiboa-version', '-f',
+    type=click.STRING,
+    help='The fiboa version to validate against. Default is the version given in the collection.',
+    default=None
+)
+@click.option(
+    '--collection', '-c',
+    type=click.Path(exists=True),
+    help='Points to the STAC collection that defines the fiboa version and extensions.',
+    default=None
+)
+@click.option(
+    '--data', '-d',
+    is_flag=True,
+    type=click.BOOL,
+    help='Validate the data in the file. Enabling this might be slow. Default is False.',
+    default=False
+)
+def validate(files, schema, fiboa_version, collection, data):
+    """
+    Validates a fiboa GeoParquet file.
+    """
+    log(f"fiboa CLI {__version__} - Validator\n", "success")
+    config = {
+        "schema": schema,
+        "fiboa_version": fiboa_version,
+        "collection": collection,
+        "data": data
+    }
+    for file in files:
+        log(f"Validating {file}", "info")
+        try:
+            result = validate_(file, config)
+            if result:
+                log("  => VALID\n", "success")
+            else:
+                log("  => INVALID\n", "error")
+                sys.exit(1)
+        except Exception as e:
+            log(f"  => UNKNOWN: {e}\n", "error")
+            sys.exit(2)
 
 @click.command()
 @click.argument('files', nargs=-1, callback=valid_files_for_cli)
-@click.option('--out', '-o', type=click.Path(exists=False), help='File to write the file to. If not provided, prints the file to the STDOUT.', required=True)
-@click.option('--collection', '-c', callback=valid_file_for_cli, help='Points to the STAC collection that defines the fiboa version and extensions.', required=True)
-@click.option('--schema', '-s', type=click.Path(exists=True), help='fiboa Schema to work against. If not provided, uses the fiboa version from the collection to load the schema for the released version.')
+@click.option(
+    '--out', '-o',
+    type=click.Path(exists=False),
+    help='File to write the file to. If not provided, prints the file to the STDOUT.',
+    required=True
+)
+@click.option(
+    '--collection', '-c',
+    callback=valid_file_for_cli,
+    help='Points to the STAC collection that defines the fiboa version and extensions.',
+    required=True
+)
+@click.option(
+    '--schema', '-s',
+    type=click.Path(exists=True),
+    help='fiboa Schema to work against. If not provided, uses the fiboa version from the collection to load the schema for the released version.'
+)
 def create(files, out, collection, schema):
-  """
-  Create a fiboa file from GeoJSON file(s).
-  """
-  config = {
-    "files": files,
-    "out": out,
-    "schema": schema,
-    "collection": collection
-  }
-  create_(config)
+    """
+    Create a fiboa file from GeoJSON file(s).
+    """
+    config = {
+        "files": files,
+        "out": out,
+        "schema": schema,
+        "collection": collection
+    }
+    try:
+        create_(config)
+    except Exception as e:
+        log(e, "error")
+        sys.exit(1)
 
 cli.add_command(validate)
 cli.add_command(create)
 
 if __name__ == '__main__':
-  cli()
+    cli()
diff --git a/fiboa_cli/const.py b/fiboa_cli/const.py
@@ -3,79 +3,79 @@
 import pandas as pd
 
 GP_TYPE_MAP = {
-  "boolean": "bool",
-  "int8": "int8",
-  "uint8": "uint8",
-  "int16": "int16",
-  "uint16": "uint16",
-  "int32": "int32",
-  "uint32": "uint32",
-  "int64": "int64",
-  "uint64": "uint64",
-  "float": "float32",
-  "double": "float64",
-  "binary": "bytearray", # todo: check
-  "string": "str",
-  "array": None, # todo: object?
-  "object": "object",
-  "enum": "category", # todo: check
-  "date": "datetime64[D]",
-  "date-time": lambda x: pd.to_datetime(x),
-  "geometry": None, # not a column, don't convert geometry
-  "bounding-box": None # todo
+    "boolean": "bool",
+    "int8": "int8",
+    "uint8": "uint8",
+    "int16": "int16",
+    "uint16": "uint16",
+    "int32": "int32",
+    "uint32": "uint32",
+    "int64": "int64",
+    "uint64": "uint64",
+    "float": "float32",
+    "double": "float64",
+    "binary": "bytearray", # todo: check
+    "string": "str",
+    "array": None, # todo: object?
+    "object": "object",
+    "enum": "category", # todo: check
+    "date": "datetime64[D]",
+    "date-time": lambda x: pd.to_datetime(x),
+    "geometry": None, # not a column, don't convert geometry
+    "bounding-box": None # todo
 }
 
 PA_TYPE_MAP = {
-  "boolean": pa.bool_(),
-  "int8": pa.int8(),
-  "uint8": pa.uint8(),
-  "int16": pa.int16(),
-  "uint16": pa.uint16(),
-  "int32": pa.int32(),
-  "uint32": pa.uint32(),
-  "int64": pa.int64(),
-  "uint64": pa.uint64(),
-  "float": pa.float32(),
-  "double": pa.float64(),
-  "binary": pa.binary(),
-  "string": pa.string(),
-  "array": lambda type: pa.list_(type),
-  "object": None, # todo: lambda type: pa.map_(pa.string(), type)
-  "enum": None, # todo: ENUM (BYTE_ARRAY)
-  "date": pa.date32(),
-  "date-time": pa.timestamp("ms", tz="UTC"),
-  "geometry": pa.binary(),
-  "bounding-box": None # todo
+    "boolean": pa.bool_(),
+    "int8": pa.int8(),
+    "uint8": pa.uint8(),
+    "int16": pa.int16(),
+    "uint16": pa.uint16(),
+    "int32": pa.int32(),
+    "uint32": pa.uint32(),
+    "int64": pa.int64(),
+    "uint64": pa.uint64(),
+    "float": pa.float32(),
+    "double": pa.float64(),
+    "binary": pa.binary(),
+    "string": pa.string(),
+    "array": lambda type: pa.list_(type),
+    "object": None, # todo: lambda type: pa.map_(pa.string(), type)
+    "enum": None, # todo: ENUM (BYTE_ARRAY)
+    "date": pa.date32(),
+    "date-time": pa.timestamp("ms", tz="UTC"),
+    "geometry": pa.binary(),
+    "bounding-box": None # todo
 }
 
 PA_TYPE_CHECK = {
-  "boolean": pat.is_boolean,
-  "int8": pat.is_int8,
-  "uint8": pat.is_uint8,
-  "int16": pat.is_int16,
-  "uint16": pat.is_uint16,
-  "int32": pat.is_int32,
-  "uint32": pat.is_uint32,
-  "int64": pat.is_int64,
-  "uint64": pat.is_uint64,
-  "float": pat.is_float32,
-  "double": pat.is_float64,
-  "binary": pat.is_binary,
-  "string": pat.is_string,
-  "array": pat.is_list,
-  "object": pat.is_map,
-  "enum": None, # ENUM (BYTE_ARRAY)
-  "date": pat.is_date32,
-  "date-time": pat.is_timestamp,
-  "geometry": pat.is_binary, # todo: check more?
-  "bounding-box": None # todo
+    "boolean": pat.is_boolean,
+    "int8": pat.is_int8,
+    "uint8": pat.is_uint8,
+    "int16": pat.is_int16,
+    "uint16": pat.is_uint16,
+    "int32": pat.is_int32,
+    "uint32": pat.is_uint32,
+    "int64": pat.is_int64,
+    "uint64": pat.is_uint64,
+    "float": pat.is_float32,
+    "double": pat.is_float64,
+    "binary": pat.is_binary,
+    "string": pat.is_string,
+    "array": pat.is_list,
+    "object": pat.is_map,
+    "enum": None, # ENUM (BYTE_ARRAY)
+    "date": pat.is_date32,
+    "date-time": pat.is_timestamp,
+    "geometry": pat.is_binary, # todo: check more?
+    "bounding-box": None # todo
 }
 
 LOG_STATUS_COLOR = {
-  "info": "white",
-  "warning": "yellow",
-  "error": "red",
-  "success": "green"
+    "info": "white",
+    "warning": "yellow",
+    "error": "red",
+    "success": "green"
 }
 
-SUPPORTED_PROTOCOLS = ["http", "https", "s3", "gs"]
+SUPPORTED_PROTOCOLS = ["http", "https", "s3", "gs"]