cu-cli 0.1.0b1__tar.gz → 0.1.0b2__tar.gz

{cu_cli-0.1.0b1 → cu_cli-0.1.0b2}/PKG-INFO

@@ -1,7 +1,7 @@
 Metadata-Version: 2.4
 Name: cu-cli
-Version: 0.1.0b1
-Summary: Author, validate, and run Azure Content Understanding custom analyzers from the terminal — advanced document layout, industry-leading OCR, and grounded field extraction with confidence, all backed by Azure Content Understanding.
+Version: 0.1.0b2
+Summary: Run Azure Content Understanding prebuilt analyzers, and/or author, validate, and run custom analyzers from the terminal — advanced document layout, industry-leading OCR, and grounded field extraction with confidence, all backed by Azure Content Understanding.
 Author: Microsoft Corporation
 License: MIT
 Project-URL: Homepage, https://github.com/Azure/content-understanding-toolkit/tree/main/cu-cli
@@ -21,7 +21,7 @@ Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
 Classifier: Topic :: Text Processing
 Requires-Python: >=3.10
 Description-Content-Type: text/markdown
-Requires-Dist: cu-cli-core<0.2.0,>=0.1.0b1
+Requires-Dist: cu-cli-core<0.2.0,>=0.1.0b2
 Requires-Dist: azure-ai-contentunderstanding>=1.2.0b3
 Requires-Dist: azure-identity>=1.19
 Requires-Dist: azure-mgmt-cognitiveservices>=13.5
@@ -44,7 +44,8 @@ Requires-Dist: vcrpy>=6.0; extra == "dev"
 Foundry Tools. Use it to provision the required Microsoft Foundry resource,
 optionally deploy selected supported large language models (LLMs) and embeddings
 models, configure Content Understanding defaults, analyze local files, manage
-analyzers, and manage local configuration.
+analyzers, and manage local configuration. Analyze inputs can also be supplied
+as HTTPS or Azure Blob SAS URLs.
 
 > [!IMPORTANT]
 > CU CLI is in preview. Commands and package contracts may change before general
@@ -72,8 +73,8 @@ The Content Understanding documentation uses these terms:
   structured fields to extract.
 
 CU CLI lets you configure a Microsoft Foundry resource, select an analyzer,
-submit local files, and save analyzer results without calling the REST API
-directly.
+submit local files or HTTPS URLs, and save analyzer results without calling the
+REST API directly.
 
 Further reading:
 
@@ -134,6 +135,47 @@ checks the API version, endpoint, authentication, service connectivity, and
 Content Understanding defaults. It exits nonzero when a required check fails,
 so it can serve as a readiness gate.
 
+## Use `cu` and `az cu` interchangeably
+
+The standalone CU CLI and the Azure CLI extension are two frontends over the
+same `cu-cli-core` operations. Install either frontend, or install both and move
+between them for the same analyzer, analysis, defaults, profile, diagnostics,
+and infrastructure-generation workflows.
+
+Both frontends read and write the same `[cu]` profile settings in the active
+Azure CLI configuration file (`~/.azure/config` by default, or the file under
+`AZURE_CONFIG_DIR`). An endpoint, API version, default analyzer, active profile,
+or model-deployment mapping saved with one frontend is immediately available to
+the other. For example:
+
+```bash
+# Save the endpoint with the standalone frontend.
+cu profile set endpoint https://<resource-name>.services.ai.azure.com/
+
+# Use the same default profile with the Azure CLI extension.
+az cu analyzer list --output table
+
+# Change the default analyzer with the Azure CLI extension.
+az cu profile set --key default_analyzer --value prebuilt-layout
+
+# Use that setting with the standalone frontend.
+cu analyze ./document.pdf
+```
+
+The command names and capabilities overlap, but frontend conventions differ:
+
+| Standalone CU CLI | Azure CLI extension |
+| --- | --- |
+| Starts commands with `cu` (`cu-cli` on macOS). | Starts commands with `az cu`. |
+| Supports standalone positional shortcuts and Rich/JSON output options. | Uses explicit options plus global Azure CLI `--output`, `--query`, and `--subscription`. |
+| Uses the profile's `auth_mode` and can use a saved API key. | Always uses the active `az login` identity; shared API keys and `auth_mode` do not override Azure CLI host authentication. |
+
+Run `cu <command> --help` or `az cu <command> --help` when translating a command
+between frontends. Profile values are shared; authentication sessions are not,
+so sign in with `az login` before using `az cu`. For Azure CLI-specific
+installation and command examples, see the
+[Azure Content Understanding extension README](https://github.com/Azure/content-understanding-toolkit/blob/main/cu-cli/packages/azure-cli-extension/README.md).
+
 ## Supported Content Understanding API versions
 
 Content Understanding API version. Known versions: 2025-11-01 (GA) and
@@ -182,6 +224,23 @@ model or embeddings model. `-a` is the short form of `--analyzer`:
 cu analyze ./document.pdf -a prebuilt-layout
 ```
 
+Analyze a remote file with `--url` without downloading it through the CLI.
+Quote URLs that contain SAS query parameters so the shell preserves `&` characters:
+
+```bash
+cu analyze --url "https://storage.example.net/container/video.mp4?sv=<version>&sp=r&sig=<signature>" \
+  --analyzer prebuilt-videoSearch
+```
+
+Repeat `--url` for multiple URLs and specify `--output-dir` for their results.
+`--url` cannot be combined with positional inputs, `--file`, or `--source`.
+Positional URLs remain available as a standalone shortcut.
+
+Remote inputs require HTTPS. Use a short-lived, read-only SAS and avoid placing
+production SAS URLs in shared shell history. CU CLI removes URL query strings
+from its status output and reports, while sending the original URL to Content
+Understanding unchanged.
+
 Markdown generated by the Content Understanding SDK's `to_llm_input()` helper is
 the default output format. The helper formats field extraction results as
 Markdown with YAML frontmatter so they can be used as generative AI model input.
@@ -271,7 +330,7 @@ Further reading:
 
 | Command | Purpose |
 | --- | --- |
-| `cu analyze` | Analyze local files and return analyzer results. |
+| `cu analyze` | Analyze local files or HTTPS URLs and return analyzer results. |
 | `cu analyzer` | List, show, create, copy, delete, and test analyzers; create and validate local analyzer schemas. |
 | `cu defaults` | Read or configure Content Understanding defaults that map models to deployments. |
 | `cu profile` | Manage local CU CLI endpoint, authentication, API, and model settings. |

{cu_cli-0.1.0b1 → cu_cli-0.1.0b2}/README.md

@@ -4,7 +4,8 @@
 Foundry Tools. Use it to provision the required Microsoft Foundry resource,
 optionally deploy selected supported large language models (LLMs) and embeddings
 models, configure Content Understanding defaults, analyze local files, manage
-analyzers, and manage local configuration.
+analyzers, and manage local configuration. Analyze inputs can also be supplied
+as HTTPS or Azure Blob SAS URLs.
 
 > [!IMPORTANT]
 > CU CLI is in preview. Commands and package contracts may change before general
@@ -32,8 +33,8 @@ The Content Understanding documentation uses these terms:
   structured fields to extract.
 
 CU CLI lets you configure a Microsoft Foundry resource, select an analyzer,
-submit local files, and save analyzer results without calling the REST API
-directly.
+submit local files or HTTPS URLs, and save analyzer results without calling the
+REST API directly.
 
 Further reading:
 
@@ -94,6 +95,47 @@ checks the API version, endpoint, authentication, service connectivity, and
 Content Understanding defaults. It exits nonzero when a required check fails,
 so it can serve as a readiness gate.
 
+## Use `cu` and `az cu` interchangeably
+
+The standalone CU CLI and the Azure CLI extension are two frontends over the
+same `cu-cli-core` operations. Install either frontend, or install both and move
+between them for the same analyzer, analysis, defaults, profile, diagnostics,
+and infrastructure-generation workflows.
+
+Both frontends read and write the same `[cu]` profile settings in the active
+Azure CLI configuration file (`~/.azure/config` by default, or the file under
+`AZURE_CONFIG_DIR`). An endpoint, API version, default analyzer, active profile,
+or model-deployment mapping saved with one frontend is immediately available to
+the other. For example:
+
+```bash
+# Save the endpoint with the standalone frontend.
+cu profile set endpoint https://<resource-name>.services.ai.azure.com/
+
+# Use the same default profile with the Azure CLI extension.
+az cu analyzer list --output table
+
+# Change the default analyzer with the Azure CLI extension.
+az cu profile set --key default_analyzer --value prebuilt-layout
+
+# Use that setting with the standalone frontend.
+cu analyze ./document.pdf
+```
+
+The command names and capabilities overlap, but frontend conventions differ:
+
+| Standalone CU CLI | Azure CLI extension |
+| --- | --- |
+| Starts commands with `cu` (`cu-cli` on macOS). | Starts commands with `az cu`. |
+| Supports standalone positional shortcuts and Rich/JSON output options. | Uses explicit options plus global Azure CLI `--output`, `--query`, and `--subscription`. |
+| Uses the profile's `auth_mode` and can use a saved API key. | Always uses the active `az login` identity; shared API keys and `auth_mode` do not override Azure CLI host authentication. |
+
+Run `cu <command> --help` or `az cu <command> --help` when translating a command
+between frontends. Profile values are shared; authentication sessions are not,
+so sign in with `az login` before using `az cu`. For Azure CLI-specific
+installation and command examples, see the
+[Azure Content Understanding extension README](https://github.com/Azure/content-understanding-toolkit/blob/main/cu-cli/packages/azure-cli-extension/README.md).
+
 ## Supported Content Understanding API versions
 
 Content Understanding API version. Known versions: 2025-11-01 (GA) and
@@ -142,6 +184,23 @@ model or embeddings model. `-a` is the short form of `--analyzer`:
 cu analyze ./document.pdf -a prebuilt-layout
 ```
 
+Analyze a remote file with `--url` without downloading it through the CLI.
+Quote URLs that contain SAS query parameters so the shell preserves `&` characters:
+
+```bash
+cu analyze --url "https://storage.example.net/container/video.mp4?sv=<version>&sp=r&sig=<signature>" \
+  --analyzer prebuilt-videoSearch
+```
+
+Repeat `--url` for multiple URLs and specify `--output-dir` for their results.
+`--url` cannot be combined with positional inputs, `--file`, or `--source`.
+Positional URLs remain available as a standalone shortcut.
+
+Remote inputs require HTTPS. Use a short-lived, read-only SAS and avoid placing
+production SAS URLs in shared shell history. CU CLI removes URL query strings
+from its status output and reports, while sending the original URL to Content
+Understanding unchanged.
+
 Markdown generated by the Content Understanding SDK's `to_llm_input()` helper is
 the default output format. The helper formats field extraction results as
 Markdown with YAML frontmatter so they can be used as generative AI model input.
@@ -231,7 +290,7 @@ Further reading:
 
 | Command | Purpose |
 | --- | --- |
-| `cu analyze` | Analyze local files and return analyzer results. |
+| `cu analyze` | Analyze local files or HTTPS URLs and return analyzer results. |
 | `cu analyzer` | List, show, create, copy, delete, and test analyzers; create and validate local analyzer schemas. |
 | `cu defaults` | Read or configure Content Understanding defaults that map models to deployments. |
 | `cu profile` | Manage local CU CLI endpoint, authentication, API, and model settings. |

{cu_cli-0.1.0b1 → cu_cli-0.1.0b2}/pyproject.toml

@@ -4,8 +4,8 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "cu-cli"
-version = "0.1.0b1"
-description = "Author, validate, and run Azure Content Understanding custom analyzers from the terminal — advanced document layout, industry-leading OCR, and grounded field extraction with confidence, all backed by Azure Content Understanding."
+version = "0.1.0b2"
+description = "Run Azure Content Understanding prebuilt analyzers, and/or author, validate, and run custom analyzers from the terminal — advanced document layout, industry-leading OCR, and grounded field extraction with confidence, all backed by Azure Content Understanding."
 readme = "README.md"
 requires-python = ">=3.10"
 license = { text = "MIT" }
@@ -27,7 +27,7 @@ classifiers = [
     "Topic :: Text Processing",
 ]
 dependencies = [
-    "cu-cli-core>=0.1.0b1,<0.2.0",
+    "cu-cli-core>=0.1.0b2,<0.2.0",
     "azure-ai-contentunderstanding>=1.2.0b3",
     "azure-identity>=1.19",
     # Management-plane clients used by `cu analyzer copy` to resolve
@@ -68,15 +68,6 @@ where = ["src"]
 
 [tool.setuptools.package-data]
 "cu_cli" = ["py.typed"]
-"cu_cli.resources" = [
-    "azd_template/azure.yaml",
-    "azd_template/README.md",
-    "azd_template/infra/*.bicep",
-    "azd_template/infra/*.json",
-    "azd_template/infra/modules/*.bicep",
-    "azd_template/hooks/*.ps1",
-    "azd_template/hooks/*.sh",
-]
 
 [tool.pytest.ini_options]
 testpaths = ["tests"]

{cu_cli-0.1.0b1 → cu_cli-0.1.0b2}/src/cu_cli/apiversion.py

@@ -18,11 +18,14 @@ Resolution precedence (highest -> lowest):
 from __future__ import annotations
 
 import os
-import re
 from dataclasses import dataclass
 from typing import Optional
 
-from cu_cli_core.service_options import DEFAULT_API_VERSION
+from cu_cli_core.service_options import (
+    DEFAULT_API_VERSION,
+    SUPPORTED_API_VERSIONS,
+    is_supported_api_version,
+)
 
 from .errors import CuCliError
 
@@ -43,8 +46,6 @@ API_VERSIONS = (
     ),
 )
 
-SUPPORTED_API_VERSIONS = tuple(item.value for item in API_VERSIONS)
-_PREVIEW_VERSION_PATTERN = re.compile(r"^\d{4}-\d{2}-\d{2}-preview$")
 _KNOWN_VERSION_LABELS = tuple(
     f"{item.value} ({item.label})" for item in API_VERSIONS
 )
@@ -68,11 +69,15 @@ def _supported_list() -> str:
 
 def is_preview_version(version: Optional[str]) -> bool:
     """Return whether *version* has the forward-compatible preview shape."""
-    return bool(version and _PREVIEW_VERSION_PATTERN.fullmatch(version))
+    return bool(
+        version
+        and version not in SUPPORTED_API_VERSIONS
+        and is_supported_api_version(version)
+    )
 
 
 def is_supported(version: Optional[str]) -> bool:
-    return version in SUPPORTED_API_VERSIONS or is_preview_version(version)
+    return is_supported_api_version(version)
 
 
 def ensure_supported(version: Optional[str]) -> str:

{cu_cli-0.1.0b1 → cu_cli-0.1.0b2}/src/cu_cli/cli.py

@@ -25,6 +25,7 @@ from .commands.infra import infra_group
 from .commands.upgrade import cmd_upgrade
 from .commands._help import common_commands
 from .commands._infra_models import cmd_infra_models
+from .commands._postprovision import cmd_postprovision_v1
 
 
 def _force_utf8_io() -> None:
@@ -132,6 +133,7 @@ main.add_command(analyzer_group)
 main.add_command(defaults_group)
 main.add_command(cmd_upgrade)
 main.add_command(cmd_infra_models)
+main.add_command(cmd_postprovision_v1)
 
 
 if __name__ == "__main__":

{cu_cli-0.1.0b1 → cu_cli-0.1.0b2}/src/cu_cli/commands/_command_spec.py

@@ -11,12 +11,18 @@ from typing import Any
 
 import rich_click as click
 
-from cu_cli_core.command_spec import ArgumentSpec, ArgumentValueType, CommandSpec
+from cu_cli_core.command_spec import (
+    ArgumentSpec,
+    ArgumentValueType,
+    CommandSpec,
+    SurfaceClassification,
+)
 
 
 _SIMPLE_CLICK_TYPES: dict[ArgumentValueType, Any] = {
     ArgumentValueType.STRING: str,
     ArgumentValueType.BOOLEAN: bool,
+    ArgumentValueType.TRISTATE_BOOLEAN: click.BOOL,
 }
 
 
@@ -74,12 +80,17 @@ def with_command_arguments(spec: CommandSpec):
     """Attach the command-specific arguments declared by ``spec``."""
 
     def decorate(fn: Callable[..., Any]) -> Callable[..., Any]:
+        arguments = tuple(
+            argument
+            for argument in spec.arguments
+            if argument.classification is not SurfaceClassification.AZURE_ONLY
+        )
         fields_with_alternates = {
             argument.field
-            for argument in spec.arguments
-            if sum(item.field == argument.field for item in spec.arguments) > 1
+            for argument in arguments
+            if sum(item.field == argument.field for item in arguments) > 1
         }
-        for argument in reversed(spec.arguments):
+        for argument in reversed(arguments):
             option = (
                 _click_argument(argument)
                 if argument.positional

{cu_cli-0.1.0b1 → cu_cli-0.1.0b2}/src/cu_cli/commands/_infra_models.py

@@ -6,9 +6,9 @@
 from __future__ import annotations
 
 import sys
-from pathlib import Path
 
 import rich_click as click
+from cu_cli_core.command_spec import INFRA_MODELS
 
 from ..client import build_client, resolve
 from ..profile import Profile
@@ -26,6 +26,7 @@ from ..errors import CuCliError, friendly_errors
 from ..output import console
 from ._help import common_commands
 from ._options import print_runtime_context, with_auth_options
+from ._command_spec import with_command_arguments
 
 NO_MODEL_ANALYZERS = (
     "prebuilt-digitalParse",
@@ -119,13 +120,7 @@ def _client(endpoint, api_key, api_version, entra, profile_name, show_runtime_co
         ),
     ),
 )
-@click.option("--resource-group", required=True)
-@click.option("--account", "account_name", required=True)
-@click.option("--subscription", "subscription_id", required=True)
-@click.option("--selection", required=True,
-              help="prompt, recommended, none, or comma-separated model selectors.")
-@click.option("--out", "out_path", required=True, type=click.Path(path_type=Path))
-@click.option("--deploy/--no-deploy", default=True)
+@with_command_arguments(INFRA_MODELS)
 @with_auth_options
 @friendly_errors
 def cmd_infra_models(