XML Validation against Rosreestr Schema

The XML Validation against Rosreestr Schema dialog is designed for automated analysis of XML documents for compliance with XSD schemas using language models from Yandex AI Studio. The tool helps identify structural errors, data type mismatches, constraint violations, and other rule breaches defined in the target schema.

Controls and Parameters

1. XML document field — displays the path to the selected file.

• The “…” button opens a standard file selection dialog to load an XML document (.xml extension).

2. XSD schema field — displays the root schema file.

• The “…” button opens a dialog to select an XSD file (.xsd extension). All dependent schemas (referenced via xs:include, xs:import, xs:redefine) are collected automatically if they are stored locally.

3. API key field — mandatory. The entered access key for Yandex AI Studio is masked. When settings are saved to the Panorama INI file, the key and folder ID are encrypted.

4. Folder ID field — mandatory. Specifies the folder ID of your cloud project in Yandex Cloud.

5. Model dropdown list — contains the list of available text models in Yandex AI Studio. The list updates automatically after a valid API key and folder ID are entered. By default, the model qwen3-235b-a22b-fp8/latest is selected. When the model is changed, the Maximum characters field is filled with the recommended value corresponding to the selected model's context window.

6. Maximum characters in XML and XSD field — sets a limit on the total volume of text (XML plus all collected XSD schemas) sent to the model. Integer values from 1,000 to 10,000,000 are allowed. If the limit is exceeded, validation is not performed and an appropriate error message is displayed.

7. Action buttons:

• Execute — starts the validation process. The button is locked during execution.

• Save report… — saves the contents of the results area to a text file. The button is active only after successful validation.

• Exit — closes the dialog and saves the current settings to the Panorama INI file.

• Help — opens the help page in a browser, respecting the interface language.

8. Results output area — displays the response from Yandex AI Studio, containing the verdict, a list of errors with XPath locations, actual values, violated rules, warnings, and a final summary. The field is read‑only; its content can be copied.

Step-by-Step Instructions

1. Click the “…” button next to the XML document field and select the target .xml file using the standard file dialog.

2. Similarly, click the “…” button next to the XSD schema field and select the root .xsd file. All local dependent schemas (referenced via xs:include, xs:import, xs:redefine) will be automatically located and loaded.

3. Copy your access key from your Yandex Cloud account and paste it into the API Key field.

4. Enter the Folder ID of your cloud project in Yandex Cloud.

5. If the model list is empty, it will refresh automatically within a few seconds after entering a valid API key and Folder ID. If needed, select a different model from the dropdown list.

6. The Maximum characters in XML and XSD field is automatically populated with the recommended value for the selected model. Manual adjustment within the range of 1,000 to 10,000,000 is allowed.

7. Click Execute. The process will start: collecting all dependent schemas, building the request, and sending it to Yandex AI Studio. Wait for the processing to complete.

8. The verdict and detailed list of errors (if any) will appear in the results output area.

9. After successful validation, the Save report… button becomes active. Click it, specify a folder and file name (.txt is suggested by default).

10. Close the dialog by clicking Exit or the standard window close button. All settings (API key, folder ID, selected model, limits) are saved to the Panorama INI file and will be restored on the next launch.

Features and Limitations

• Credential encryption

The API key and folder ID are stored encrypted using Windows DPAPI. Decryption is possible only on the same computer and under the same user account.

• Automatic dependency collection

The script recursively processes all xs:include, xs:import, and xs:redefine references, loading only local files. External URLs (e.g., http://...) are not fetched.

• Preliminary XML syntax check

If the source XML document is not well‑formed, XSD validation is not performed, and a syntax error message is immediately shown in the results area.

• Recommendatory nature of the verdict

The verdict produced by the language model is not an official conclusion from a certified XSD validator. It is strongly recommended to additionally validate the document using specialized tools (e.g., a built‑in validator).

• Response language

Regardless of the Panorama interface language, the model always responds in Russian.

• Settings persistence

All settings except file paths are remembered when the dialog is closed and restored upon the next opening.

Response Structure

The model's response contains the following sections:

• VERDICT – one of: "COMPLIANT", "NON‑COMPLIANT", or "INSUFFICIENT DATA".

• ERRORS – a numbered list indicating:

— XPath to the problematic element/attribute;

— actual value (if applicable);

— violated rule or constraint;

— remediation recommendation.

• WARNINGS – additional non‑critical notes.

• SUMMARY – a brief conclusion based on the validation results.