Release v1.3.0

CHANGELOG.md

@@ -7,6 +7,15 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [1.3.0] - 2026-04-22
+
+### Added
+
+- `check <file>` workbook and sheet quality scans with a stable JSON summary, stats, and findings contract.
+- Fixed v1.3.0 rule registry covering `blank_headers`, `duplicate_headers`, `blank_rows`, `blank_columns`, `null_ratio`, `duplicate_values`, `type_drift`, and `formula_presence`.
+- TUI findings panel support for browsing quality-check results without leaving the terminal.
+- Regression coverage for rule positives and negatives, workbook and sheet targeting, threshold filtering, finding order, and exit-code behavior.
+
 ## [1.2.0] - 2026-04-21
 
 ### Added
@@ -21,12 +30,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - JSON Lines output for stream-friendly record processing.
 - Read-only TUI query preview with `:preview` and `:pv`.
 - Regression coverage for filtering, output shapes, invalid query errors, no-match results, help text, and query preview behavior.
-- CI formatting enforcement with `cargo fmt --check`.
 
 ### Changed
 
 - Enriched read metadata now reports applied filters, selected columns, returned row count, truncation, and output shape.
-- Package version updated to `1.2.0`.
 
 ## [1.1.0] - 2026-04-21
 
@@ -35,7 +42,6 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - `inspect columns <file> --sheet <name>` to profile column headers, generated safe names, duplicate headers, inferred types, non-null ratios, formula ratios, and sample values.
 - `inspect tables <file> --sheet <name>` to detect table-like regions with ranges, header rows, dimensions, and confidence scores.
 - Regression coverage for structure inspection cases including duplicate headers, blank headers, preamble sections, late headers, multi-table sheets, mixed-type columns, formula columns, and non-ASCII column names.
-- Release branch CI triggers for `release/**` branches.
 
 ## [1.0.0] - 2026-04-20
 
@@ -195,7 +201,8 @@ This is the initial release of excel-cli, a lightweight terminal-based Excel vie
 - Copy, cut, and paste functionality with `y`, `d`, and `p` keys
 - Support for pipe operator when exporting to JSON
 
-[Unreleased]: https://github.com/fuhan666/excel-cli/compare/v1.2.0...HEAD
+[Unreleased]: https://github.com/fuhan666/excel-cli/compare/v1.3.0...HEAD
+[1.3.0]: https://github.com/fuhan666/excel-cli/compare/v1.2.0...v1.3.0
 [1.2.0]: https://github.com/fuhan666/excel-cli/compare/v1.1.0...v1.2.0
 [1.1.0]: https://github.com/fuhan666/excel-cli/compare/v1.0.0...v1.1.0
 [1.0.0]: https://github.com/fuhan666/excel-cli/releases/tag/v1.0.0

Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "excel-cli"
-version = "1.2.0"
+version = "1.3.0"
 edition = "2021"
 description = "Excel CLI for AI, scripting, and terminal users. Headless JSON API for automation, plus a Vim-like TUI for interactive browsing and editing."
 license = "MIT"

README.md

@@ -9,9 +9,11 @@ An Excel CLI for AI, scripting, and terminal users. Inspect and read headlessly
 - Edit cell contents directly in the terminal
 - Export data to JSON format
 - Select, filter, paginate, and stream read results for automation
+- Run workbook and sheet quality checks with stable JSON findings
 - Delete rows and columns
 - Search functionality with highlighting
 - Read-only query preview in the TUI
+- Review quality-check findings inside the TUI
 - Command mode for advanced operations
 
 ## Installation & Uninstallation
@@ -95,6 +97,15 @@ excel-cli read records path/to/your/file.xlsx --sheet Orders \
   --limit 50 \
   --output-shape jsonl
 
+# Check workbook data quality with the full rule set
+excel-cli check path/to/your/file.xlsx
+
+# Check one sheet with selected rules
+excel-cli check path/to/your/file.xlsx --sheet Orders --rules duplicate_values,type_drift
+
+# Return only warning and error findings
+excel-cli check path/to/your/file.xlsx --severity-threshold warning
+
 # Open interactive TUI browser
 excel-cli ui path/to/your/file.xlsx
 ```
@@ -154,6 +165,28 @@ excel-cli read records report.xlsx --sheet Orders --output-shape jsonl
 
 Invalid selected columns, unknown filter columns, unsupported operators, malformed filters, invalid numeric comparisons, and invalid regular expressions return structured `invalid_query` errors with exit code `6`.
 
+### Quality Checks
+
+`check` runs the fixed quality-rule registry against a whole workbook or a single sheet and emits the same stable JSON envelope as the other headless commands. By default it scans every sheet, returns `info`, `warning`, and `error` findings, exits `1` when the filtered result set is non-empty, and exits `0` when no findings remain after filtering.
+
+Supported rules:
+- `blank_headers`: flag blank header cells inside the detected header row
+- `duplicate_headers`: flag duplicate normalized header names
+- `blank_rows`: flag fully blank rows inside the used range
+- `blank_columns`: flag fully blank columns inside the used range
+- `null_ratio`: flag columns with blank-value ratios above the built-in thresholds
+- `duplicate_values`: flag repeated values in the candidate identifier column
+- `type_drift`: flag mixed dominant and drift data types in a column
+- `formula_presence`: report sheets that still contain formulas in the checked data region
+
+Use `--sheet <name>` to limit the scan to one sheet, `--rules <comma,separated,ids>` to run a subset in registry order, and `--severity-threshold <info|warning|error>` to filter returned findings. `data.summary` counts only the returned findings, while `data.stats.finding_count_before_threshold` preserves the total before threshold filtering and `data.stats.rules_run` records the normalized rule order.
+
+```bash
+excel-cli check report.xlsx
+excel-cli check report.xlsx --sheet 客户 --rules blank_headers,duplicate_headers
+excel-cli check report.xlsx --rules null_ratio,duplicate_values,type_drift --severity-threshold warning
+```
+
 ### Exit Codes
 
 | Code | Meaning |