DataDev Utilities

A lightweight, browser-based toolkit that bundles several everyday utilities for developers and data analysts into one terminal-inspired, fully responsive interface — no installation, no backend, everything runs client-side.

🔗 Live demo: skematografi.github.io/auto-utility-tool

About the Project

Features

The app is organized into tabs, each a self-contained tool:

• calc (Calculator) — Sum a list of numbers (one per line) with live totals, item count, and one-click copy. Accepts Indonesian/European number formats: a comma is treated as the decimal separator and dots as thousands separators when a comma is present (e.g. 100,120 → 100.120, 100.120,55 → 100120.55); plain dot-decimals like 120.55 are unchanged. Any line containing +, -, *, or / (other than a leading negative sign) is evaluated as a formula with operator precedence and added to the total (e.g. 1*5 + 5 → 10; 10-5 + 10/2 + 5 → 15).
• dupes (Duplicates) — Detect duplicate entries in a list and show how many times each appears, with three copy options: copy dupes (only the duplicated values), copy unique (every value de-duplicated, one each), and copy non-dupes (only values that appear exactly once).
• ascii (Non-ASCII) — Detect and highlight non-ASCII / hidden Unicode characters, list their code points, and copy a cleaned (ASCII-only) version of the text. Also accepts an Excel/CSV upload and scans it by row — reporting only the rows/cells that actually contain non-ASCII characters (with their location and code points), or a "file is clean" message when there are none.
• diff (Compare) — Compare two blocks of text line-by-line and highlight the differences between them.
• sql (SQL Generator) — Generate SQL from an uploaded Excel / CSV file:
  • Upload .xlsx, .xls, or .csv and preview the detected columns.
  • Delete — build DELETE ... WHERE col IN (...) with automatic per-column de-duplication, from one or more conditions.
  • Update — build UPDATE ... SET ... WHERE ..., grouping rows with identical SET values into a single IN (...) statement where possible.
  • Template — write any SQL with {ColumnName} / {1} placeholders and it is filled once per row — for complex cases the structured modes can't express (e.g. updating a detail table via a join).
  • Smart quoting: numbers stay unquoted; strings containing a single quote are wrapped in double quotes (and vice versa).
  • The result is previewed and downloaded directly as a .sql file.
• log (Log Viewer) — Upload or paste an application log and read it as a searchable table (time, level, user, IP, category, message):
  • Understands the common 2026-07-25 11:06:39 [ip][user][session][level][category] message layout, and keeps multi-line dumps (e.g. $_POST = [ … ]) attached to their entry.
  • Full-text search across every column with match highlighting, plus a level filter and per-level counts.
  • The message column is a show detail button — click it to open the full message in a centered, scrollable dialog (with copy, Esc to close), search matches highlighted inside.
  • Results are paged (50/100/250 per page) so large files stay responsive.
• split (Split File) — Split an uploaded Excel / CSV into multiple files, downloaded together as a .zip:
  • Set the max rows per file — the number of output files is calculated automatically.
  • Optionally pick a column so rows sharing the same value stay in the same file; leave it empty to split purely by row count.
  • Each output file keeps the header row and matches the uploaded format (.csv → CSV, Excel → .xlsx).
• dummy (Dummy File) — Generate a blank file of a given type and size, for testing uploads and size limits (max 100 MB):
  • Supported formats: xlsx, docx, pdf, csv, txt, json, xml, jpg, png, gif — or enter any custom extension.
  • Set the target size in KB or MB, and optionally a file name.
  • Each file stays valid for its format: padding is placed where the format allows ignorable data (a PNG text chunk, JPEG/GIF comment blocks, a PDF content stream, or a spreadsheet cell / document paragraph) instead of being blindly appended — so images, PDFs, and Office documents all still open.
  • A custom extension that matches a supported format (e.g. pdf, jpeg, docx) produces a real file of that format; anything else produces a blank text file, and the status line says so.
  • Sizes are byte-exact, except where a format has a larger minimum structure (e.g. an .xlsx cannot go below ~16 KB).
• in() (WHERE IN Generator) — Paste a list (one item per line) and get values ready to drop into a SQL WHERE col IN (...) clause:
  • Auto-detects the list type: if every line is numeric (integer or decimal, e.g. 500.0000) you get two outputs (unquoted numbers and quoted strings); otherwise only the quoted string output is shown.
  • Decimal values are tidied without changing their value (500.0000 → 500, 3.5000 → 3.5) so the unquoted output matches integer columns for faster queries.
  • Single quotes inside string values are escaped (' → '') so the output runs safely in SQL; double quotes need no escaping.
  • Optional query template: write any SQL with a {values} placeholder (e.g. ... WHERE productId IN ({values});) to get complete statements instead of a bare list.
  • Two copy buttons per output — copy (all values as one list/statement) and copy chunked (split into groups of at most 500 values per IN; with a template each chunk becomes its own statement, otherwise each is labelled with a -- chunk n/N header).
  • When the list exceeds 1000 items the chunked version is shown in the output box by default.
  • Output lines wrap at 200 characters, breaking to a new line after a comma.
• chars (Character Count) — Paste any text and get live counts plus database storage estimates:
  • Total characters (counted per Unicode code point, so emoji count as 1), lines, words, sentences, paragraphs, and spaces.
  • size_utf8 — actual UTF-8 byte size, matching storage in MySQL utf8mb4 (VARCHAR / TEXT) and PostgreSQL.
  • size_utf16 — 2 bytes per character, matching SQL Server NVARCHAR.
  • Sizes are shown human-readable (B / KB / MB / GB) with the exact byte count below; per-row overhead is not included.
• restore (JSON → SQL) — Paste JSON (a single object or an array of records) and generate INSERT ... SELECT statements to restore the data:
  • Head — all root fields that are not arrays go into one head table (you provide its name).
  • Detail — map one or more array keys (e.g. Details) to their own tables; each array element becomes a select ... union all row.
  • Dates are normalized (...T00:00:00... → YYYY-MM-DD, otherwise YYYY-MM-DD HH:MM:SS); null stays null; single quotes are escaped (' → '').
  • Two output styles: with column names it emits INSERT INTO t (cols) VALUES (...), (...) (columns camelCase, any ...ID stays capital, e.g. branchID); without, it emits INSERT INTO t SELECT ... UNION ALL SELECT .... Optionally null-out the identity ID column for a clean restore.
  • Output can be copied or downloaded as a .sql file.
• sort (Sort) — Paste a list or sentence (e.g. 52,91,4,5,11) and instantly get it sorted ascending and descending, both copyable. The separator is configurable (default ,) and newlines are always accepted. Sorting is natural / numeric-aware, so 1,10,2,5 becomes 1,2,5,10 (not lexicographic).
• merge (Merge SQL) — Upload multiple .sql files and combine them into a single .sql (downloaded, merged in the order selected):
  • Optional clear SQL comments — strips --, #, and /* */ comments while preserving string literals.
  • A boundary comment is inserted before each file (default: the filename); use {filename} as a placeholder to customize it.
• json (JSON Formatter) — Paste any JSON and it formats live, no button required:
  • code — pretty-printed, 2-space-indented JSON.
  • tree — an expandable/collapsible tree (native <details>, no extra JS needed to toggle), with string/number/boolean/null values color-coded.
  • Shows a valid / invalid status badge and the total key count, and surfaces the parser's exact error message on invalid input.
  • copy grabs the pretty-printed JSON regardless of which view is active.
• csv (CSV Viewer) — Paste delimited text or upload a .csv file and browse it as a spreadsheet-style table, processed live:
  • The first row is always treated as the header; RFC4180-style quoted fields (embedded delimiter, quotes, newlines) always work.
  • Delimiter is configurable — comma, semicolon, pipe, tab, or any single custom character.
  • Filter — a search box matches across every column, plus a per-column filter box in the header row (Excel-style, all active filters combine); results are paginated (50/100/250 per page) so large files stay responsive.
  • Bordered cells, row numbers, and a sticky header keep it readable like a spreadsheet.
  • copy filtered copies the currently filtered rows (with header) back out as delimited text.

Sticky note

Alongside the tabs there is a sticky note panel — a scratchpad for parking values while you move between tools (a total from calc into in(), a column spotted in log into sql), since the clipboard only holds one item at a time.

It is hidden by default and opens from the floating button in the bottom-right corner. The content is saved automatically in your browser (localStorage), so it survives tab switches and reloads, and never leaves your device. On wide screens the panel slides in beside the card without covering it; on narrower screens it overlays the page.

Command palette

Press Ctrl+K (or Cmd+K on macOS), or click the ⌘ icon in the title bar, to open a searchable command palette: type to fuzzy-filter and jump straight to any of the tabs above, or run a quick action (toggle theme, toggle the background animation, toggle the sticky note) without touching the mouse. Navigate with the arrow keys, confirm with Enter, dismiss with Esc or a click outside.

Focus mode

For screen-sharing or demos, click the maximize icon in the footer (or run toggle focus mode from the command palette) to hide the hero text, the tab grid, and the floating sticky-note / command-palette buttons — leaving just the title bar and the active tool. Switching tabs still works via Ctrl+K. Off by default; the toggle itself always stays visible so you can turn it back off.

Run Locally

Clone this repository:

git clone https://github.com/Skematografi/auto-utility-tool.git

Then simply open index.html in your browser — no installation or setup required.