excel-html-converter

Turn Excel pivot tables, plain tables, and charts into browsable HTML reports — no Excel and no plugins needed to view them, just a browser. A small self-hosted Python web app does the work: upload an .xlsx, get a shareable report page.

📺 Video walkthrough

Screenshots

| Report list | Table (sort & filter) | |---|---| | | |

| Pivot grid | Line chart | |---|---| | | |

| Bar chart | Pie chart | |---|---| | | |

Quick start

npm install excel-html-converter
python node_modules/excel-html-converter/server/server.py

Open http://localhost:8000 in your browser.

What it does

• Reads classic (worksheet-based) pivot tables, plain worksheet tables, and charts (line / bar / pie) directly from .xlsx files.
• Respects what Excel already shows: hidden filter values, autofilter-hidden rows, and date groupings (Year / Quarter / Month) are read from the file, not re-derived.
• Renders each as its own standalone page — a PivotGrid-based grid for pivot tables (with expand/collapse), a sortable/filterable table for plain data, and an SVG chart for line/bar/pie — all read-only snapshots of what was already built in Excel.
• A tiny web UI lists all generated reports: create, open, update, delete. Report generation runs in a separate OS process, so a large file never blocks the server or takes the whole app down if it runs out of memory.
• Update re-runs the conversion for an existing report from a newly uploaded .xlsx — the report's id, name, and position in the list stay the same, only its content is replaced.

Configuration

This project has two independent config files — nothing else. Changing one does not affect the other; if you want both the site and the reports in the same language, set LANGUAGE in both.

server/.env (copy from server/.env.example) — server-side settings, restart the server after changing:

• PORT, HOST — where the server listens.
• LANGUAGE — ru or en. Controls server error messages and everything baked into a generated report page: footers, the "back to reports" nav link, the table's sort/filter UI (search, "Contains"/"Starts with", column headers), and the pivot grid widget itself.
• REPORTS_DIR — where uploaded files and generated reports are stored. Defaults to a reports/ folder next to wherever you run the server from — deliberately not inside the package itself, since node_modules gets wiped and recreated by npm install at any time, and that would take your reports with it.
• TABLE_DEFAULT_ROW_LIMIT, PIVOT_DEFAULT_ROW_LIMIT — covered in their own section below.

client/js/config.js — client-side settings for the report-list website itself (index.html / new.html). Edited directly in the file, no .env involved on this side, and no restart needed — just reload the page:

• LANGUAGE — ru or en. Controls only the site's own UI (the list page and the create-report form). Independent of the server's LANGUAGE above.
• POLL_INTERVAL_MS (default 3000) — how often, in milliseconds, the report list re-checks /api/reports while at least one report is still "processing", so its status flips to "ready" on its own, without a manual page reload.

Keeping your config outside node_modules

npm install puts both files inside node_modules, which gets wiped on every reinstall. Keep your config elsewhere with two optional flags:

python node_modules/excel-html-converter/server/server.py \
  --server-config /path/to/your/server.env \
  --client-config /path/to/your/config.js

Both flags are optional and independent of each other — pass either one, both, or neither (falls back to the bundled files). --server-config takes a .env-style file; --client-config takes a config.js-style file (same const CONFIG = {...} format as the bundled one). Run with --help to see them listed.

Project layout

server/
  server.py        - REST API + static file serving
  converters/       - the actual xlsx -> html parsers/generators
  reports/          - generated reports (created automatically)
client/
  index.html, new.html, js/, css/  - the small report-list web UI
  table-tools/, chart-tools/       - vanilla-JS rendering libraries used by generated reports
  node_modules/pivotgrid-js/       - pivot grid rendering engine (npm dependency)

Excel file requirements

• One kind of content per sheet. A sheet is either a plain table, or hosts pivot table(s), or hosts chart(s) — not a mix. Multiple pivot tables or charts can share one sheet (each becomes its own page in the report); a plain table cannot share a sheet with anything else.
• Only classic (worksheet-based) pivot tables are supported. Pivot tables built on the Data Model / Power Pivot are skipped (with a message in the server log) — rebuild them as a regular pivot on a cell range if you need them in the report.
• Supported chart types: line, bar (column), pie. Other chart types are skipped.
• Excel's own hard ceiling is 1,048,576 rows per sheet — not something this tool adds, just a fact about .xlsx itself.

Limits & configuration

Two different kinds of limit, both configurable in server/.env (copy from server/.env.example), server restart required after changing:

• TABLE_DEFAULT_ROW_LIMIT (default 5000) — a hard architectural ceiling, not just a nicety. Sorting/filtering a plain table happens directly on the DOM in the browser, with no virtualization; raising this a lot will make the browser itself noticeably slow on sort/filter, regardless of server power.
• PIVOT_DEFAULT_ROW_LIMIT (default 500000) — a soft safety default, not a hard limit. It exists because decoding raw pivot cache rows takes real server time and memory. The pivot grid itself renders fine at any size — feel free to raise this if your server has the time/memory to spare.

If a file has more rows than the limit in effect, the generated report footer says so explicitly ("shown: first N, source may have more"); if everything fit, no such note appears.

License

Free for personal, educational, or non-commercial use — see LICENSE. For commercial use, see LICENSE.commercial or contact [email protected].