ExcelReader

High-performance Excel reading and writing for .NET 10. Reads .xlsx, .xlsb, .xls, and .csv; writes .xlsx, .xlsb, .xls, and .csv.

ExcelReader is built for streaming spreadsheet workloads where low allocations matter. It reads worksheet rows as lightweight ref struct values, resolves shared strings, recognizes date styles, and handles sparse cells. Workbooks open from a path, a Stream, or a caller-owned ReadOnlyMemory<byte> — the in-memory path decompresses parts directly, without a ZipArchive or an intermediate stream.

Install

dotnet add package ExcelReader.NET

using ExcelReader.Core.Reader;

using IExcelRowReader reader = Excel.Open("book.xlsx");
foreach (Row row in reader)
{
    foreach (RowCell cell in row.Cells)
    {
        Console.Write(cell.Value.GetString());
        Console.Write('\t');
    }
    Console.WriteLine();
}

Guide

• Reading rows — opening a workbook, enumerating rows, auto-detecting the format, async reads, prefetch decompression, and the ADO.NET IDataReader bridge.
• Typed parsing — binding rows to your own types with attributes, the compile-time generator (Native AOT / trimming), the fluent API, required columns, custom converters, and the zero-copy ref struct path.
• Writing workbooks — XLSX, XLSB (BIFF12) and XLS (BIFF8) writers, cell styles, typed records, and prefetch compression.
• CSV — reading, dialect sniffing, parallel parsing, and writing delimited text.
• Encrypted workbooks — opening password-protected packages and encrypting written ones.
• Benchmarks — throughput and allocation against Sep, Sylvan, OfficeIMO and SpreadCheetah. Live results.

Command line

dotnet tool install --global ExcelReader.NET.Cli

excelreader sheets  book.xlsb                        # 0<TAB>Sheet1
excelreader schema  book.xlsb --sample-size 500      # 0<TAB>Id<TAB>Int64
excelreader convert book.xlsb --output data.csv      # convert to another format, by extension
excelreader convert book.xlsb --output data.xlsx     # .xlsx, .xlsb, .xls and .csv all work
excelreader convert book.xlsb --format xlsx | head -c 4   # to stdout, --format picks it instead

Flags: --sheet|-s <name|index>, --header-row N (0 = no header), --sample-size N, --output|-o <file>, --format|-f <xlsx|xlsb|xls|csv> (defaults to --output's extension, or csv for stdout), --delimiter|-d <char> (csv output), --input-delimiter <char> (csv input). Run excelreader <command> --help for the full list.

--sheet matches a sheet name first and falls back to a 0-based index, so a sheet literally named 2 is reachable by name. A .csv input has its dialect sniffed unless --input-delimiter names one, so a ;-separated export reads as columns rather than as one field; --delimiter stays on the output side, so convert data.csv -o out.csv -d ';' still means "read as sniffed, write semicolons". sheets and schema write plain tab-separated text. convert reports progress on stderr while it runs, unless stderr is redirected. Exit codes are 0 ok and 1 failure; results go to stdout and errors to stderr, so convert is safe to pipe.

Notes

• Reads .xlsx, .xlsb (BIFF12), .xls (BIFF8), and .csv; writes .xlsx, .xlsb, .xls, and .csv.
• Excel.Open/OpenAsync auto-detect XLSX/XLSB/XLS from the file signature. CSV has none, so pass ExcelFileFormat.Csv to open delimited text through the same entry point; ExcelReaderOptions.Csv carries the dialect, and CsvReaderOptions.SniffDialect infers it from the source.
• Reads one sheet at a time (XLSX/XLSB/XLS); use MoveToSheet(index) or TryMoveToSheet(name) to switch sheets. CSV has no sheets.
• SheetVisibility/SheetVisibilityAt(index) report whether a sheet is hidden (Visible, Hidden, VeryHidden). Hidden sheets still enumerate their rows — filter on the value if you want them skipped. The writers take the same value via AddSheet(name, visibility), and reject a workbook whose sheets would all be hidden.
• Missing cells in sparse rows are exposed as empty cells.
• String conversion allocates only when you call GetString().
• The XLSX scanner accepts the SpreadsheetML shapes commonly emitted by non-Excel producers, including single-quoted attributes, comments in sheetData, and CDATA text runs.
• Readers bound untrusted input by default: 512 MB total decompressed ZIP data, 32 MB per cell/row value buffer, and 128 MB for shared strings. Pass ExcelReaderOptions to the Excel.From*/Excel.Open* factories to tune these limits; set a limit to 0 to opt out and restore unlimited behavior for that limit. CsvReader has its own CsvReaderOptions.MaxCellBytes (default 32 MB) for the same purpose.
• The XLSX writer emits a compact workbook with strings, numbers, booleans, dates, and blank cells; shared strings are opt-in.
• The XLSB writer emits BIFF12 workbook parts inside the standard XLSB ZIP package; shared strings are opt-in.
• The XLS writer buffers records in memory and assembles the OLE container at EndAsync; choose it when write throughput matters more than peak allocation.
• Excel.EncryptPackage/EncryptPackageAsync wrap a written XLSX/XLSB package in agile ECMA-376 encryption; see Encrypted workbooks.

Build

dotnet restore ExcelReader.slnx
dotnet build ExcelReader.slnx --configuration Release
dotnet test --project tests/ExcelReader.Tests/ExcelReader.Tests.csproj --configuration Release

Other languages

ExcelReader ships a NativeAOT shared library with a C ABI, so non-.NET languages can read and write XLSX, XLSB, XLS and CSV without a .NET runtime installed.

• C ABI header: src/ExcelReader.Native/include/excelreader.h
• Python package: python/
• C++ package: header-only CMake wrapper, xl::Workbook/xl::parse_sheet/xl::write_sheet over the same ABI — see cpp/README.md.
• Rust crate: safe Workbook/parse_sheet/write_sheet bindings, downloadable via cargo add excelreader — see rust/excelreader/README.md.

from excelreader import open_workbook

with open_workbook("book.xlsx") as workbook:
    for row in workbook.rows():
        print([cell.value for cell in row])

Writing goes through one export, xl_write_typed: a whole sheet in a single call, from columnar buffers the ABI borrows rather than copies. All three bindings expose it — Python as write_workbook/write_arrow/write_pandas/write_polars, C++ as xl::write_columns and xl::write_sheet<T>, Rust as writer::write_columns and writer::write_sheet. In C++ and Rust the same struct mapping drives both directions, so reading a sheet and writing it back needs one mapping, not two:

use excelreader::writer::write_sheet;
use excelreader::XL_FORMAT_XLSX;

write_sheet("out.xlsx", XL_FORMAT_XLSX, &rows, None)?;

auto written = xl::write_sheet("out.xlsx", rows);   // format inferred from the extension

Row-by-row decoded reads are available from all three bindings — Python as Workbook.rows(), C++ as xl::Workbook::rows(), Rust as Workbook::rows(). Python additionally exposes read_all_columnar() over xl_read_all_blob, which the other two do not wrap.

The Arrow export is available from Python (to_arrow/to_record_batch), C++ (xl::parse_arrow<T>, in the separate <xl/excelreader_arrow.hpp> header — no Apache Arrow C++ dependency, you get the raw C Data Interface pair), and Rust (excelreader::arrow::parse_arrow, behind the arrow cargo feature, returning an arrow::array::RecordBatch).

All three bindings can also read a sheet a batch at a time instead of the whole sheet in one call, so peak memory is one batch rather than one sheet — Python as iter_parse_typed/ to_record_batch_reader (plus iter_pandas/iter_polars), C++ as xl::typed_reader<T>/ xl::arrow_stream<T>, Rust as Workbook::typed_chunks/arrow::parse_arrow_stream. The .NET reader above already streams row-by-row by construction, so there is nothing to add here for it; see each binding's README for the chunked-reading semantics (one chunked read per workbook, batch_size meaning, and what invalidates a live one).

Encrypting a written package goes through one export too, xl_encrypt_package — wrap a finished plaintext XLSX/XLSB package in agile ECMA-376 encryption, given a password. All three bindings expose it: Python as encrypt_package, C++ as xl::encrypt_package, Rust as writer::encrypt_package; see each binding's README for the encrypted-workbooks section.

Contributing

See ARCHITECTURE.md for a map of the codebase, STYLEGUIDE.md for the code style, and CONTRIBUTING.md for build expectations and how to submit a change. Security issues should go through the private channel in SECURITY.md, not a public issue.

License

ExcelReader is licensed under the MIT License. See LICENSE.