Extract Tables and Records

Tables come out of every format in the same shape: a ReaderTable with column names, rows of strings, its location in the source, and whether it was truncated. That means the same code exports a table from a Word document, a PDF page, or an Excel range.

Extract tables

C#

using OfficeIMO.Reader;
using OfficeIMO.Reader.Pdf;
using OfficeIMO.Reader.Word;

OfficeDocumentReader reader = new OfficeDocumentReaderBuilder()
    .AddWordHandler()
    .AddPdfHandler()
    .Build();

OfficeDocumentReadResult document = reader.ReadDocument("Quote.pdf");
int index = 0;
foreach (ReaderTable table in document.ExtractTables()) {
    index++;
    Console.WriteLine($"{table.Title ?? "Table"}: {table.Columns.Count} columns, {table.Rows.Count} rows");
    File.WriteAllText($"quote-table-{index}.csv", table.ToCsv());
    File.WriteAllText($"quote-table-{index}.md", table.ToMarkdownTable());
}

ReaderOptions.MaxTableRows (200 by default) bounds rows per table; TotalRowCount and Truncated tell you when a table was cut.

PowerShell

Read-OfficeDocument -Path .\Quote.pdf -As Tables

Extract key-value records and sections

ExtractStructured returns key-value records (such as "Invoice number: 123" lines and two-column tables) and the document's sections.

C#

using OfficeIMO.Reader;
using OfficeIMO.Reader.Word;

OfficeDocumentReader reader = new OfficeDocumentReaderBuilder().AddWordHandler().Build();
OfficeDocumentStructuredExtractionResult structured = reader.ReadDocument("Contract.docx").ExtractStructured();
Console.WriteLine($"{structured.Records.Count} records, {structured.Sections.Count} sections");

PowerShell

Read-OfficeDocument -Path .\Contract.docx -As Structure

Limits and reference

Limits: table detection in PDFs works from the drawn layout; see Extract text and tables from PDF. See the compatibility matrix for per-format coverage.

Reference: OfficeIMO.Reader API