Extract Text and Tables from PDF

ExtractText returns the text of the whole document in reading order. Read returns a logical model: pages with paragraphs and detected tables, plus a Markdown rendering. Both work on PDFs produced by any application; scanned pages without a text layer need OCR first (see PDF workflows in the CLI).

Extract all text

C#

using OfficeIMO.Pdf;

PdfDocument pdf = PdfDocument.Load("statement.pdf");
string text = pdf.ExtractText();
File.WriteAllText("statement.txt", text);

PowerShell

Read-OfficeDocument -Path .\statement.pdf -As Text
Read-OfficeDocument -Path .\statement.pdf -As Text -ByPage

CLI

officeimo read statement.pdf --format markdown --output statement.md

Read pages, paragraphs, and tables

C#

using OfficeIMO.Pdf;

PdfDocumentReadResult result = PdfDocument.Load("statement.pdf").Read();

foreach (PdfLogicalPage page in result.Pages) {
    Console.WriteLine($"Page {page.PageNumber}: {page.Paragraphs.Count} paragraphs, {page.Tables.Count} tables");
    foreach (PdfLogicalTable table in page.Tables) {
        PdfLogicalTableData data = PdfLogicalTableAnalysis.Extract(table);
        foreach (IReadOnlyList<string> row in data.Rows) {
            Console.WriteLine(string.Join(" | ", row));
        }
    }
}

File.WriteAllText("statement.md", result.ExtractMarkdown());

PowerShell

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

CLI

officeimo read statement.pdf --format json --output statement.json

Convert PDF tables to Excel

To move every detected table into a workbook, use the PDF-to-Excel conversion route:

officeimo convert statement.pdf statement.xlsx

For searchable chunks with page locations, read the PDF through OfficeIMO.Reader instead.

Limits and reference

Limits: table detection works from the drawn layout, so tables without ruling lines or with merged cells may be split or merged differently from the source. See the PDF package row and the PDF to Excel row.

Reference: OfficeIMO.Pdf API · Read-OfficeDocument