Keep the exponent with its value

Check the saved and reopened value so the receiving application preserves the magnitude you intended to hand over.

Before paying for technical-document extraction, inspect how the source meaning survives in the output. ParseMill’s public material-register example gives you a synthetic PDF and downloadable evidence to compare. Bring your own document type and acceptance checks to a scope discussion; the self-service invoice plan is a separate offer.

Treat the coefficient, multiplication sign, exponent base and exponent sign as meaningful components. A change from 10⁻³ to 10³ changes magnitude substantially while leaving most characters unchanged. Source review must include the exponent, not just the leading digits.

Use a notation round-trip fixture

Raw sourceField roleExpected handling
3.2E-4Measured resultDerived 0.00032 after verification
3.2 × 10⁻⁴Measured resultPreserve raw; explicit conversion rule needed
12E04Product identifierText; never infer measurement
3.2 × 10?Unclear exponentUnresolved; no numeric guess

These invented values demonstrate different input forms. The first derived number follows ordinary arithmetic. The second may require a parser-specific conversion that should be tested rather than assumed. The identifier remains exactly 12E04 regardless of whether a spreadsheet can interpret it numerically.

Microsoft: Number.FromText documents support for common numeric strings including exponent notation. It does not imply support for every typographic scientific expression copied from a PDF.

Test the conversion path

  1. Preserve the source string and unit in text fields. Inspect superscripts, exponent signs and any line breaks in the rendered source.
  2. Declare identifiers as text before import. Apply numeric conversion only to confirmed measurement fields with supported notation.
  3. Compare the normalized value with an independently calculated expected value for a small fixture. Include positive and negative exponents.
  4. Export, reopen and compare both raw strings and normalized numbers. Record the application and import settings used for the check.

Microsoft: Keeping leading zeros and large numbers describes automatic conversions involving the letter E and text-preserving import. Use explicit types to keep identifier-like strings from becoming scientific numbers.

Recognize unsupported cases

A superscript lost during extraction may leave an ordinary digit in an ambiguous position. Do not repair it solely from expected magnitude. Return to the source and record the corrected transcription. If the exponent is unreadable, preserve the partial text and hold the numeric field.

Very small or large values also need a destination type that can represent them. A displayed zero can reflect rounding or a representation limit rather than a source zero. Check the stored value and the raw expression before accepting the export.

Verify a meaningful outcome

The fixture passes when 3.2E-4 remains traceable to 0.00032, the typographic form retains its original expression, and 12E04 remains an identifier. The unreadable exponent stays unresolved. A successful numeric parse is only one part of this check.

Scientific reports require a separate assessment of exponent notation and output types. Inspect the public material-register example, then bring this fixture to a scoped discussion. The self-service invoice trial does not demonstrate scientific-report extraction.

Put it to work with ParseMill

Check the example against the work you need done

Compare the synthetic report, extracted rows and source references. For your own supplier reports, discuss the document type and review requirements before agreeing to a pilot.

Also working with supplier invoices? Try the invoice-to-Excel workspace. Any wider document pilot needs an agreed scope and fee.

Invoice processing: $49 USD/month · 500 PDF pages · one user. Buy only after you have checked the fit on your own invoices.

Sources and editorial notes

Written by ParseMill. Worked examples are illustrative process examples, not customer results. Software availability can vary by subscription and region; check the linked documentation for your account. Review decisions and approvals remain with your team.

Explore the wider workflow