Audit right-to-left lines and embedded fields
Complete source
A full-width crop that keeps right-to-left lines and embedded numbers together.
Direction relationship
Script direction controls word order, while numbers and Latin terms may run differently.
Identity check
Check that punctuation, line order, and labels remain attached to the right text.
Direction must be checked separately
Check logical line order and displayed order separately, especially around numbers and Latin text.
Capture complete lines
Do not crop so tightly that the beginning or end of a right-to-left line is lost. Keep headings and adjacent labels visible, especially when a form places a value to one side of its label.
Arabic letter shapes connect and can change by position. Hebrew points and Arabic vowel marks are small and may be omitted in everyday text, but when present they need sharp capture because they can affect reading.
Audit mixed-direction content
Phone numbers, dates, currencies, URLs, and Latin brand names may run left to right inside a right-to-left sentence. Compare their visual order with the source instead of assuming the plain-text display preserved it.
Brackets and neutral punctuation can appear on an unexpected side when direction metadata is missing. Verify the words enclosed by the punctuation, not only the visible bracket orientation.
Pages that combine right-to-left lines with Latin labels also benefit from the workflow for vertical text and mixed scripts.
Preserve names and official forms
Keep the original script for personal names, addresses, institutions, and place names beside any transliteration. More than one Latin spelling may be reasonable, while an official document may require one established form.
For legal, identity, immigration, medical, or religious material, use a qualified translator familiar with the variety and purpose of the text. Script recognition alone does not resolve dialect, specialized meaning, or official naming rules.
When small vowel or pronunciation marks affect a name or word, apply the checks for diacritics and accented text in OCR.
Handle right-to-left Arabic or Hebrew text, mixed numbers, Latin names, punctuation, and line order in image translation. Continue with the languages and scripts guide collection for adjacent workflows.