Arabic tashkeel and text editing online | Itqan

Arabic text work rarely fails for lack of ideas — it fails in the messy middle: pasted CAPS from a PDF, tashkeel you need for a learner then must strip for search, mixed Arabic-Indic digits, and word counts that disagree between tools. This checklist-only guide walks an editing path on Itqan with case converter (tashkeel, cleanup, Latin case), translator for short bilingual drafts, and word counter for length gates — so you finish publishable Arabic without fighting your clipboard.

Checklist mindset: one job per pass

Treat each pass as a single responsibility. Mixing “add tashkeel + translate + count + title-case English headings” in one frantic paste creates undiagnosable errors. The cards on the case converter are designed for chaining with Apply: each apply becomes the new input for the next card. Keep translator and word counter as sibling tabs for language and metrics, not as substitutes for cleanup.

Intake checklist — before you change a character

• □ Paste into case converter and skim for mojibake (اͅ style garbage) — if present, re-copy from a Unicode-safe source.
• □ Note whether you need full tashkeel, partial marks, or a bare skeleton for indexing.
• □ Separate English fragments you must title-case from Arabic body text you must not “uppercase.”
• □ Decide if length limits are words, characters with spaces, or characters without — then open word counter with the same definition later.
• □ If the source is a scan, run OCR first; do not diacritize OCR noise.

Cleanup checklist — make text predictable

On the case converter, use the cleanup cards before linguistic tools:

• □ Trim repeated spaces and odd line breaks from chat or PDF copy.
• □ Normalize punctuation variants that break search (mixed Arabic/Latin commas and question marks) according to your house style.
• □ Remove leftover control characters from rich-text pastes when the preview looks “empty but long.”
• □ Apply cleanup results back to input so every later card sees the cleaned string.
• □ Keep a raw backup in a second notepad until the chain is finished.

Why cleanup first? Automatic tashkeel and translation amplify whatever you feed them. Clean input yields fewer invented marks and fewer mistranslated fragments.

Tashkeel checklist — add or remove with intent

Arabic diacritics (harakat / tashkeel) guide pronunciation and disambiguate lookalike undiacritized forms. Itqan’s case converter includes Arabic tools for this workflow; automatic diacritization calls the server-side Mishkal path, while most other cards stay local in the browser.

• □ Add tashkeel for learners, Qur’anic study aids you are allowed to process, children’s content, or TTS prep — then human-review ambiguous proper names.
• □ Strip tashkeel before search indexes, many ERP fields, and hashtags that break on combining marks.
• □ Never publish automatic diacritics on legal or liturgical text without a qualified reviewer.
• □ After diacritize, read aloud or use text to speech on a sample paragraph to catch absurd vocalizations.
• □ If you only needed three words marked, prefer manual marks over full-document automation.

Deeper card-by-card behavior lives in the case converter guide; this topic checklist focuses on sequence.

Latin case checklist — bilingual documents

Arabic letters do not have uppercase/lowercase. Case cards mainly reshape English (or other cased scripts) inside mixed text:

• □ UPPERCASE for warning labels and OCR headings that arrived screaming.
• □ lowercase as a neutral baseline after caps-lock disasters.
• □ Title Case for English headlines — then fix small words your style guide leaves lowercase.
• □ Do not expect case tools to “fix” Arabic shape; use Arabic cards instead.
• □ Apply case on English-only selections when possible; full-document title case can mangle brand acronyms (use judgment).

Translation checklist — short text, clear purpose

When a paragraph must move between Arabic and another language, open translator rather than forcing PDF translate on three sentences.

• □ Clean and (if needed) strip tashkeel before translation if marks confuse the engine for your domain.
• □ Set source and target explicitly; do not rely on auto-detect for bilingual paste.
• □ Translate meaning units (paragraphs), not half-lines broken by PDF copy.
• □ Re-check numbers, dates, and names against the source after translation.
• □ For whole PDFs with layout, switch to PDF translate instead — see that guide.
• □ Machine output is a draft; certified needs stay with human professionals.

Tool-page walkthrough: translator guide.

Word count checklist — gate length honestly

Open word counter when forms, SEO fields, subtitles, or assignments impose limits.

• □ Paste the final cleaned text — counting mid-chain drafts wastes decisions.
• □ Confirm whether the gate counts words or characters (with/without spaces).
• □ Remember Arabic clitics and punctuation can shift “word” splits versus English expectations.
• □ If you must cut, trim filler sentences before stripping meaning-bearing tashkeel you still need for learners.
• □ Save a screenshot of the counter result when a portal disputes your length.

More detail: word counter guide.

Recommended Apply chains (copy these)

Chain A — Learner paragraph with vocalization

1. Paste raw text → cleanup cards → Apply.
2. Run Arabic tashkeel → review names → Apply if accepted.
3. Word-count for the worksheet limit.
4. Optional: translator for an English teacher key on a copy, not on the only file.

Chain B — Web article from messy PDF paste

1. Cleanup repeated spaces and broken lines → Apply.
2. Strip tashkeel if the CMS search hates marks → Apply.
3. Title Case English subheads only (or fix manually).
4. Word counter for meta description length.
5. Translator only for a short bilingual abstract.

Chain C — Support reply Arabic → English

1. Cleanup chat paste.
2. Strip decorative tatweel/kashida if present.
3. Translator to English for the agent’s twin ticket.
4. Count characters if SMS/email templates are capped.

Four checklist scenarios

University worksheet

Instructor needs vocalized examples. Intake → cleanup → tashkeel → human review of verbs → word count for page budget → print. Do not translate the vocalized Arabic away before students see it.

E-commerce titles

Strip tashkeel for URL slugs and search facets; keep a vocalized marketing blurb elsewhere if branding wants it. Count characters for marketplace title caps.

HR policy bilingual summary

Cleanup Arabic source → translate a one-page summary → counsel reviews both → counter for intranet teaser length. Full policy PDFs use the PDF pipeline, not this text checklist alone.

Social caption from ALL CAPS English + Arabic body

lowercase/Title Case the English line → cleanup Arabic → count for platform limits → publish. Skip tashkeel unless the brand voice requires it.

Anti-checklist — habits that waste an hour

• Diacritizing OCR garbage, then “fixing” marks by hand for an hour.
• Title-casing an entire Arabic article and wondering why nothing changed (or why English brands broke).
• Counting words before cleanup so spaces and blank lines inflate the score.
• Translating with wrong source language on bilingual paste.
• Uploading confidential HR text to any online tool against policy — read privacy and minimize.
• Using reverse-text “for fun” on a production string and saving it by accident.

Pre-publish master checklist

1. □ Encoding looks correct; no replacement characters.
2. □ Cleanup applied; no double spaces or broken RTL isolates you can see.
3. □ Tashkeel decision explicit: present, partial, or stripped — and reviewed if added automatically.
4. □ English case matches house style; Arabic unaffected wrongly.
5. □ Translation (if any) verified for numbers and names.
6. □ Word/character gate passed on the final string.
7. □ Backup of pre-chain text stored.
8. □ Sensitive content allowed to be processed per team policy.

FAQ

Related guides

• Case converter — cards and Apply chain
• Translator guide
• Word counter guide
• Text to speech for hearing diacritized drafts
• AI PDF summarize & translate for full documents
• Arabic OCR when the source is still a scan

Arabic editing on Itqan is a checklist sport: intake, clean, decide tashkeel, fix Latin case, translate only what must move, count the final string. Start on case converter, keep translator and word counter ready for the next single responsibility, and ship text that reads as intentional — not as a clipboard accident.

When source material is still a scan, run scan to searchable PDF and Arabic OCR before you paste recognized text into the converter — garbage recognition poisons every checklist step that follows.

Content ops that treat tashkeel as an afterthought pay in brand trust. Put the Monday–Friday rhythm on a shared calendar, assign one “diacritic owner” per language pair, and refuse to ship bilingual campaigns until both the Arabic edit checklist and the word-count check are green. Tools on Itqan make the clicks cheap; the checklist makes the decisions repeatable.

A weekly rhythm for solo Arabic editors

Many freelancers only diacritize on delivery day, which raises error pressure. Use a light cadence: Monday collect the week’s drafts in one folder without tashkeel. Wednesday run cleanup and strip accidental diacritics in the case converter, then add tashkeel only to teaching passages, poetry, or what the client explicitly ordered. Thursday hand-check hamza, tāʾ marbūṭa, and proper names. Friday count the final string in the word counter before handoff, and keep both undiacritized and diacritized copies when the contract asks for two. The rhythm keeps tools as helpers, not substitutes for editorial judgment.

If source material arrives as a scan, do not paste straight into the diacritic tool. Finish scan to searchable PDF and review OCR using the Arabic OCR guide, then move only clean text. For partial translation, use the translator on sentences that must move languages — not the whole file by default. Card-level steps live in the case converter how-to.