Extract Data Tables from Arabic PDFs
Pull structured data from Arabic research reports and keep the numbers accurate
Arabic PDFs scramble when copied. Numbers switch between Arabic-Indic digits and Western numerals, decimal separators disappear, and right-to-left formatting breaks table structure.
That is the gap this drop closes. It is a data & analysis skill built for Kimi, and it takes about ten minutes to set up the first time. After that it runs in under a minute.
Who should use it
Analysts and researchers working with Arabic-language reports, industry studies, or government publications that contain data tables
How it works
The skill file does five things, in this order.
1. Specify the digit system you need. Tell Kimi whether your output needs Eastern Arabic numerals (٠-٩) or Western numerals (0-9) for consistency with your spreadsheet
2. Request CSV format with explicit separators. Ask for comma-separated values and specify that periods mark decimals to avoid confusion with Arabic comma usage
3. Name the columns in both languages. Include Arabic headers and English equivalents so you can work with the data in either language without re-labeling
4. Flag units and their Arabic abbreviations. Ask Kimi to preserve measurement units exactly as written because abbreviations differ (مليون vs mn, ألف vs k)
5. Request a row count verification. Have Kimi state how many data rows it extracted so you can cross-check against the original table before importing
What comes back
A clean CSV file ready for Excel or Google Sheets, with numbers in the format you need and no mixed scripts breaking your formulas
The mistake to avoid
If Kimi mixes digit systems within one column, paste the CSV output back and ask it to standardize all numbers to one system. This second pass catches inconsistencies the first extraction missed.
Running it
Paste the prompt block into a new Kimi chat and attach your files in the same message. Kimi's long context means you should upload everything at once rather than in batches, splitting the input is what causes it to lose cross-references between documents.
Where this fits
On its own, one skill saves an hour a week. The compounding happens when three or four of them run in sequence on the same input, the same transcript that produces a scope of work also produces the follow-up email and the project brief. That is the point at which it stops being a prompt and starts being an internal tool. If you want that wired into the systems your team already uses, that is the work 67 Digital does.
In the file
I have an Arabic PDF report with data tables. Please extract the table on page {page_number} that shows {brief_description_of_table_content}.
Format requirements:
- Output as CSV
- Use {Western numerals 0-9 / Eastern Arabic numerals ٠-٩} for all numbers
- Use periods for decimal separators
- Use commas to separate values
- Preserve all measurement units exactly as written in the original
Column headers:
- Provide both the original Arabic header and an English translation
- Format: Arabic Header (English Translation)
Data handling:
- If a cell contains a range (e.g., ٥-١٠), keep it as text
- If a cell is empty or shows a dash, use empty quotes
- Do not add any formatting, currency symbols, or percentage signs unless they appear in the original
After extraction:
- State the total number of data rows you extracted
- Note any cells where the value was unclear or ambiguous
Here is the document: {attach or paste relevant pages}