The useful result
Preserve identifier strings exactly while normalizing only permitted numeric columns.
- Preserve identifier strings exactly while normalizing only permitted numeric columns.
An identifier that looks numeric may still be text. If a cleanup turns 0017 into 17, a record can stop matching its source even though the column looks tidier. Give Fable a column contract before asking for normalization, and check the exported text as well as what a spreadsheet displays.
Fable 5.1 requires eligible paid access; plan inclusion and usage credits differ. Check the linked account guide. You can practise the technique on the supplied fictional inputs with an available compatible assistant first. This lesson is documentation-based; examples are expected results, not a claim that we ran the named model.
Before you start
- A fictional CSV: id,quantity with rows 0017,2 and 1E10,3.
- A plain-text viewer and, optionally, a spreadsheet with explicit import column types.
Review evidence
Documentation-based review · 2026-10-18
Provider documentation checked on 18 October 2026. Examples and checks are original illustrative exercises, not recorded platform tests. Cover artwork is conceptual and does not imply provider sponsorship.
Declare the column contract
Mark id as an exact string and quantity as an integer. Do not let the assistant infer data types from appearance. Keep a copy of the original CSV and identify which transformations are allowed, such as trimming outer spaces from quantity.
Example / expected result
Illustrative example; not a recorded model result.
Protected IDs: “0017” and “1E10”. Allowed quantity values: integers 2 and 3.
Check your result
The contract explicitly protects leading zeros and scientific-looking identifiers.
Request a transformation log
Ask for the cleaned CSV plus a record of permitted changes and unresolved values. Require a separate report if a proposed change would affect an identifier. Keep descriptions outside the CSV block.
Clean this fictional CSV using id as an exact string and quantity as an integer: id,quantity / 0017,2 / 1E10,3. Preserve every id character, including leading zeros. Return one CSV block and a separate change log. Do not infer missing rows or silently repair identifiers.Example / expected result
Illustrative example; not a recorded model result.
id,quantity "0017",2 "1E10",3 Change log: no identifier changes.
Check your result
IDs remain byte-for-byte identical; no numeric expansion of 1E10 occurs.
Check export and import independently
Inspect the raw file in a text viewer. If opening it in a spreadsheet, import the identifier column as Text; CSV quotes alone do not force every spreadsheet to preserve its type. Compare the resulting values with the original IDs before using joins or lookups.
Check your result
Both raw CSV and imported cells retain 0017 and 1E10 exactly.
When it goes wrong
An application may change types after the AI has produced a correct file. CSV has no universal column-type metadata. Dates, long numbers and postal codes also need explicit handling; do not repair a real ID from guesswork.
Make it your own
Add ID 00003 and a quantity of 4. Keep the three IDs unchanged, then check that the quantity total is 9 without summing or transforming the identifier column.
Useful next resources
Sources and review
Last checked Oct 18, 2026. Features and account limits can change. Verify important results against your original material.
Did this help you complete your task?
Tell us what needs improving. No account or private output required.
Report incorrect information or a broken linkGenius Forges Editorial
Practical AI lessons with checked sources, labelled examples and human verification.
