> ## Documentation Index
> Fetch the complete documentation index at: https://parser.xdr.ooo/llms.txt
> Use this file to discover all available pages before exploring further.
# Auto Parser: source values have no names
> Keep all extracted values and enter trustworthy field names when available.
# My log has values separated by commas, but no field names
**Situation:** You pasted lines of values and clicked **Analyze Logs**. Parser Studio says that the values were found, but their names are not trustworthy. That is expected when the source contains only positions.
*The app identifies five columns but does not guess what they mean.*
## Practice sample
```csv theme={null}
2026-10-18T13:00:00Z,allow,192.0.2.10,198.51.100.22,443
2026-10-18T13:00:01Z,deny,192.0.2.11,203.0.113.7,22
```
For this fictional practice data, assume the product's specification says the column order is **timestamp, action, src\_ip, dst\_ip, dst\_port**.
## Walkthrough
1. Select **Auto Parser**, paste the complete sample and choose **Analyze Logs**.
2. If a warning says values were found but there are no names, look for **Names to give each value**.
3. If you **have the vendor field document**, type the field names in exactly the published order.
4. Under **Character between the values**, select the actual delimiter (for this example, **Comma (,)**).
5. Click **I Confirm These CSV Column Names** when the application enables it.
6. If you **do not** have a trustworthy definition, click **Keep all 5 values with temporary names**. Temporary names such as `field_001` preserve the data, but are not official field meanings.
7. Continue to **Review Fields**, compare column positions against the raw line, then enter **Map Fields** and correct the destinations.
8. Only then continue to **Create Parser** and the final test.
*When you do not have trustworthy names, preserve all five columns with temporary names and confirm the review before mapping.*
## Expected review outcome
All five values are preserved in their original order on every record. Source-backed names are used only if you actually confirmed what they mean.
## Recovery path
If the column count changes between rows, stop. Some logs from the same product have different shapes. Follow [the multiple-log-types guide](/scenarios/mixed-events) instead of stretching one column definition over every row.
A field named `field_001` is intentionally neutral. Do not label it "timestamp" merely because a sample value looks like a date unless the source definition supports that interpretation.
This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.