Parser
On this page
Purpose
Parser formats Data or DataFrame values with a template, or converts the input to a string. It does not parse arbitrary documents, HTML, XML, CSV, or PDF files.
When to use it
- Format structured rows for a prompt.
- Turn a DataFrame into predictable text.
- Convert an already structured value to a message.
How to use it
- Connect Data or a DataFrame.
- Select Parser to apply a template, or Stringify to convert the input to text.
- In Parser mode, enter a template such as
Name: {name}. - Send the Message output to a Prompt Template, Language Model, or output component.
Configuration
| Setting | What it controls | Recommended starting point |
|---|---|---|
| Mode | Selects Parser or Stringify. The default is Parser. | Use Parser when field formatting matters. |
| Template | Defines the formatted text in Parser mode. The default is Text: {text}. | Replace it when the input has no exact text field. |
| Separator | Joins formatted DataFrame rows. The default is a newline. | Keep the default for readable multi-row output. |
| Clean Data | Does not affect Parser or Stringify output. | Leave it off. |
Expected result
Parser returns one Message containing the formatted or stringified value. Template placeholders are case-sensitive and must exactly match Data keys or DataFrame columns. Missing fields and unsupported input types return an error.
Reference details
Use a file, web, or data component to obtain structured content first. Use Parser after that step when a downstream prompt needs stable text.