Save To Zip¶
Aggregates outputs from multiple upstream nodes, organizes them into a structured folder layout, and creates a zip archive. The node intelligently names folders/files (with collision-avoidance) and, when available, uses metadata from inputs to preserve meaningful filenames and group related results.
Usage¶
Use this node at the end of a workflow when you want to package results into a single downloadable zip. Connect any outputs (e.g., sequences, PDB/A3M dictionaries, text, or other data) to results_1 and subsequent results_N inputs. The node will produce a URL/path to the created zip, which you can connect to the Output node to download.
Inputs¶
| Field | Required | Type | Description | Example |
|---|---|---|---|---|
| results_1 | True | WILDCARD | First input to include in the zip. Accepts any output from any node. Once connected, results_2 becomes available. | Any node output, e.g., a dictionary of PDBs, an A3M dict, or a string. |
| results_2 ... results_101 | False | WILDCARD | Additional inputs to include in the zip. Each appears after connecting the previous one, allowing up to 100 additional inputs. | Any node output, e.g., additional PDB/A3M dicts, FASTA strings, or other text values. |
Outputs¶
| Field | Type | Description | Example |
|---|---|---|---|
| url | STRING | URL or path to the folder containing the created zip archive for this workflow run. Connect to the Output node to click and download. | https://download/ |
Important Notes¶
- Sequential input reveal: results_2 only appears after connecting results_1, results_3 after results_2, and so on (up to results_101).
- Metadata-aware saving: Inputs produced by nodes that embed metadata (e.g., file_name, folder_name) are saved using those names, with automatic numbering to avoid clashes.
- Structured handling for PDB/A3M: When inputs are PDB or A3M dictionaries, each entry is saved as a separate file within a subfolder.
- Non-metadata inputs: Values without metadata are saved as text files (e.g., output_
.txt) in a folder named after their source node class. - Seeds and configs capture: If provided via input metadata, consolidated seeds.json and configs.json are added at the top level of the zip content.
- Collision handling: If a file/folder name already exists, a numeric suffix is appended to avoid overwriting.
- Execution priority: Designed to run at the end of a workflow to capture all upstream outputs.
- Output location: The node returns a URL/path where the zip can be downloaded; connect it to the Output node to expose a clickable link.
Troubleshooting¶
- Cannot see results_2 or further: Connect results_1 first; additional inputs are revealed only after the previous one is connected.
- Download link not clickable: Ensure the url output is connected to an Output node to render a clickable link.
- Unexpected file names or folders: If metadata is missing on an input, the node falls back to generic names (e.g., source node class and output_
.txt). To influence naming, use nodes that provide metadata-aware outputs. - Missing files in zip: Verify that each desired upstream output is connected to a results_N input and that the inputs are not None.
- Permission or path issues: If the generated path cannot be accessed or saved, check environment permissions and the configured output directory.
- Large inputs or many files: Packaging many or large outputs may take longer; allow the node to complete before attempting to download.
Example Pipelines¶
