Output Format

Overview

The gws CLI is designed for both human and machine consumption. Every response — success, error, or download metadata — is structured JSON by default. This makes it trivial to pipe output to jq, parse in scripts, or process in AI agent workflows.

Unlike traditional CLIs that mix human-readable text with machine-parseable output, gws guarantees all stdout is valid JSON (or NDJSON for paginated responses).

Default: Pretty-Printed JSON

Without any flags, API responses are printed as pretty-printed JSON:

gws drive files list --params '{"pageSize": 2}'

Output:

{
  "kind": "drive#fileList",
  "files": [
    {
      "kind": "drive#file",
      "id": "1a2b3c4d5e6f",
      "name": "My Document",
      "mimeType": "application/vnd.google-apps.document"
    },
    {
      "kind": "drive#file",
      "id": "7g8h9i0j1k2l",
      "name": "Spreadsheet Q1 2025",
      "mimeType": "application/vnd.google-apps.spreadsheet"
    }
  ],
  "nextPageToken": "CAESBQgBIAEoAQ"
}

Alternative Formats

Use the --format flag to change output formatting:

gws drive files list --params '{"pageSize": 2}' --format table
gws drive files list --params '{"pageSize": 2}' --format yaml
gws drive files list --params '{"pageSize": 2}' --format csv

Table Format

id              name                   mimeType
──────────────  ─────────────────────  ────────────────────────────────────
1a2b3c4d5e6f    My Document            application/vnd.google-apps.document
7g8h9i0j1k2l    Spreadsheet Q1 2025    application/vnd.google-apps.spreadsheet

Features:

Automatically extracts array data from list responses
Nested objects are flattened into dot-notation columns (e.g., owner.displayName)
Columns are auto-sized (capped at 60 chars)
Multi-byte UTF-8 characters are handled safely

Implementation in src/formatter.rs:163-271:

fn format_array_as_table(arr: &[Value], emit_header: bool) -> String {
    if arr.is_empty() {
        return "(empty)\n".to_string();
    }

    // Flatten each row so nested objects become dot-notation columns.
    let flat_rows: Vec<Vec<(String, String)>> = arr
        .iter()
        .map(|item| match item {
            Value::Object(obj) => flatten_object(obj, ""),
            _ => vec![(String::new(), value_to_cell(item))],
        })
        .collect();

    // Collect all unique column names (preserving insertion order).
    let mut columns: Vec<String> = Vec::new();
    for row in &flat_rows {
        for (key, _) in row {
            if !columns.contains(key) {
                columns.push(key.clone());
            }
        }
    }
    // ... calculate widths, render table
}

YAML Format

kind: "drive#fileList"
files:
  - kind: "drive#file"
    id: "1a2b3c4d5e6f"
    name: "My Document"
    mimeType: "application/vnd.google-apps.document"
  - kind: "drive#file"
    id: "7g8h9i0j1k2l"
    name: "Spreadsheet Q1 2025"
    mimeType: "application/vnd.google-apps.spreadsheet"
nextPageToken: "CAESBQgBIAEoAQ"

Features:

Single-line strings are always double-quoted to avoid YAML parser ambiguity (# comments, : mappings)
Multi-line strings use block scalar (|) syntax
Nested structures preserve hierarchy

CSV Format

id,name,mimeType
1a2b3c4d5e6f,My Document,application/vnd.google-apps.document
7g8h9i0j1k2l,Spreadsheet Q1 2025,application/vnd.google-apps.spreadsheet

Features:

Automatically extracts array data from list responses
Column order preserved based on first appearance
Values containing ,, ", or newlines are properly escaped

Error Format

All errors are output as structured JSON with consistent fields:

{
  "error": {
    "code": 403,
    "message": "The user does not have sufficient permissions for file 1a2b3c4d5e6f.",
    "reason": "insufficientPermissions"
  }
}

Error types and their reason codes:

Error Type	Code	Reason	Example
Validation	400	`validationError`	Missing required parameter
Authentication	401	`authError`	No credentials provided
API Error	Varies	`<api-specific>`	`insufficientPermissions`, `notFound`, etc.
Discovery Error	500	`discoveryError`	Failed to fetch Discovery Document
Internal Error	500	`internalError`	Unexpected error

Implementation in src/error.rs:42-93:

impl GwsError {
    pub fn to_json(&self) -> serde_json::Value {
        match self {
            GwsError::Api {
                code,
                message,
                reason,
                enable_url,
            } => {
                let mut error_obj = json!({
                    "code": code,
                    "message": message,
                    "reason": reason,
                });
                if let Some(url) = enable_url {
                    error_obj["enable_url"] = json!(url);
                }
                json!({ "error": error_obj })
            }
            GwsError::Validation(msg) => json!({
                "error": {
                    "code": 400,
                    "message": msg,
                    "reason": "validationError",
                }
            }),
            // ... other error types
        }
    }
}

`accessNotConfigured` Errors

When a required API is not enabled for your GCP project, the error includes a direct link to enable it:

{
  "error": {
    "code": 403,
    "message": "Gmail API has not been used in project 549352339482 before or it is disabled. Enable it by visiting https://console.developers.google.com/apis/api/gmail.googleapis.com/overview?project=549352339482 then retry.",
    "reason": "accessNotConfigured",
    "enable_url": "https://console.developers.google.com/apis/api/gmail.googleapis.com/overview?project=549352339482"
  }
}

A human-readable hint is also printed to stderr (not stdout, so it doesn’t pollute JSON parsing):

💡 API not enabled for your GCP project.
   Enable it at: https://console.developers.google.com/apis/api/gmail.googleapis.com/overview?project=549352339482
   After enabling, wait a few seconds and retry your command.

Piping to `jq`

jq is the standard tool for filtering JSON output:

# Extract just file names
gws drive files list --params '{"pageSize": 10}' | jq -r '.files[].name'

# Filter files by MIME type
gws drive files list --params '{"pageSize": 100}' \
  | jq '.files[] | select(.mimeType == "application/pdf")'

# Count total files
gws drive files list --params '{"pageSize": 1000}' | jq '.files | length'

# Extract specific fields into CSV
gws drive files list --params '{"pageSize": 50}' \
  | jq -r '.files[] | [.id, .name, .mimeType] | @csv'

Pagination and NDJSON

Many Google Workspace APIs return paginated results with a nextPageToken. The CLI supports auto-pagination with the --page-all flag:

gws drive files list --params '{"pageSize": 100}' --page-all

With --page-all, output is NDJSON (newline-delimited JSON) — one JSON object per line:

{"files":[{"id":"1","name":"file1.pdf"}],"nextPageToken":"token1"}
{"files":[{"id":"2","name":"file2.pdf"}],"nextPageToken":"token2"}
{"files":[{"id":"3","name":"file3.pdf"}]}

This format is:

Streamable: You can process each line as it arrives
Memory-efficient: No need to load the entire result set into memory
jq-compatible: jq -r '.files[].name' works on NDJSON

Implementation in src/formatter.rs:79-88:

pub fn format_value_paginated(value: &Value, format: &OutputFormat, is_first_page: bool) -> String {
    match format {
        OutputFormat::Json => serde_json::to_string(value).unwrap_or_default(),
        OutputFormat::Csv => format_csv_page(value, is_first_page),
        OutputFormat::Table => format_table_page(value, is_first_page),
        OutputFormat::Yaml => format!("---\n{}", format_yaml(value)),
    }
}

Pagination Options

Flag	Description	Default
`--page-all`	Auto-paginate, one JSON line per page (NDJSON)	off
`--page-limit <N>`	Max pages to fetch	10
`--page-delay <MS>`	Delay between pages (in milliseconds)	100 ms

Example:

# Fetch up to 50 pages with 200ms delay between each
gws drive files list --params '{"pageSize": 100}' \
  --page-all --page-limit 50 --page-delay 200

Extracting Data from NDJSON

Process each page’s file array:

gws drive files list --params '{"pageSize": 100}' --page-all \
  | jq -r '.files[].name'

Or concatenate all pages into a single array:

gws drive files list --params '{"pageSize": 100}' --page-all \
  | jq -s '[.[].files[]]'

Binary Downloads

When downloading file content (e.g., gws drive files get --params '{"fileId":"...", "alt":"media"}'), the CLI:

Detects the Content-Type header
Streams the response to a file (default: download.<ext>)
Outputs JSON metadata about the download:

{
  "status": "success",
  "saved_file": "download.pdf",
  "mimeType": "application/pdf",
  "bytes": 245829
}

You can specify the output path with --output:

gws drive files get --params '{"fileId":"abc123","alt":"media"}' \
  --output report.pdf

Implementation in src/executor.rs:295-341:

async fn handle_binary_response(
    response: reqwest::Response,
    content_type: &str,
    output_path: Option<&str>,
    output_format: &crate::formatter::OutputFormat,
    capture_output: bool,
) -> Result<Option<Value>, GwsError> {
    let file_path = if let Some(p) = output_path {
        PathBuf::from(p)
    } else {
        let ext = mime_to_extension(content_type);
        PathBuf::from(format!("download.{ext}"))
    };

    let mut file = tokio::fs::File::create(&file_path)
        .await
        .context("Failed to create output file")?;

    let mut stream = response.bytes_stream();
    let mut total_bytes: u64 = 0;

    while let Some(chunk) = stream.next().await {
        let chunk = chunk.context("Failed to read response chunk")?;
        file.write_all(&chunk)
            .await
            .context("Failed to write to file")?;
        total_bytes += chunk.len() as u64;
    }

    file.flush().await.context("Failed to flush file")?;

    let result = json!({
        "status": "success",
        "saved_file": file_path.display().to_string(),
        "mimeType": content_type,
        "bytes": total_bytes,
    });

    if capture_output {
        return Ok(Some(result));
    }

    println!("{}", crate::formatter::format_value(&result, output_format));

    Ok(None)
}

CSV and Table Pagination

When using --page-all with --format csv or --format table, the CLI only emits column headers on the first page. Subsequent pages contain only data rows, so the combined output is machine-parseable:

gws drive files list --params '{"pageSize": 2}' --page-all --format csv

Output:

id,name,mimeType
1a2b3c4d5e6f,My Document,application/vnd.google-apps.document
7g8h9i0j1k2l,Spreadsheet Q1 2025,application/vnd.google-apps.spreadsheet
3m4n5o6p7q8r,Presentation Deck,application/vnd.google-apps.presentation
9s0t1u2v3w4x,Budget 2025,application/vnd.google-apps.spreadsheet

No duplicate header rows between pages.

Use Cases

Scripting

#!/bin/bash
for file_id in $(gws drive files list --params '{"q":"mimeType='application/pdf'"}' \
  | jq -r '.files[].id'); do
  echo "Processing $file_id"
  gws drive files get --params "{\"fileId\":\"$file_id\",\"alt\":\"media\"}" \
    --output "pdfs/$file_id.pdf"
done

AI Agents

AI agents can reliably parse all output without regex hacks or brittle text parsing:

import subprocess
import json

result = subprocess.run(
    ["gws", "drive", "files", "list", "--params", '{"pageSize": 10}'],
    capture_output=True,
    text=True
)
data = json.loads(result.stdout)
files = data["files"]
for f in files:
    print(f"File: {f['name']} (ID: {f['id']})")

Monitoring

Check if a specific file exists (exit code 0 = success, 1 = error):

if gws drive files get --params '{"fileId":"abc123"}' > /dev/null 2>&1; then
  echo "File exists"
else
  echo "File not found or inaccessible"
fi

Dynamic Discovery

How the CLI generates commands at runtime

Pagination

Deep dive into auto-pagination strategies

Get Started

Core Concepts

Authentication

Command Reference

Advanced Usage

AI Integration

Development

Troubleshooting

Overview

Default: Pretty-Printed JSON

Alternative Formats

Table Format

YAML Format

CSV Format

Error Format

`accessNotConfigured` Errors

Piping to `jq`

Extracting Data from NDJSON

Binary Downloads

Use Cases

Scripting

AI Agents

Monitoring

Dynamic Discovery

Pagination

Get Started

Core Concepts

Authentication

Command Reference

Advanced Usage

AI Integration

Development

Troubleshooting

Documentation Index

​Overview

​Default: Pretty-Printed JSON

​Alternative Formats

​Table Format

​YAML Format

​CSV Format

​Error Format

​accessNotConfigured Errors

​Piping to jq

​Pagination and NDJSON

​Pagination Options

​Extracting Data from NDJSON

​Binary Downloads

​CSV and Table Pagination

​Use Cases

​Scripting

​AI Agents

​Monitoring

​Related Concepts

Dynamic Discovery

Pagination

Overview

Default: Pretty-Printed JSON

Alternative Formats

Table Format

YAML Format

CSV Format

Error Format

`accessNotConfigured` Errors

Piping to `jq`

Pagination and NDJSON

Pagination Options

Extracting Data from NDJSON

Binary Downloads

CSV and Table Pagination

Use Cases

Scripting

AI Agents

Monitoring

Related Concepts