xleak — Excel Data Extraction

To limit exported rows, pipe the output:

# JSON: use jq to slice the data array
xleak file.xlsx --export json | jq '{sheet, rows, columns, headers, data: .data[:100]}'

# CSV: use head (add 1 for header row)
xleak file.xlsx --export csv | head -n 101   # 100 data rows + 1 header

JSON Output Structure

When using --export json, xleak produces:

{
  "sheet": "Sheet1",
  "rows": 18081,
  "columns": 35,
  "headers": ["Col1", "Col2", ...],
  "data": [
    ["val1", 42, null, ...],
    ...
  ]
}

• rows / columns: total counts in the sheet (useful for knowing size before reading all data)
• headers: column names from the first row
• data: array of arrays, with typed values (strings, numbers, booleans, null)

Recommended Workflow

1. Start with a preview to understand the file structure:

   xleak file.xlsx -n 5
   

   This shows a formatted table with sheet name, row/column counts, and sample data.

2. For programmatic use, export as JSON and slice if the file is large:

   # Small files (< 1000 rows): export everything
   xleak file.xlsx --export json
   
   # Large files: slice to what you need
   xleak file.xlsx --export json | jq '{sheet, rows, columns, headers, data: .data[:50]}'
   

3. For multi-sheet workbooks, check available sheets first:

   xleak file.xlsx -n 1          # shows default sheet info
   xleak file.xlsx --sheet 2 -n 1  # check second sheet
   

4. For named tables (.xlsx only):

   xleak file.xlsx --list-tables
   xleak file.xlsx --table "TableName" --export json
   

jq Filtering Patterns

When piping xleak JSON through jq, zsh escapes != as \!= which breaks jq filters. This is the most common pitfall when working with xleak output programmatically.

Safe patterns (work in both bash and zsh):

# Filter rows where column index 12 has a value (truthy check — avoids != entirely)
xleak file.xlsx --sheet "Field Dictionary" --export json \
  | jq -r '.data[] | select(.[12]) | [.[1], .[3], .[12]] | @tsv'

# Filter non-null with pipe syntax (zsh-safe alternative to != null)
xleak file.xlsx --export json | jq '.data[] | select(.[5] | . != null)'

# Count non-empty values in a column
xleak file.xlsx --export json | jq '[.data[] | select(.[12])] | length'

# Extract specific columns as CSV-like output
xleak file.xlsx --export json \
  | jq -r '.data[] | [.[0], .[3], .[7]] | @csv'

Patterns to AVOID (break in zsh):

# BAD: != gets escaped to \!= by zsh
xleak file.xlsx --export json | jq '.data[] | select(.[12] != null and .[12] != "")'
#                                                        ^^ zsh escapes this

# GOOD: use truthy check instead
xleak file.xlsx --export json | jq '.data[] | select(.[12])'

The truthy select(.[N]) pattern filters out both null and "" in one shot, so it's both simpler and shell-safe.

Gotchas

• jq != breaks in zsh — see jq Filtering Patterns above. Use select(.[N]) truthy checks or pipe syntax select(.[N] | . != null) instead of inline !=.
• Large file JSON export is slow — xleak file.xlsx --export json reads the entire file into memory. For files with 10k+ rows, prefer CSV export or use jq to slice early.
• Column truncation in terminal mode — columns wider than 30 chars are truncated by default. Use -w 50 to increase, or --wrap to wrap text.
• No write support — xleak is read-only. To modify Excel files, use other tools.
• ODS limitations — .ods files don't support named tables or formulas display.