Glob Patterns

Glob Patterns

Globs are used throughout Copybara to match files and directories.

Basic Syntax

glob(["pattern1", "pattern2"])

Pattern Characters

Pattern	Matches
*	Any characters except /
**	Any characters including / (recursive)
?	Single character
[abc]	Character class
[!abc]	Negated character class

Examples

Match All Files

glob(["**"])

Match by Extension

# Single extension
glob(["**/*.py"])

# Multiple extensions
glob(["**/*.py", "**/*.pyi"])

# Or with braces
glob(["**/*.{py,pyi}"])

Match Directory

# All files in src/
glob(["src/**"])

# Only immediate children
glob(["src/*"])

# Specific subdirectory
glob(["src/lib/**"])

Match Specific Files

# Exact file
glob(["README.md"])

# Files in any directory
glob(["**/README.md"])

# File pattern
glob(["**/test_*.py"])

Include and Exclude

glob(
    include = ["**"],
    exclude = ["**/internal/**"],
)

Common Patterns

# Everything except tests
glob(
    include = ["**"],
    exclude = ["**/*_test.go", "**/test_*.go"],
)

# Source files only
glob(
    include = ["src/**"],
    exclude = [
        "**/internal/**",
        "**/testdata/**",
        "**/*.test.ts",
    ],
)

# Documentation
glob(
    include = ["docs/**", "*.md"],
    exclude = ["docs/internal/**"],
)

Usage Contexts

origin_files

What to read from origin:

core.workflow(
    origin_files = glob(
        include = ["src/**", "docs/**"],
        exclude = ["**/internal/**"],
    ),
    ...
)

destination_files

What Copybara can modify:

core.workflow(
    destination_files = glob(
        include = ["**"],
        exclude = ["README.md", ".github/**"],
    ),
    ...
)

Transformation Paths

Filter where transforms apply:

core.replace(
    before = "old",
    after = "new",
    paths = glob(["**/*.md"]),
)

core.remove(glob(["**/*.bak"]))

Precedence

Exclude patterns take precedence over include:

glob(
    include = ["src/**"],           # Include all of src/
    exclude = ["src/internal/**"],  # But not internal/
)
# Result: src/* except src/internal/*

Order Independence

Include patterns are combined (union):

# These are equivalent:
glob(["src/**", "docs/**"])
glob(include = ["src/**", "docs/**"])

Exclude patterns are also combined:

glob(
    include = ["**"],
    exclude = ["*.tmp", "*.bak"],  # Excludes both
)

Special Cases

Root Files Only

glob(["*"])  # Only files in root

Specific Depth

glob(["src/*"])      # One level deep
glob(["src/*/*"])    # Two levels deep
glob(["src/**"])     # Any depth

Hidden Files

glob(["**/.*"])     # Hidden files
glob(["**/.git/**"])  # Git directory

Common Exclusions

# Build artifacts
exclude = [
    "**/node_modules/**",
    "**/dist/**",
    "**/build/**",
    "**/__pycache__/**",
    "**/target/**",
]

# IDE files
exclude = [
    "**/.idea/**",
    "**/.vscode/**",
    "**/*.iml",
]

# Internal content
exclude = [
    "**/internal/**",
    "**/*.internal.*",
    "**/INTERNAL_*",
]

# Tests
exclude = [
    "**/*_test.go",
    "**/test_*.py",
    "**/*.test.ts",
    "**/testdata/**",
]

Debugging Globs

Test which files match:

# Use folder destination to see results
java -jar copybara.jar migrate copy.bara.sky workflow \
  --folder-destination /tmp/output

# List the files
find /tmp/output -type f

Tip

When unsure, test with a narrow include pattern first, then expand.

Next Steps

• Authoring configuration
• Transformations