basic operations

copy(conanfile, pattern, src, dst, keep_path=True, excludes=None, ignore_case=True)

Copy the files matching the pattern (fnmatch) at the src folder to a dst folder.

  • conanfile – The current recipe object. Always use self.
  • pattern – (Required) An fnmatch file pattern of the files that should be copied. It must not start with .. relative path or an exception will be raised.
  • src – (Required) Source folder in which those files will be searched. This folder will be stripped from the dst parameter. E.g., lib/Debug/x86.
  • dst – (Required) Destination local folder. It must be different from src value or an exception will be raised.
  • keep_path – (Optional, defaulted to True) Means if you want to keep the relative path when you copy the files from the src folder to the dst one.
  • excludes – (Optional, defaulted to None) A tuple/list of fnmatch patterns or even a single one to be excluded from the copy.
  • ignore_case – (Optional, defaulted to True) If enabled, it will do a case-insensitive pattern matching. will do a case-insensitive pattern matching when True

list of copied files


def package(self):
    copy(self, "*.h", self.source_folder, os.path.join(self.package_folder, "include"))
    copy(self, "*.lib", self.build_folder, os.path.join(self.package_folder, "lib"))


The files that are symlinks to files or symlinks to folders with be treated like any other file, so they will only be copied if the specified pattern matches with the file.

At the destination folder, the symlinks will be created pointing to the exact same file or folder, absolute or relative, being the responsibility of the user to manipulate the symlink to, for example, transform the symlink into a relative path before copying it so it points to the destination folder.

Check here the reference of tools to manage symlinks.

load(conanfile, path, encoding='utf-8')

Utility function to load files in one line. It will manage the open and close of the file, and load binary encodings. Returns the content of the file.

  • conanfile – The current recipe object. Always use self.
  • path – Path to the file to read
  • encoding – (Optional, Defaulted to utf-8): Specifies the input file text encoding.

The contents of the file


from import load

content = load(self, "myfile.txt")

save(conanfile, path, content, append=False, encoding='utf-8')

Utility function to save files in one line. It will manage the open and close of the file and creating directories if necessary.

  • conanfile – The current recipe object. Always use self.
  • path – Path of the file to be created.
  • content – Content (str or bytes) to be write to the file.
  • append – (Optional, Defaulted to False): If True the contents will be appended to the existing one.
  • encoding – (Optional, Defaulted to utf-8): Specifies the output file text encoding.


from import save

save(self, "path/to/otherfile.txt", "contents of the file")

rename(conanfile, src, dst)

Utility functions to rename a file or folder src to dst with retrying. os.rename() frequently raises “Access is denied” exception on Windows. This function renames file or folder using robocopy to avoid the exception on Windows.

  • conanfile – The current recipe object. Always use self.
  • src – Path to be renamed.
  • dst – Path to be renamed to.


from import rename

def source(self):
    rename(self, "lib-sources-abe2h9fe", "sources")  # renaming a folder

replace_in_file(conanfile, file_path, search, replace, strict=True, encoding='utf-8')

Replace a string search in the contents of the file file_path with the string replace.

  • conanfile – The current recipe object. Always use self.
  • file_path – File path of the file to perform the replacing.
  • search – String you want to be replaced.
  • replace – String to replace the searched string.
  • strict – (Optional, Defaulted to True) If True, it raises an error if the searched string is not found, so nothing is actually replaced.
  • encoding – (Optional, Defaulted to utf-8): Specifies the input and output files text encoding.


from import replace_in_file

replace_in_file(self, os.path.join(self.source_folder, "folder", "file.txt"), "foo", "bar")

rm(conanfile, pattern, folder, recursive=False)

Utility functions to remove files matching a pattern in a folder.

  • conanfile – The current recipe object. Always use self.
  • pattern – Pattern that the files to be removed have to match (fnmatch).
  • folder – Folder to search/remove the files.
  • recursive – If recursive is specified it will search in the subfolders.


from import rm

rm(self, "*.tmp", self.build_folder, recursive=True)

mkdir(conanfile, path)

Utility functions to create a directory. The existence of the specified directory is checked, so mkdir() will do nothing if the directory already exists.

  • conanfile – The current recipe object. Always use self.
  • path – Path to the folder to be created.


from import mkdir

mkdir(self, "mydir") # Creates mydir if it does not already exist
mkdir(self, "mydir") # Does nothing

rmdir(conanfile, path)


from import rmdir

rmdir(self, "mydir") # Remove mydir if it exist
rmdir(self, "mydir") # Does nothing

chdir(conanfile, newdir)

This is a context manager that allows to temporary change the current directory in your conanfile

  • conanfile – The current recipe object. Always use self.
  • newdir – Directory path name to change the current directory.


from import chdir

def build(self):
    with chdir(self, "./subdir"):

This function extract different compressed formats (.tar.gz, .tar, .tzb2, .tar.bz2, .tgz, .txz, tar.xz, and .zip) into the given destination folder.

It also accepts gzipped files, with extension .gz (not matching any of the above), and it will unzip them into a file with the same name but without the extension, or to a filename defined by the destination argument.

from import unzip

unzip(self, "")
# or to extract in "myfolder" sub-folder
unzip(self, "", "myfolder")

You can keep the permissions of the files using the keep_permissions=True parameter.

from import unzip

unzip(self, "", "myfolder", keep_permissions=True)

Use the pattern argument if you want to filter specific files and paths to decompress from the archive.

from import unzip

# Extract only files inside relative folder "small"
unzip(self, "", pattern="small/*")
# Extract only txt files
unzip(self, "", pattern="*.txt")
unzip(conanfile, filename, destination='.', keep_permissions=False, pattern=None, strip_root=False)

Extract different compressed formats

  • conanfile – The current recipe object. Always use self.
  • filename – Path to the compressed file.
  • destination – (Optional, Defaulted to .) Destination folder (or file for .gz files)
  • keep_permissions – (Optional, Defaulted to False) Keep the zip permissions. WARNING: Can be dangerous if the zip was not created in a NIX system, the bits could produce undefined permission schema. Use this option only if you are sure that the zip was created correctly.
  • pattern – (Optional, Defaulted to None) Extract only paths matching the pattern. This should be a Unix shell-style wildcard, see fnmatch documentation for more details.
  • strip_root – (Optional, Defaulted to False) If True, and all the unzipped contents are in a single folder it will flat the folder moving all the contents to the parent folder.

This function reads the conandata.yml inside the exported folder in the conan cache, if it exists. If the conandata.yml does not exist, it will create it. Then, it updates the conandata dictionary with the provided data one, which is updated recursively, prioritizing the data values, but keeping other existing ones. Finally the conandata.yml is saved in the same place.

This helper can only be used within the export() method, it can raise otherwise. One application is to capture in the conandata.yml the scm coordinates (like Git remote url and commit), to be able to recover it later in the source() method and have reproducible recipes that can build from sources without actually storing the sources in the recipe.


from conan import ConanFile
from import update_conandata

class Pkg(ConanFile):
    name = "pkg"
    version = "0.1"

    def export(self):
        # This is an example, doesn't make sense to have static data, instead you
        # could put the data directly in a conandata.yml file.
        # This would be useful for storing dynamic data, obtained at export() time from elsewhere
        update_conandata(self, {"mydata": {"value": {"nested1": 123, "nested2": "some-string"}}})

    def source(self):
        data = self.conan_data["sources"]["mydata"]
update_conandata(conanfile, data)

Tool to modify the conandata.yml once it is exported. It can be used, for example:

  • To add additional data like the “commit” and “url” for the scm.
  • To modify the contents cleaning the data that belong to other versions (different from the exported) to avoid changing the recipe revision when the changed data doesn’t belong to the current version.
  • conanfile – The current recipe object. Always use self.
  • data – (Required) A dictionary (can be nested), of values to update

collect_libs(conanfile, folder=None)

Returns a sorted list of library names from the libraries (files with extensions .so, .lib, .a and .dylib) located inside the conanfile.cpp_info.libdirs (by default) or the folder directory relative to the package folder. Useful to collect not inter-dependent libraries or with complex names like libmylib-x86-debug-en.lib.

For UNIX libraries staring with lib, like libmath.a, this tool will collect the library name math.

  • conanfile – The current recipe object. Always use self.
  • folder – (Optional, Defaulted to None): String indicating the subfolder name inside conanfile.package_folder where the library files are.

A list with the library names


This tool collects the libraries searching directly inside the package folder and returns them in no specific order. If libraries are inter-dependent, then package_info() method should order them to achieve correct linking order.


from import collect_libs

def package_info(self):
    self.cpp_info.libdirs = ["lib", "other_libdir"]  # Default value is 'lib'
    self.cpp_info.libs = collect_libs(self)

For UNIX libraries starting with lib, like libmath.a, this tool will collect the library name math. Regarding symlinks, this tool will keep only the “most generic” file among the resolved real file and all symlinks pointing to this real file. For example among files below, this tool will select libmath.dylib file and therefore only append math in the returned list:

-rwxr-xr-x libmath.1.0.0.dylib lrwxr-xr-x libmath.1.dylib -> libmath.1.0.0.dylib
lrwxr-xr-x libmath.dylib -> libmath.1.dylib