So my folder structure is like this
.
├── eulerlibs
│ ├── EulerLibs
│ └── eulerlibs.cabal
├── flake.lock
├── flake.nix
├── p001
│ ├── Main.hs
│ └── p001.cabal
├── p002
│ ├── Main.hs
│ └── p002.cabal
├── p003
│ ├── Main.hs
│ └── p003.cabal
├── p004
│ ├── Main.hs
│ └── p004.cabal
├── p005
│ ├── Main.hs
│ └── p005.cabal
What I am trying to do is import the library modules in the ./eulerlibs (library) cabal directory to lets say ./p005(executable) cabal directory
I do understand that I can include built libraries (*.so), but can find no way of linking two cabal projects
I'm very new to cabal as well as haskell so some insight will be much appreciated. If nothing works out I'll have to add the library to p005 itself which I really want to avoid.
You can do this with a cabal.project file. Place it at the top level (in the . directory) with these contents:
packages: */*.cabal
Then you will be able to use all those libraries in each other's cabal files in the build-dependencies, e.g. in p001.cabal:
...
library
...
build-depends: base, eulerlibs
...
...
Here is the full documentation of cabal.project files: https://cabal.readthedocs.io/en/latest/cabal-project.html
Related
I have a package structure like this:
├── LICENSE
├── README.md
├── main
│ ├── __init__.py
│ ├── application.py
│ ├── core_function
│ │ ├── __init__.py
│ │ └── maps
│ │ ├── Taiwan
│ │ ├── Taiwan_detailed
│ │ └── taiwan.txt
└── setup.py
I try to wrap this package by python setup.py develop.
When it runs Taiwan = gpd.read_file(pkg_resources.resource_stream(__name__, 'maps/Taiwan/COUNTY_MOI_1090820.shp')) in application.py,
fiona.errors.DriverError: '/vsimem/9b633f8a8a3f457eadf710539afd2a22' not recognized as a supported file format. or
fiona._err.CPLE_OpenFailedError: '/vsimem/9b633f8a8a3f457eadf710539afd2a22' not recognized as a supported file format. would occur.
It reads perfectly when I run it as a script on my machine, but it fails as a package.
Knowing that the shp file should be read along with all the files in that folder, in my setup.py I also include them
packages= setuptools.find_packages(),
package_data={'maps': ['main/core_function/maps/*','core_function/maps/Taiwan/*']},
I was thinking the problem is about the path, but taiwan.txt can be read.
Any suggestion is appreciated. Thanks in advance.
So far I have not found the reason.
But I instead used the to_file method and work with only one file.
.to_file("package.gpkg", driver="GPKG")
This works in my package. The problem can be due to reading multiple files.
I'm using cargo-bundle to create an .app bundle for my Rust application built with fltk-rs.
There are assets in the application like images. While I'm developing, accessing these assets is no problem.
main.rs:
let mut my_img = SharedImage::load("imgs/smiley.png").unwrap();
.
├── Cargo.lock
├── Cargo.toml
├── app_icon_design.psd
├── imgs
│ └── smiley.PNG <--get this image
├── src
│ ├── app_icon.png
│ ├── app_icon#2x.png
│ ├── icon32x32.png
│ └── main.rs
└── target
├── CACHEDIR.TAG
├── debug
└── release
I then bundle to an .app with cargo build --release which gives me the following directory structure inside of my .app:
.
├── Info.plist
├── MacOS
│ └── build_to_app_bundle <--my executable
└── Resources
└── imgs
└── smiley.PNG<--get this image now
Now my application needs to get the image file from the /Resources folder:
println!("my resources folder is: {:?}", std::env::current_exe().unwrap().parent().unwrap().join(Path::new("Resources")));
let my_resources_path = std::env::current_exe().unwrap().parent().unwrap().join(Path::new("Resources"));
How can I make it so that every time I want to load an image from a path I do not need to explicitly reference the Resources value?
My abridged sonar-project.properties files is as follows:
# Sources
sonar.sources=felix
sonar.sources.inclusions=**/**.py
sonar.exclusions=**/test_*.py,**/**.pyc,felix/utils/*,**/*.iml
# Linter
sonar.python.pylint=/usr/local/bin/pylint
sonar.python.pylint_config=.pylintrc
sonar.python.pylint.reportPath=pylint-report.txt
# Coverage / Unit Tests
sonar.tests=./tests
sonar.test.inclusions=**/test_**.py
sonar.python.xunit.skipDetails=false
#DEFAULT VALUES: sonar.python.xunit.reportPath=xunit-reports/xunit-result-*.xml
#DEFAULT VALUES: sonar.python.coverage.reportPath=coverage-reports/*coverage-*.xml
The abridged source code tree is like so:
├── felix
│ ├── __init__.py
│ ├── __pycache__
│ │ ├── __init__.cpython-35.pyc
│ │ ├── process.cpython-35.pyc
│ │ └── spark.cpython-35.pyc
│ ├── felix.iml
│ ├── process.py
│ ├── spark.py
│ └── utils
│ └── utils.py
├── requirements.txt
├── setup.py
├── sonar-project.properties
├── tests
│ ├── __init__.py
│ ├── __pycache__
│ │ ├── __init__.cpython-35.pyc
│ │ └── test_process.cpython-35-PYTEST.pyc
│ ├── cia-spark.iml
│ ├── data
│ └── test_process.py
└── tox.ini
I'm getting the following warning, though, when I run the sonar-scanner: WARN: The resource for '' is not found, drilling down to the details of this test won't be possible
Could someone, please, let me know why I'm getting this warning and how can I get rid of / fix it? Thanks.
I received this warning because I was excluding test files from analysis. I see than in your properties you are excluding your tests as well:
sonar.exclusions=**/test_*.py,**/**.pyc,felix/utils/*,**/*.iml
This will prevent sonar from calculating the number of tests and their pass/fail status as shown in the open source here
The issue turned out to be with the Pylint integrated into the Pytest call.
The parent Pytest call generated a unit testing report, which had empty classnames for additional "empty" tests that Pylint came up with.
SonarQube warned about those empty classnames.
I ended up removing Pylint Pytest integration and just running Pylint as a separate step from Pytest.
The directory layout of a Rust project should look like this (source)
.
├── Cargo.lock
├── Cargo.toml
├── benches
│ └── large-input.rs
├── examples
│ └── simple.rs
├── src
│ ├── bin
│ │ └── another_executable.rs
│ ├── lib.rs
│ └── main.rs
└── tests
└── some-integration-tests.rs
What is the file simple.rs under examples? How do I execute it? How should the file look like?
Examples are useful in library crates to show how the crate is used.
An example can be an executable with a main method or a library; it can either be in a single file examples/example-name.rs or consist of several files in a subdirectory examples/example-name/, with the main method in main.rs. To compile a library example you need to specify its crate type in Cargo.toml:
[[example]]
name = "example-name"
crate-type = ["lib"]
Examples are compiled by cargo test to ensure that they are up to date with the crate. You can run a specific executable example by
cargo run --example <example-name>
and selectively build any example with
cargo build --example <example-name>
This is documented in the Cargo Reference.
I'm writing a book about embedded Rust using mdbook as one git repository and then I have another repository created by cargo where I place the code.
I'd like to structure the code so it corresponds with the chapters in the book and therefore is in separate directory.
The structure for the book:
├── book
├── book.toml
└── src
├── chapter_1.md
├── chapter_2.md
├── chapter_3.md
├── chapter_4.md
├── chapter_5.md
├── chapter_6.md
└── SUMMARY.md
And the structure for the code:
├── aarch64-unknown-none.json
├── Cargo.lock
├── Cargo.toml
├── layout.ld
├── Readme.md
├── chapter1
│ └── main.rs
├── chapter2
│ ├── boot.rs
│ └── main.rs
└── chapter3
├── boot.rs
├── console.rs
└── main.rs
I'd prefer this structure as the reader can then look directly at the code for the chapter and not search git commits. I also sometimes need to modify something later therefore git commits are not a solution.
Is there a way to specify this format in Cargo.toml? To either build all the directories or specify which one on the command line.
The exact solution can be found in the second edition of the Rust book with an example.
I restructured the repository like this:
├── aarch64-unknown-none.json
├── Cargo.lock
├── Cargo.toml
├── layout.ld
├── Readme.md
├── chapter1
│ ├── Cargo.toml
│ └── main.rs
├── chapter2
│ ├── boot.rs
│ ├── Cargo.toml
│ └── main.rs
└── chapter3
├── boot.rs
├── Cargo.toml
├── console.rs
└── main.rs
Cargo.toml files in the chapter directories remain without any modification. Only the Cargo.toml in the root is modified to contain the following:
[workspace]
members = ["chapter1", "chapter2", "chapter3"]
One small drawback to this solution is that the members must have different crate names in their Cargo.toml as the output of all members is stored in the target dir in the root of the workspace. This is only a small issue and I appreciate the flexibility Cargo offers.