icpc-style Contest pReparation Insights tool For anyone (X).
Project description
CRIFX
CRIFX is the Contest pReparation Insights tool For anyone (X).
Its purpose is to automatically produce pdf reports that give insights into tasks that are completed and tasks that are still outstanding in preparing an icpc-style competitive programming contest.
Installation
Install crifx with pip install crifx
. Building the report will require some
TeX packages. For example, on Ubuntu 22.04 installing the texlive-latex-extra
apt package with apt-get install texlive-latex-extra
should be sufficient.
If you are using a different operating system, and you are able to run crifx, then please share your experience by opening an Issue and I would be pleased to update installation instructions accordingly.
Using crifx
In the root directory for a problemset, run crifx
from within a problemset nested
directory, or run crifx /path/to/problemset/root
from anywhere to generate a report
for the problemset at the specified path.
If there is no crifx.toml
configuration file in the problemset root directory,
then a report will be created using default configuration values.
Crifx can be configured by adding a crifx.toml
file to the root of the problemset
directory. The configuration can be used to define requirements on things like
the number of indepenedent AC submissions for each problem, groups of programming
languages and the number of submissions needed from each group or the number of
distinct groups required. The configuration file can also be used to specify the
names of people involved in contest preparation such that filenames and special
strings in judge submissions can be used to associate submissions with those
people.
Counting independent submissions
How crifx decides the author of a submission
In order of precedence, crifx determines the author of a file by:
- The presence of a crifx!(author="name") string in the text of a submission file.
- The presence of an underscore separated name string that matches an alias in the
crifx.toml
configuration file. - The git user associated with the largest number of lines in the git blame. Roughly speaking this will be the git user that created/changed the largest number of lines in the current revision.
The crifx.toml
file is used to define the aliases of a submission author and associate them with
a git username. More details to come, including an example.
If you encounter a scenario where crifx is incorrectly attributing a submission to the wrong person
and it cannot easily be fixed by adding a user and alias to the crifx.toml
file.
Crifx does not yet support disambiguating git users with the same git name.
Configuration
Top-level
The top-level configuration of crifx
is defined in a TOML text file named
crifx.toml
placed in the directory containing the directories for each problem.
github_repo_url
. Optional. String. Default:null
. The GitHub repository url for the problem set. This should be the root url for the repository and not the url to some directory in the repository tree.
[review_requirements]
independent_ac
. Optional. Integer. Default:3
.language_groups
. Optional. Integer. Default:2
.submissions_wa
. Optional. Integer. Default:1
.submissions_tle
. Optional. Integer. Default:1
.statement_reviewers
. Optional. Integer. Default:3
.validator_reviewers
. Optional. Integer. Default:2
.data_reviewers
. Optional. Integer. Default:2
.
[[judge]]
The judge
array of tables is used to associate judge names and aliases. The
judge name can also optionally be associated with a git name.
primary_name
. Required. String. A name to identify the judge with. This name will be used in thecrifx
report when listing who wrote each submission and who is ineligible to provide further contributions for a problem.git_name
. Optional. String. This is the git name for the judge if they have one. Note that this is not the GitHub username. From the command line, you can see your git name with the commandgit config user.name
. By specifying agit_name
, submissions from a judge that are uploaded or edited by someone else can be associated with submissions that the judge themself uploads. For the submission uploaded by someone else, the submission will need to be identified by file name or acrifx!(author=name)
string in the submission file.aliases
. Optional. Array of Strings. List of aliases that can be used to identify the judge. Submission file names are split into parts by underscores. If any part matches an alias of a judge, then the submission is assumed to belong to the judge with the alias and associated primary name/git name. Aliases can also be used incrifx!(author=name)
strings inside submission files to identify a judge.
[[language_group]]
name
. Required. String. A name to use for the language group. E.g.,"C/C++"
languages
. Required. Array of Strings. A list of languages to include in the group. The languages must be known tocrifx
.required_ac_count
. Optional. Integer. Default:0
. The number of Accepted submissions that are required for this language group. E.g., a value of2
means that each problem must have at least2
accepted submissions from languages in this language group.
Per problem configuration
For each problem, a TOML text file placed in the root directory for the problem
(i.e., at the same level as the problem.yaml
file and the submissions
directory),
called crifx-problem-status.toml
. This file can be used to track who has reviewed
different parts of a problem. It can also be used to define some per-problem configuration.
github_issue_id
. Optional. Integer. Default:null
. The GitHub Issue id number for a GitHub Issue that may have been created to discuss the problem. This is used to generate links to the GitHub Issue url from within thecrifx
report.
[review_status]
statement_reviewed_by
. Optional. Array of Strings. A list of judge aliases for judges that have reviewed the problem statement and are happy with the state of the problem statement for use in a contest. The judge's name should only be added to this list if they think that the statement is clear and unambiguous.validators_reviewed_by
. Optional. Array of Strings. A list of judge aliases for judges that have reviewed the input (and output) validators. The judge's name should only be added to this list if they have verified that the input validator checks all guarantees made in the problem statement about the input data.data_reviewed_by
. Optional. Array of Strings. A list of judge aliases for judges that have reviewed the test data (secret and sample). The judge's name should only be added to this list if they have inspected the generated test data and after doing so they cannot think of another test case that should be added to ensure that correct submissions will be judged as correct and incorrect submissions will be judged as incorrect.
Example
Below is an example crifx.toml
file.
# Crifx configuration file example.
# The github repository url for this problemset. This can be used to include
# hyperlinks to GitHub Issues for each problem in the report.
github_repo_url = "https://github.com/FinnLidbetter/crifx"
# Section for defining minimum numbers of submissions and manual reviews.
[review_requirements]
# There must be AC submissions from at least 3 distinct judges.
independent_ac = 3
# There must be at least 2 distinct language groups with at least one AC
# submission written in a language in the group.
language_groups = 2
# There must be at least one wrong_answer submission.
submissions_wa = 1
# There must be at least one time_limit_exceeded submission.
submissions_tle = 1
# There must be at least 3 judges who have reviewed the problem statement.
statement_reviewers = 3
# There must be at least 2 judges who have reviewed the input validators.
validator_reviewers = 2
# There must be at least 2 judges who have reviewed the test data.
data_reviewers = 2
# Define judge names in `[[judge]]` tables.
[[judge]]
# Primary name is required.
primary_name = "Jane Doe"
# git_name is not required.
# git_name = "janedoe"
# Aliases is a list of zero or more alternative names.
aliases = ["jd", "jane"]
[[judge]]
primary_name = "Homer Simpson"
git_name = "homers123"
aliases = []
# Tables for defining language groups.
[[language_group]]
# Non alphanumeric characters can be used in the language group name.
name = "c/c++"
# The names of the languages in the group. Language names are cast to
# lowercase before comparison to language names known to crifx.
languages = ["C", "C++"]
# The number of AC submissions required for this group.
required_ac_count = 0
[[language_group]]
name = "java/kotlin"
languages = ["Java", "Kotlin"]
required_ac_count = 0
[[language_group]]
name = "python"
languages = ["Python"]
required_ac_count = 0
Here is an example crifx-problem-status.toml
file inside the
examples/example_problemset/helloworld/
directory.
# GitHub Issue id number associated with this problem.
github_issue_id = 1
[review_status]
# List of names of judges who have reviewed the problem statement.
statement_reviewed_by = [
"Alice",
"Bob",
"Jane Doe",
]
# List of names of judges who have reviewed the input validator(s).
validators_reviewed_by = [
"Homer Simpson",
]
# List of names of judges who have reviewed the test data.
data_reviewed_by = [
"Homer Simpson",
"Bob",
]
This example report screenshot is produced from the example in examples/example_problemset
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file crifx-0.1.3.tar.gz
.
File metadata
- Download URL: crifx-0.1.3.tar.gz
- Upload date:
- Size: 654.1 kB
- Tags: Source
- Uploaded using Trusted Publishing? Yes
- Uploaded via: twine/5.1.1 CPython/3.12.7
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | dfe3d562af51eb47266040936d0213fb911c271b77aa6bcee723c4266326bbc6 |
|
MD5 | 351165ce6e7e47b7690be4fd0df23ad0 |
|
BLAKE2b-256 | 76b4dca66d7c6329095ab2909b121850fbaccbbb34cd409c82a0486957a14eb4 |
File details
Details for the file crifx-0.1.3-py3-none-any.whl
.
File metadata
- Download URL: crifx-0.1.3-py3-none-any.whl
- Upload date:
- Size: 26.4 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? Yes
- Uploaded via: twine/5.1.1 CPython/3.12.7
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 07acec12f1caa890f2c4f6af064adb25c4de2a490fb5d621346a2c566d71eb0b |
|
MD5 | 7e8e42997a36d086f94923dea133918e |
|
BLAKE2b-256 | aa738b55833038f0f7b145dbde5477b7ddd391e19c5f490f4d0f4f8273ece31c |