# pkg-analyzer

**Repository Path**: shiloong/pkg-analyzer

## Basic Information

- **Project Name**: pkg-analyzer
- **Description**: A tool help to analyze the package similarity among given Linux distro, by checking the NVR of package.
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2022-07-17
- **Last Updated**: 2022-07-20

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# The package similarity analyzer
A simple analyzer help to check the similarity among Linux distroes.
Now it only support the rpm packages, by comparing the Name and Version.

## Requirements
- python3 is needed to run this tool;
- `repoquery' command is needed, which can be install by 'dnf install yum-utils'; 
- root permission is needed for distro in ISO image.

## Usage
```
$./pkg_analyzer -h
usage: pkg_analyzer [-h] -a DISTROA -b DISTROB [DISTROB ...]

Package Similarity Analyzer

optional arguments:
  -h, --help            show this help message and exit
  -a DISTROA, --distroA DISTROA
                        distro A source repo, ISO or URL
  -b DISTROB [DISTROB ...], --distroB DISTROB [DISTROB ...]
                        distro B source repo, ISO or URL
```

## Examples
### 1. Check the package similarity between distro by ISO:
`./pkg_analyzer -a Fedora-Server-dvd-x86_64-33-1.2.iso -b Fedora-Server-dvd-x86_64-29-1.2.iso`

The results:
```bash
===============================================================
             PACKAGES : TOTAL   SAME_NAME   SAME_NAME_&_VERSION
----------------------:----------------------------------------
            distro  A :  1770          --                    --
            distro  B :  2980        1550                   504
[SIMILARITY: A vs B ] :    --      87.57%                28.47%
===============================================================
```
### 2. Check the package similarity between distro by remote yum repo:
`./pkg_analyzer -a https://dl.fedoraproject.org/pub/archive/fedora/linux/releases/30/Server/source/tree/ -b https://dl.fedoraproject.org/pub/archive/fedora/linux/releases/29/Server/source/tree/`

The results:
```bash
===============================================================
             PACKAGES : TOTAL   SAME_NAME   SAME_NAME_&_VERSION
----------------------:----------------------------------------
            distro  A :  1690          --                    --
            distro  B :  1666        1602                  1149
[SIMILARITY: A vs B ] :    --      94.79%                67.99%
===============================================================
```
### 3. Check the package similarity among multiple distroes:
`./pkg_analyzer -a https://dl.fedoraproject.org/pub/archive/fedora/linux/releases/29/Everything/source/tree/ -b https://dl.fedoraproject.org/pub/archive/fedora/linux/releases/30/Everything/source/tree/ https://dl.fedoraproject.org/pub/archive/fedora/linux/releases/31/Everything/source/tree/https://dl.fedoraproject.org/pub/archive/fedora/linux/releases/32/Everything/source/tree/ https://dl.fedoraproject.org/pub/archive/fedora/linux/releases/33/Everything/source/tree/`
```bash
===============================================================
             PACKAGES : TOTAL   SAME_NAME   SAME_NAME_&_VERSION
----------------------:----------------------------------------
            distro  A : 21847          --                    --
            distro B0 : 21292       20619                 16279
[SIMILARITY: A vs B0] :    --      94.38%                74.51%
----------------------:----------------------------------------
            distro B1 : 21192          10                     6
            distro B2 : 20934          10                     1
            distro B3 : 20965           6                     0
[SIMILARITY: A vs B*] :    --      94.50%                74.55%
===============================================================
```

## TODO
1. to support multiple repo in url for the repo without "Everything" repo;
2. add more dimension of similarity calculation.