app.py

from datetime import datetime, timedelta
from collections import Counter
import streamlit as st
import requests
import plotly.graph_objects as go
from analysis import display_analysis

# GitHub API endpoint for searching repositories
GITHUB_API_URL = "https://api.github.com/search/repositories"

# Function to fetch repositories. Only repos with stars > 50 and updated in the last year are shown
def fetch_uml_repos(query="uml", sort="stars", order="desc", per_page=100, max_pages=10):
    query += " stars:>=" + "50" + " pushed:>=" + (datetime.now() - timedelta(days=365)).strftime("%Y-%m-%d")
    all_repos = []
    for page in range(1, max_pages + 1):
        params = {
            "q": query,
            "sort": sort,
            "order": order,
            "per_page": per_page,
            "page": page
        }
        response = requests.get(GITHUB_API_URL, params=params)
        if response.status_code == 200:
            repos = response.json()["items"]
            if not repos:
                break
            all_repos.extend(repos)
        else:
            st.error(f"Error fetching data from GitHub API: {response.status_code}")
            break
    return all_repos


# Fetch repositories
if 'repos' not in st.session_state:
    st.session_state.repos = fetch_uml_repos()

# List of excluded repositories
excluded_repos = {
    "awesome-low-level-design", "Books-Free-Books", "awesome-diagramming", "plantuml-examples", "plantuml-examples", "hogwarts-artifacts-online", "-Enterprise-Architect-16-Crack-renewal-", "UoM-Applied-Informatics", "UML-Best-Practices",
"design-pattern-examples-in-python", "design-pattern-examples-in-crystal", "FreeTakServer", "plantuml-icon-font-sprites", "snow-owl", "StarUML-CrackedAndTranslate", "tiro-notes", "QuickUMLS"
}

# Filter out excluded repositories
st.session_state.repos = [repo for repo in st.session_state.repos if repo['name'] not in excluded_repos]

repos = st.session_state.repos


# Display the table
st.set_page_config(layout="wide")
st.title("Dashboard of Open-Source UML Tools in GitHub")
st.subheader("Maintained by the [BESSER team](https://github.com/BESSER-PEARL/BESSER)")

# Add table of contents
st.markdown("""
## Table of Contents
- [Quick Notes](#quick-notes)
- [Repository Filters](#repository-filters)
- [Repository Table](#repository-table)
- [Selection Method](#selection-method)
- [Global Statistics](#global-statistics)
- [Tools covering as well other concepts](#repository-analysis)
    - [UML and no-code](#analysis-for-nocode)
    - [UML and low-code](#analysis-for-lowcode)
    - [UML and AI](#analysis-for-ai)
    - [UML tools in the PlantUML ecosystem](#analysis-for-plantuml)

""")

st.markdown("<a name='quick-notes'></a>", unsafe_allow_html=True)
st.write("## Quick notes:")
st.write("- Use the sliders to filter the repositories. Click on a column header to sort the table.")
st.write("- Hover over the table to search for specific reports or export the table as a CSV file.")
st.write("- A few global stats are also available at the bottom of the page.")
st.write("- Suggest improvements via the [GitHub repository of this dashboard](https://github.com/jcabot/oss-uml-tools)")

# Add anchors before each section
st.markdown("<a name='repository-filters'></a>", unsafe_allow_html=True)
st.write("## Repository Filters")

# Add star filter slider
min_stars = st.slider("Minimum Stars", min_value=50, max_value=50000, value=50, step=50)

# Add a date filter slider
# Calculate date range, also storing the value in the session to avoid the slider resetting all the time due to
# streamlit thinking the min max value have changed and need to restart

if 'today' not in st.session_state:
    st.session_state.today = datetime.today()

today = st.session_state.today
one_year_ago = today - timedelta(days=365)

# Date slider
min_date = st.slider(
    "Last Commit",
    min_value=one_year_ago,
    max_value=today,
    value=one_year_ago,
    step=timedelta(days=1)
)



if repos:
    # Create a table with repository information. Only repos with stars >= min_stars and last commit >= min_date are shown
    table_data = []

    filtered_repos = [repo for repo in repos if repo['stargazers_count'] >= min_stars and datetime.strptime(repo['pushed_at'].split('T')[0], '%Y-%m-%d').date() >= min_date.date()]
  
    for repo in filtered_repos:
        table_data.append({
            "Name": repo["name"],
            "Stars⭐": repo['stargazers_count'],
            "Last Updated": repo['pushed_at'].split('T')[0],
            "First Commit": repo['created_at'].split('T')[0],
            "URL": repo['html_url'],
            "Forks": repo['forks'],
            "Issues": repo['open_issues'],
            "Language": repo['language'],
            "License": repo['license']['name'] if repo['license'] else "No license",
            "Description": (repo["description"] or "No description")[:200],
            "Topics": repo['topics']
        })
    
    st.write(f"Showing {len(table_data)} repositories")
    st.dataframe(
        table_data,
        column_config={
            "URL": st.column_config.LinkColumn("URL")
        },
        use_container_width=True,
        height=(len(table_data)+1)*35+3,
        hide_index=True
    )

    st.markdown("<a name='selection-method'></a>", unsafe_allow_html=True)
    st.subheader("Selection method")

    #Write the selection method


    st.write("The selection method is based on the following inclusion criteria:")
    st.write("- Repositories that declare themselves as UML projects")
    st.write("- Repositories with more than 50 stars")
    st.write("- Active repositories (last commit is no more than 1 year ago")
    st.write("- Tool aims to render, edit or generate from UML models")
    st.write("and exclusion criteria:")
    st.write("- Repositories with no information in English")
    st.write("- Repositories that were just created to host the source code of a published article")
    st.write("- Repositories that are awesome lists or collection of resources or examples")

    st.write("The final list is the intersection of the above criteria. The final list has also been manually curated to remove projects that use UML in a different sense of what we mean by UML in software engineering.")
    st.write("For more information about UML tools:")
    st.write("- See this list of [UML tools](https://modeling-languages.com/uml-tools/)")
    st.write("- Check out these [UML books](https://modeling-languages.com/list-uml-books/)")
    st.write("- Play with UML via our open source low-code tool [BESSER](https://github.com/BESSER-PEARL/BESSER) that comes with a web-based UML editor")
    st.write("- And learn about the role of [UML in modern development approaches](https://lowcode-book.com/)")

    st.markdown("<a name='global-statistics'></a>", unsafe_allow_html=True)
    st.subheader("Some global stats")

    # Create a list of first commit dates
    first_commit_dates = [datetime.strptime(repo['created_at'].split('T')[0], '%Y-%m-%d').date() for repo in
                          filtered_repos]

    # Grouping the data by year
    years = [date.year for date in first_commit_dates]
    year_counts = Counter(years)

    # Plotting the distribution of first commit dates by year
    year_bar_chart = go.Figure(
        data=[
            go.Bar(
                x=list(year_counts.keys()),
                y=list(year_counts.values()),
            )
        ]
    )
    year_bar_chart.update_layout(
        title="Distribution of First Commit Dates by Year",
        xaxis_title="Year of First Commit",
        yaxis_title="Number of Repositories",
        xaxis=dict(tickangle=45)
    )

    # Create a list of star counts
    star_counts = [repo['stargazers_count'] for repo in filtered_repos]

    # Plotting the distribution of repositories by star count using a boxplot
    star_box_plot = go.Figure(
        data=[
            go.Box(
                x=star_counts,
                boxpoints="outliers",  # Show only outliers as points
                jitter=0.5,
            )
        ]
    )
    star_box_plot.update_layout(
        title="Distribution of Repositories by Star Count",
        xaxis_title="",
        yaxis_title="Number of Stars",
        xaxis=dict(showticklabels=False)
    )

    # Create a list of languages from filtered_repos
    languages = [repo['language'] for repo in filtered_repos if repo['language']]

    # Count the occurrences of each language
    language_counts = Counter(languages)

    # Plotting the aggregation of repositories by language
    language_bar_chart = go.Figure(
        data=[
            go.Bar(
                x=list(language_counts.keys()),
                y=list(language_counts.values()),
            )
        ]
    )
    language_bar_chart.update_layout(
        title="Aggregation of Repositories by Language",
        xaxis_title="Programming Language",
        yaxis_title="Number of Repositories",
        xaxis=dict(tickangle=45)
    )

    cols = st.columns(2)
    with cols[0]:
        st.plotly_chart(year_bar_chart, use_container_width=True)
        st.plotly_chart(language_bar_chart, use_container_width=True)
    with cols[1]:
        st.plotly_chart(star_box_plot, use_container_width=True)

else:
    st.write("No repositories found or there was an error fetching data.")


if 'repos' in st.session_state and st.session_state.repos:
    st.write("## Repository Analysis")
    
    for keyword in ['nocode','lowcode', 'ai', 'plantuml']:
        st.write(f"### Analysis for '{keyword}'")
        display_analysis(st.session_state.repos, keyword)
        st.markdown("---")
else:
    st.warning("Please fetch repositories first using the search functionality above.")