app.py

from pathlib import Path

HERE = Path(__file__).parent

import sys
sys.path.append(HERE)

from utils import GParse_Paper, Get_Bibliography
from bs4 import BeautifulSoup
import solara
from solara.components.file_drop import FileInfo

app_style = (HERE / "style.css").read_text()


def Get_HTMLTop(title):
    # Top part of HTML
    html_top = f"""
    <h1>{title}</h1>
    <span typeof="schema:Person" resource="http://orcid.org/0000-0003-1279-3709">
    
    </span>
    """
    return html_top

def Get_Controls():
    controls="""
        <label for="textSize">Text Size: </label>
        <select id="textSize" name="textSize" onchange="adjustTextSize(this.value)">
            <option value="10">10px</option>
            <option value="12">12px</option>
            <option value="14">14px</option>
            <option value="16" selected>16px</option>
            <option value="18">18px</option>
            <option value="20">20px</option>
            <option value="24">24px</option>
            <option value="28">28px</option>
            <option value="32">32px</option>
            <option value="36">36px</option>
            <option value="40">40px</option>
            <option value="44">44px</option>
            <option value="48">48px</option>
            <option value="50">50px</option>
        </select>
        <script>
            function adjustTextSize(size) {
                const baseSize = parseInt(size);
                document.body.style.fontSize = baseSize + 'px';
            }
        </script>
    
        <script>
          function openDialog(event, dialogId) {
                var dialog = document.getElementById(dialogId);
                var rect = event.target.getBoundingClientRect();
                dialog.style.top = rect.top + window.scrollY + 'px';
                dialog.style.left = rect.left + window.scrollX + 'px';
                dialog.style.display = 'block';
    
                // Add an event listener to close the dialog when clicking outside of it
                document.addEventListener('click', function(event) {
                    var isClickInside = dialog.contains(event.target);
                    var isClickOnText = event.target.classList.contains('text-area');
    
                    if (!isClickInside && !isClickOnText) {
                        closeDialog(dialogId);
                    }
                }, { once: true });
            }
    
          function closeDialog(dialogId) {
              document.getElementById(dialogId).style.display = 'none';
          }
        </script>
        """
    return Controls


def Get_Sections(soup):
    # Generate sections from divs
    sections_content = ""
    sections_list = []
    raw_text=""
    
    bib = Get_Bibliography(soup)
    citation_modals = []
    
    for div in soup.find_all("div"):
        header = div.find("head")
        if header is not None:
            section_number = header.get('n', "")
            section_id = header.text.replace(" ", "_")
            sections_list.append({'num': normalize_section(section_number), 'text': section_id})
            sections_content += f"<section id='{section_id}'>"
            sections_content += f"<h2>{section_number} {header.text}</h2>"
        else:
            sections_content += f"<section id=''>"
    
        for i,paragraph in enumerate(div.find_all("p")):
            new_paragraph = ""
            for ii,element in enumerate(paragraph.contents):
                if isinstance(element, NavigableString):
                    new_paragraph += element
                elif isinstance(element, Tag) and element.name == "ref" and element.get("target")!=None:
                    ref_id = element.get("target").lstrip("#")
                    if ref_id in bib.keys():
                      citation = f"""<span class="text-area" onclick="openDialog(event, '{ref_id}')">{element.text}</span>"""
                      new_paragraph += citation
                      cit_info = bib[ref_id]
                      citation_modals.append(f"""<div id="{ref_id}" class="dialog">
                              <b>{element.text}</b><br>
                              <b>Title:</b> {cit_info['title']}<br>
                              <b>Authors:</b> {", ".join(cit_info['authors'])}<br>
                              <b>Year:</b> {cit_info['year']}<br>
                              <b>Journal:</b> {cit_info['journal']}<br>
                              <b>DOI:</b> <a href="https://doi.org/{cit_info['doi']}">{cit_info['doi']} </a><br>
                              <button class="close-button" onclick="closeDialog('{ref_id}')">Close</button>
                          </div>""")
                else:
                    new_paragraph += element.text
    
            sections_content += f"<p>{new_paragraph}</p>"
            raw_text += "\n" + paragraph.text
    
        sections_content += "</section>"
    return sections_list, sections_content

def Get_Navigation(controls,):
    # Generate navigation for sections
    navigation = "<div class='sticky-content' style='max-height: 100%; overflow-y: auto;'>" + controls + " <h2> Navigation </h2>"
    for section in sections_list:
        no_tabs = section['num'].count(".")
        if no_tabs>0:
          left = str(20*no_tabs)+"px"  # Adjust the multiplier for desired tab width
        else:
          left= "0px"
        navigation += f'<p style="margin-left: {left}; font-size: 10px;"><a href="#{section["text"]}">{section["num"]} {section["text"]}</a></p>'
    
    navigation += "</div>"



def Get_Article_HTML(pdf):

    article = GParse_Paper(pdf)
    
    soup = BeautifulSoup(article, "xml")

    html = Get_Article_HTML(soup)

    try:
        document_title = soup.find("fileDesc").find("title").text
    except:
        document_title = ""
        
    html_top = Get_HTMLTop(document_title)
    sections_list, sections_content = Get_Sections(soup)
    controls = Get_Controls()
    navigation = Get_Navigation(controls, sections_list)
    
    # Combine all parts into final HTML
    html = style + "<body><article>" + html_top + sections_content + navigation + "n".join(citation_modals)+"</body></article>"
    
    with open("article_demo.html","w") as f:
      f.write(html)
    
    return html


html = solara.reactive("<h1> Article PDF to HTML converter </h1>")

@solara.component
def Page():
    solara.Style(app_style)

    def on_file(f: FileInfo):
        html.value = Get_Article_HTML(f["file_obj"].read()) 
        
    solara.FileDrop(label="Drag and drop article pdf", on_file=on_file, lazy=True)

    solara.HTML(unsafe_innerHTML=html.value)