main.py

from fastapi import FastAPI, Request,HTTPException
from fastapi.templating import Jinja2Templates
from fastapi.staticfiles import StaticFiles
from pydantic import BaseModel, Json
from typing import Optional, List, Tuple
import json 
import spacy
from spacy import displacy
from spacy.displacy import EntityRenderer,parse_ents
from spacy.lang.en import English
import uuid
import random
import srsly
from pathlib import Path
import re
from markupsafe import Markup, escape

app = FastAPI()
app.mount("/assets", StaticFiles(directory="assets"), name="assets")

templates = Jinja2Templates(directory="templates")

#Register custom filter to convert \n to <br>
def nl2br(value):
    br = "<br>\n"
    result = "\n\n".join(
        f"<p>{br.join(p.splitlines())}</p>"
        for p in re.split(r"(?:\r\n|\r(?!\n)|\n){2,}", value)
    )
    return result
templates.env.filters['nl2br'] = nl2br

# def textrank_me(item:dict):
#     text = item['name']
#     nlp = spacy.load("en_core_web_sm")
#     nlp.add_pipe("textrank")
#     doc = nlp(text)
#     # TODO add item annotations to doc.ents

#     # examine the top-ranked phrases in the document
#     for phrase in doc._.phrases:
#         print(phrase.text)
#         print(phrase.rank, phrase.count)
#         print(phrase.chunks)



class Interview(BaseModel):
    name: str
    date: Optional[str] = ''
    location: Optional[str] = ''
    interviewer: Optional[str] = ''
    portrait: Optional[str] = ''
    thumbnail: Optional[str] = ''
    text: Optional[str] = ''
    annotations: List[dict] #for both audio, text and images
    audio: Optional[List[dict]] = None
    subjects: Optional[List[str]] = None

def load_data() -> Tuple[List[Interview], List[str]]:
    interviews = []
    subjects = []
    data_dir = Path.cwd() / 'data'  
    for item in data_dir.iterdir():
        data = srsly.read_json(item)
        if data:
            i = Interview(**data)
            i.subjects = list(set([a['label'] for a in i.annotations if a['label'] !='SENT']))
            interviews.append(i)
            subjects.extend(i.subjects)
        else:
            print('error',data)
    return interviews, list(set(subjects))


brick = [
        "Ivan_6-3-2019_Portrait.jpg",
        "Jeimmy_6-17-18_Portrait.jpg", 
        "Jesus_1_6-5-2019_Portrait.jpg", 
        "Jesus_3_6-5-2019_Portrait.jpg",
        "Joana_6-8-2019_Portrait.jpg"
        ]
photos = brick


def add_spans_to_text(interview:Interview):
    renderer = EntityRenderer()
    renderer.colors = {"PERSON": "transparent", "SENT":"transparent" }
    renderer.ent_template = """<span id="{id}" onclick="" style="background: {bg}" class="{class}" value="{label}">{text}</span>""" #data-url="{data_url}"
    parsed = dict(text=interview.text,ents=interview.annotations, settings={'lang': 'en', 'direction': 'ltr'})
    return renderer.render([parsed], page=False, minify=False).strip()

def add_audio_to_annotations(interview:Interview):
    audio = [a for a in interview.audio if a.get('start',None) and a.get('end',None)]
    interview.annotations.extend(audio)
    return interview

@app.get("/")
def index(request:Request):
    #choose three random images from the photographs with a green background
    oral_histories, photographs, teaching = random.sample(photos, 3)
    interviews, subjects = load_data()
    context = dict(
        request=request,
        interviews=interviews,
        filters= subjects,
        oral_histories= oral_histories, 
        photographs=photographs,
        subjects=subjects, 
        teaching=teaching)
    return templates.TemplateResponse("index.html", context)

@app.get("/interview/{person}.html")
def interview(request:Request,person:str):
    context = {}
    context['request'] =request
    interviews, subjects = load_data()
    person = [i for i in interviews if i.name.lower() == str(person).lower()]
    context['person'] = person[0]
    context['person'] = add_audio_to_annotations(context['person'])
    #mark_sentence_spans_with_ids(person[0])
    context['text'] = person[0].text #add_spans_to_text(person[0])
    
    return templates.TemplateResponse("interview.html", context)

@app.get("/interview_json/{person}")
def interview_json(person:str):
    interviews, subjects = load_data()
    person = [i for i in interviews if i.name.lower() == str(person).lower()]
    if person:
        return person[0]
    else: 
        raise HTTPException(status_code=404, detail="Interview not found")

@app.get("/search.html")
def search(request:Request, q:str=None):
    return templates.TemplateResponse("search.html", {'request': request})


@app.get("/add_interview")
def add_interview(request:Request):
    return templates.TemplateResponse("add_interview.html", {'request': request})