Added afunctionality to readout wich page is currently being censored; It probably helps with patience; It is at the moment only implemented in the backend; A pretty frontend is still nessecerry

app/main.py

@@ -1,12 +1,15 @@
 from typing import Annotated
 from typing import List, Dict, Tuple, Sequence
+
+from starlette.responses import StreamingResponse
 from annotated_types import IsDigit
 from fastapi import FastAPI, File, HTTPException, UploadFile, Request, Form
 from fastapi.responses import FileResponse
+
 # import multiprocessing
 # import threading
 # import concurrent.futures
-# import asyncio
+import asyncio
 
 # import fastapi
 from fastapi.staticfiles import StaticFiles
@@ -77,6 +80,9 @@ GREETINGFILE = "./app/graphics/greeting.pdf"
 # for l in cur:
 #     print(l)
 # locpaths = ["./VO_Mathematik_3.pdf"]  # replace this with a database
+censor_status_update_events: Dict[str, asyncio.Event] = {}
+censor_status_datas: Dict[str, Dict[str, int | None | str | bool]] = {}
+# censor_finished_flags: Dict[str, asyncio.Event] = {}
 
 
 def _sql_quarry(
@@ -442,22 +448,62 @@ async def get_submission(
     except ValueError as e:
         error(f"Error creating savepath: {e}")
         raise HTTPException(status_code=400, detail=f"Error creation savepath: {e}")
-    await censor_pdf(
+    # censor_finished_flags[fileId] = asyncio.Event()
-        filepath, dest, rects_p, scales_p, False if censor == "False" else True
+    censor_status_datas[fileId] = {}
+    if fileId not in censor_status_update_events:
+        censor_status_update_events[fileId] = asyncio.Event()
+    await asyncio.to_thread(
+        censor_pdf,
+        filepath,
+        dest,
+        rects_p,
+        scales_p,
+        False if censor == "False" else True,
+        fileId,
     )
+
     # return {"done": "ok"}
     # print(dest)
+    # await censor_finished_flags[fileId].wait()
+    # censor_finished_flags[fileId].clear()
     info(f"Saved file {fileId} as {dest}")
     delete_from_FIP(fileId)
     return FileResponse(dest, content_disposition_type="inline")
 
 
-async def censor_pdf(
+@app.get("/get_censor_status/{file_id}")
+async def get_censor_status(file_id: str):
+    """Yields the currrent page being censored and the total number of pages"""
+    if len(sql("Select filename from FIP where id=?", (file_id,))) < 1:
+        raise HTTPException(
+            400,
+            detail="You are trying to get a status updater for a file that dosent exist.",
+        )
+    if file_id not in censor_status_update_events:
+        censor_status_update_events[file_id] = asyncio.Event()
+    return StreamingResponse(
+        yield_censor_status(file_id), media_type="text/event-stream"
+    )
+
+
+async def yield_censor_status(file_id: str):
+    while True:
+        await censor_status_update_events[file_id].wait()
+        censor_status_update_events[file_id].clear()
+        yield f"event: censorUpdate\ndata: {json.dumps(censor_status_datas[file_id])}\n\n"
+        if censor_status_datas[file_id]["done"]:
+            del censor_status_update_events[file_id]
+            del censor_status_datas[file_id]
+            return
+
+
+def censor_pdf(
     path: str,
     destpath: str,
     rects: List[List[List[float]]],
     scales: List[Dict[str, float]],
     secure: bool,
+    file_id: str,
 ):
     """Censors pdf and runs OCR
     If Secure is True the file is converted to Pixels and then recreated; else the censored sections are just covering the text below and can be easiliy removed with e.g. Inkscape
@@ -499,13 +545,20 @@ async def censor_pdf(
                     fill=(0, 0, 0),
                 )
         if secure:
+            censor_status_datas[file_id]["page"] = i + 1
+            censor_status_datas[file_id]["pages"] = npage
+            censor_status_datas[file_id]["done"] = False
+            censor_status_update_events[file_id].set()
+
             # pages.append(page)
+            # THis Costs us dearly
             bitmap = page.get_pixmap(dpi=400)
             pdf_bytes = bitmap.pdfocr_tobytes(
                 language="deu",
                 tessdata="/usr/share/tessdata/",  # tesseract needs to be installed; this is the path to thetesseract files
             )
             output.insert_pdf(pymupdf.Document(stream=pdf_bytes))
+            # End of the costly part
             # tasks.append(asyncio.create_task(censor_page(page)))
             print(f"Page {i + 1}/{npage}: CENSORING DONE")
         else:
@@ -530,6 +583,10 @@ async def censor_pdf(
     # output.insert_pdf(pymupdf.Document(stream=await task))
     # print("CENSORING DONE")
     output.save(destpath)
+    if secure:
+        censor_status_datas[file_id]["done"] = True
+        censor_status_update_events[file_id].set()
+    # censor_finished_flags[file_id].set()
 
 
 def test_function(i: int) -> bytes:

static/app.js

@@ -281,10 +281,18 @@ function submitPdf(eve) {
 }
 async function submitForm(formData) {
   try {
+    const updateEventSource = new EventSource(
+      "http://127.0.0.1:8000/get_censor_status/" + doc.fID,
+    );
+    console.log("http://127.0.0.1:8000/get_censor_status/" + doc.fID);
+    updateEventSource.addEventListener("censorUpdate", function(eve) {
+      console.log(eve.data);
+    });
     const response = await fetch("http://127.0.0.1:8000/submit", {
       method: "POST",
       body: formData,
     });
+    updateEventSource.close();
     //let responseJSON=await response.json();
     if (response.ok) {
       console.log("Submit OK");