paradizelost/paperless-ngx
1
0
Fork 0
mirror of https://github.com/paperless-ngx/paperless-ngx.git synced 2026-02-09 23:49:29 -06:00
Code Issues Packages Projects Releases Wiki Activity

Replaces tqdm with rich

Browse Source
This commit is contained in:
Trenton H
2026-02-09 15:52:52 -08:00
parent c4ed4e7f36
commit e0b45539a6
11 changed files with 368 additions and 205 deletions
Download Patch File Download Diff File Expand all files Collapse all files

38
src/documents/management/commands/document_fuzzy_match.py
View File

@@ -3,9 +3,13 @@ import multiprocessing
from typing import Final
import rapidfuzz
import tqdm
from django.core.management import BaseCommand
from django.core.management import CommandError
from rich.progress import BarColumn
from rich.progress import Progress
from rich.progress import TaskProgressColumn
from rich.progress import TextColumn
from rich.progress import TimeRemainingColumn
from documents.management.commands.mixins import MultiProcessMixin
from documents.management.commands.mixins import ProgressBarMixin
@@ -106,19 +110,25 @@ class Command(MultiProcessMixin, ProgressBarMixin, BaseCommand):
work_pkgs.append(_WorkPackage(first_doc, second_doc))
# Don't spin up a pool of 1 process
if self.process_count == 1:
results = []
for work in tqdm.tqdm(work_pkgs, disable=self.no_progress_bar):
results.append(_process_and_match(work))
else: # pragma: no cover
with multiprocessing.Pool(processes=self.process_count) as pool:
results = list(
tqdm.tqdm(
pool.imap_unordered(_process_and_match, work_pkgs),
total=len(work_pkgs),
disable=self.no_progress_bar,
),
)
with Progress(
TextColumn("[progress.description]{task.description}"),
BarColumn(),
TaskProgressColumn(),
TimeRemainingColumn(),
disable=self.no_progress_bar,
) as progress:
task = progress.add_task("Fuzzy matching documents", total=len(work_pkgs))
if self.process_count == 1:
results = []
for work in work_pkgs:
results.append(_process_and_match(work))
progress.update(task, advance=1)
else: # pragma: no cover
with multiprocessing.Pool(processes=self.process_count) as pool:
results = []
for result in pool.imap_unordered(_process_and_match, work_pkgs):
results.append(result)
progress.update(task, advance=1)
# Check results
messages = []