pybaseutils 2.2.20__tar.gz → 2.2.22__tar.gz

{pybaseutils-2.2.20 → pybaseutils-2.2.22}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: pybaseutils
-Version: 2.2.20
+Version: 2.2.22
 Summary: pybaseutils
 Home-page: https://github.com/PanJinquan/base-utils
 Author: PanJinquan

{pybaseutils-2.2.20 → pybaseutils-2.2.22}/pybaseutils/__init__.py

@@ -8,4 +8,4 @@
 # --------------------------------------------------------
 """
 
-__version__ = '2.2.20'
+__version__ = '2.2.22'

{pybaseutils-2.2.20 → pybaseutils-2.2.22}/pybaseutils/converter/build_coco.py

@@ -120,7 +120,7 @@ class COCOBuilder():
 
     def save_coco(self, json_file):
         file_utils.create_file_path(json_file)
-        file_utils.write_json_path(json_file, self.coco)
+        file_utils.save_json(json_file, self.coco)
         print("save file:{}".format(json_file))
 
     def set_keypoints_category(self, kps_name=[], skeleton=[], cat_id=0):

{pybaseutils-2.2.20 → pybaseutils-2.2.22}/pybaseutils/converter/build_labelme.py

@@ -48,7 +48,7 @@ def maker_labelme(json_file, points, labels, image_name, image_size, image_bs64=
         "imageWidth": image_size[0]
     }
     if os.path.exists(image_name): file_utils.copy_file_to_dir(image_name, os.path.dirname(json_file))
-    file_utils.write_json_path(json_file, data)
+    file_utils.save_json(json_file, data)
     return data
 
 
@@ -60,9 +60,9 @@ def del_labelme_imagedata(anno_dir):
     """
     file_list = file_utils.get_files_lists(anno_dir, postfix=["*.json"])
     for anno_file in tqdm(file_list):
-        data_info = json_utils.read_json_data(anno_file)
+        data_info = json_utils.load_json(anno_file)
         data_info["imageData"] = None
-        json_utils.write_json_path(anno_file, data_info)
+        json_utils.save_json(anno_file, data_info)
 
 
 def copy_labelme_files(image_dir, anno_dir, out_root):
@@ -77,7 +77,7 @@ def copy_labelme_files(image_dir, anno_dir, out_root):
     out_images = file_utils.create_dir(out_root, "images")
     out_json = file_utils.create_dir(out_root, "json")
     for json_file in tqdm(json_list):
-        json_data = json_utils.read_json_data(json_file)
+        json_data = json_utils.load_json(json_file)
         image_name = json_data['imagePath']
         shapes = json_data.get('shapes', [])
         image_file = os.path.join(image_dir, image_name)

{pybaseutils-2.2.20 → pybaseutils-2.2.22}/pybaseutils/converter/concat_coco.py

@@ -103,14 +103,14 @@ class ConcatCoco(object):
         """
         if isinstance(file_dict, dict):
             for dirname, file in file_dict.items():
-                coco = file_utils.read_json_data(file)
+                coco = file_utils.load_json(file)
                 self.add_categories(copy.deepcopy(coco["categories"]))
                 self.add_images(copy.deepcopy(coco["images"]), dirname)
                 self.add_annotations(copy.deepcopy(coco["annotations"]),
                                      copy.deepcopy(coco["categories"]))
         elif isinstance(file_dict, list):
             for file in file_dict:
-                coco = file_utils.read_json_data(file)
+                coco = file_utils.load_json(file)
                 self.add_categories(copy.deepcopy(coco["categories"]))
                 self.add_images(copy.deepcopy(coco["images"]), dirname=None)
                 self.add_annotations(copy.deepcopy(coco["annotations"]),
@@ -120,7 +120,7 @@ class ConcatCoco(object):
 
     def save_coco(self, json_file):
         file_utils.create_file_path(json_file)
-        file_utils.write_json_path(json_file, self.coco)
+        file_utils.save_json(json_file, self.coco)
         print("save file:{}".format(json_file))
 
 

{pybaseutils-2.2.20 → pybaseutils-2.2.22}/pybaseutils/cvutils/video_utils.py

@@ -238,6 +238,7 @@ def video_capture(video_file: int or str, save_video: str or int = None, interva
                  title: 控制显示窗口名
     :return:
     """
+    assert os.path.exists(video_file), f"video_file={video_file}"
     video_cap = image_utils.get_video_capture(video_file)
     w, h, num_frames, fps = image_utils.get_video_info(video_cap)
     start = int(kwargs.get("start", 0) * fps)
@@ -291,6 +292,7 @@ def video_iterator(video_file: int or str, save_video: str or int = None, interv
                  speed: 播放速度
     :return: frame, count, w, h, fps =data_info['frame'],data_info['count'],data_info['w'],data_info['h'],data_info['fps']
     """
+    assert os.path.exists(video_file), f"video_file={video_file}"
     video_cap = image_utils.get_video_capture(video_file)
     w, h, num_frames, fps = image_utils.get_video_info(video_cap)
     start = int(kwargs.get("start", 0) * fps)

{pybaseutils-2.2.20 → pybaseutils-2.2.22}/pybaseutils/dataloader/parser_image_text.py

@@ -81,10 +81,10 @@ class TextDataset(Dataset):
             class_lack = sorted(class_lack.items(), key=lambda x: x[1], reverse=True)
             class_lack = {n[0]: n[1] for n in class_lack}
             class_lack.update({"mean": m})
-            file_utils.write_json_path(os.path.join(save_info, f"{self.phase}_class_dict.json"), self.class_dict)
-            file_utils.write_json_path(os.path.join(save_info, f"{self.phase}_class_count.json"), self.class_count)
+            file_utils.save_json(os.path.join(save_info, f"{self.phase}_class_dict.json"), self.class_dict)
+            file_utils.save_json(os.path.join(save_info, f"{self.phase}_class_count.json"), self.class_count)
             file_utils.write_list_data(os.path.join(save_info, f"{self.phase}_class_name.txt"), self.class_name)
-            file_utils.write_json_path(os.path.join(save_info, f"{self.phase}_class_lack.json"), class_lack)
+            file_utils.save_json(os.path.join(save_info, f"{self.phase}_class_lack.json"), class_lack)
             print("loss_labels: {}".format(class_lack))
         print("------------------------------------------------------------------")
 

{pybaseutils-2.2.20 → pybaseutils-2.2.22}/pybaseutils/file_utils.py

@@ -22,7 +22,7 @@ import argparse
 import itertools
 from datetime import datetime
 from tqdm import tqdm
-from pybaseutils import text_utils
+from pybaseutils import text_utils, thread_utils
 
 IMG_POSTFIX = ['*.jpg', '*.jpeg', '*.png', '*.tif', "*.JPG", "*.bmp"]
 VIDEO_POSTFIX = ['*.mp4', '*.avi', '*.mov', "*.flv", "*.dav"]
@@ -107,14 +107,14 @@ def combine_flags(flags: list, use_time=True, info=True):
     return out_flags
 
 
-def write_file(file, data):
+def write_file(file, data, mode='wb'):
     """写二进制数据"""
-    with open(file, 'wb') as f: f.write(data)
+    with open(file, mode) as f: f.write(data)
 
 
-def read_file(file):
+def read_file(file, mode='rb'):
     """读取二进制数据"""
-    with open(file, 'rb') as f: key = f.read()
+    with open(file, mode) as f: key = f.read()
     return key
 
 
@@ -161,7 +161,7 @@ def parser_classes(class_name):
     return class_name, class_dict
 
 
-def read_json_data(json_path):
+def load_json(json_path):
     """
     读取数据
     :param json_path:
@@ -172,7 +172,10 @@ def read_json_data(json_path):
     return json_data
 
 
-def write_json_path(json_file, json_data):
+read_json_data = load_json
+
+
+def save_json(json_file, json_data):
     """
     写入 JSON 数据
     :param json_file:
@@ -183,6 +186,30 @@ def write_json_path(json_file, json_data):
         json.dump(json_data, f, indent=4, ensure_ascii=False)
 
 
+write_json_path = save_json
+
+
+def load_json_files(files: list, max_workers=8):
+    """
+    根据url读取文件
+    :param files: 文件urls列表
+    :param max_workers: 开启线程数目
+    :return: 返回读取成功的文件数据file_list
+    """
+    # pool = thread_utils.ProcessPool(max_workers=max_workers)
+    pool = thread_utils.ThreadPool(max_workers=max_workers)
+    print("----" * 10)
+    print(f"启动{max_workers}个线程读取{len(files)}个文件,请等待....")
+    t0 = time.time()
+    file_list = pool.task_map(func=load_json, inputs=files)
+    t1 = time.time()
+    dt = (t1 - t0) * 1000
+    # 读取失败的url列表loss_list
+    loss_list = [files[i] for i in range(len(file_list)) if not file_list[i]]
+    print(f"成功:{len(file_list) - len(loss_list)},失败:{len(loss_list)},耗时:{dt:.2f}ms")
+    return file_list, loss_list
+
+
 def write_data(filename, content_list, split=",", mode='w'):
     """保存list[list[]]的数据到txt文件
     :param filename:文件名
@@ -565,8 +592,7 @@ def copy_dir(src, dst, sub=False, exclude=[]):
                 if isExclude: break
         if isExclude: continue
         dest_path = os.path.join(dst, os.path.relpath(root, src))
-        if not os.path.exists(dest_path):
-            os.makedirs(dest_path)
+        os.makedirs(dest_path, exist_ok=True)
         for filename in files:
             copy_file(
                 os.path.join(root, filename),
@@ -582,8 +608,7 @@ def move_dir(src, dst, sub=False):
     if sub: dst = os.path.join(dst, os.path.basename(src))
     for root, dirs, files in os.walk(src, topdown=False):
         dest_path = os.path.join(dst, os.path.relpath(root, src))
-        if not os.path.exists(dest_path):
-            os.makedirs(dest_path)
+        os.makedirs(dest_path, exist_ok=True)
         for filename in files:
             move_file(
                 os.path.join(root, filename),
@@ -595,8 +620,7 @@ def move_file(srcfile, dstfile):
     """ 移动文件或重命名"""
     if os.path.exists(srcfile) and os.path.isfile(srcfile):
         fpath, fname = os.path.split(dstfile)  # 分离文件名和路径
-        if not os.path.exists(fpath):
-            os.makedirs(fpath)  # 创建路径
+        os.makedirs(fpath, exist_ok=True)  # 创建路径
         shutil.move(srcfile, dstfile)
         # print("copy %s -> %s"%( srcfile,dstfile))
         # time.sleep(1 / 1000.)
@@ -615,8 +639,7 @@ def copy_file(srcfile, dstfile):
         print("%s not exist!" % (srcfile))
     else:
         fpath, fname = os.path.split(dstfile)  # 分离文件名和路径
-        if not os.path.exists(fpath):
-            os.makedirs(fpath)  # 创建路径
+        os.makedirs(fpath, exist_ok=True)  # 创建路径
         shutil.copyfile(srcfile, dstfile)  # 复制文件
         # print("copy %s -> %s"%( srcfile,dstfile))
         # time.sleep(1 / 1000.)
@@ -627,8 +650,7 @@ def copy_file_to_dir(srcfile, des_dir):
         print("%s not exist!" % (srcfile))
     else:
         fpath, fname = os.path.split(srcfile)  # 分离文件名和路径
-        if not os.path.exists(des_dir):
-            os.makedirs(des_dir)  # 创建路径
+        os.makedirs(des_dir, exist_ok=True)  # 创建路径
         dstfile = os.path.join(des_dir, fname)
         shutil.copyfile(srcfile, dstfile)  # 复制文件
 
@@ -638,8 +660,7 @@ def move_file_to_dir(srcfile, des_dir):
         print("%s not exist!" % (srcfile))
     else:
         fpath, fname = os.path.split(srcfile)  # 分离文件名和路径
-        if not os.path.exists(des_dir):
-            os.makedirs(des_dir)  # 创建路径
+        os.makedirs(des_dir, exist_ok=True)  # 创建路径
         dstfile = os.path.join(des_dir, fname)
         # shutil.copyfile(srcfile, dstfile)  # 复制文件
         move_file(srcfile, dstfile)  # 复制文件
@@ -653,22 +674,6 @@ def move_file_list(file_list, dst_dir):
     [move_file_to_dir(file, dst_dir) for file in file_list]
 
 
-def merge_dir(src, dst, sub, merge_same):
-    src_dir = os.path.join(src, sub)
-    dst_dir = os.path.join(dst, sub)
-
-    if not os.path.exists(src_dir):
-        print("\nno src path:{}".format(src))
-        return
-    if not os.path.exists(dst_dir):
-        os.makedirs(dst_dir)
-    elif not merge_same:
-        t = get_time()
-        dst_dir = os.path.join(dst, sub + "_{}".format(t))
-        print("have save sub:{}".format(dst_dir))
-    copy_dir(src_dir, dst_dir)
-
-
 def create_dir(parent_dir, dir1=None, filename=None):
     """
     create directory
@@ -681,8 +686,7 @@ def create_dir(parent_dir, dir1=None, filename=None):
     if dir1:
         out_path = os.path.join(parent_dir, dir1)
     if not out_path: return out_path
-    if not os.path.exists(out_path):
-        os.makedirs(out_path)
+    os.makedirs(out_path, exist_ok=True)
     if filename:
         out_path = os.path.join(out_path, filename)
     return out_path
@@ -700,7 +704,7 @@ def create_file_path(filename):
     return out_path
 
 
-def get_sub_paths(input_dir):
+def get_sub_paths(input_dir, abspath=False):
     """
     当前路径下所有子目录
     :param input_dir:
@@ -714,6 +718,7 @@ def get_sub_paths(input_dir):
     # print(dirs)   # 当前路径下所有子目录
     # print(files)  # 当前路径下所有非目录子文件
     sub_list.sort()
+    if abspath: sub_list = [os.path.join(input_dir, p) for p in sub_list]
     return sub_list
 
 
@@ -1373,6 +1378,10 @@ def copy_move_voc_dataset(data_file, data_root=None, out_root=None, file_map={},
 
 
 if __name__ == '__main__':
-    dir = "/home/dm/nasdata/dataset-dmai/handwriting/word-class/trainval/unknown"
-    file_list, label_list = get_files_labels(dir)
-    print(label_list)
+    from pybaseutils import time_utils
+
+    path = "/home/PKing/Downloads/tmp"
+    file_list = get_files_list(path, postfix=["*.json"])
+    with time_utils.Performance():
+        file_data, loss_list = load_json_files(file_list, max_workers=1) # 12433.96ms
+    print(len(file_data))

{pybaseutils-2.2.20 → pybaseutils-2.2.22}/pybaseutils/font_style/__init__.py

@@ -6,7 +6,7 @@
 """
 import os
 
-# TODO 如果字体缺失，可以把字体拷贝到系统字体目录：cp -r fonts/* ~/.local/share/fonts/
+# TODO 如果字体缺失，可以把字体拷贝到系统字体目录：cp -r font_style/*
 # windows: C:\Windows\Fonts可查看系统支持的字体
 # Linux  : /usr/share/fonts/truetype 或者~/.local/share/fonts中 可查看系统支持的字体
 

{pybaseutils-2.2.20 → pybaseutils-2.2.22}/pybaseutils/font_utils.py

@@ -58,9 +58,6 @@ class FontType(object):
         return self.font_type
 
 
-# font_type = FontType()
-
-
 def get_font_type(size, font=""):
     """
     Windows字体路径      : /usr/share/fonts/楷体.ttf
@@ -101,8 +98,7 @@ def draw_image_text(image, point, text, style="楷体", size=20, color=(255, 255
         font = ImageFont.truetype(style, size=size, encoding="utf-8")
     else:
         # simhei.ttf 是字体，你如果没有字体，需要下载
-        font_type.set_font_style(style=style, size=size)
-        font = font_type.get_font_type()
+        font = get_font_type(size=size)
     draw.text(point, text, fill=color, font=font)
     image = np.asarray(image)
     return image
@@ -211,6 +207,24 @@ def get_font_char(font_file, only_chinese=False):
     return fonts
 
 
+def set_pyplot_font(font="simhei"):
+    """
+    设置 seaborn，matplotlib等默认字体
+    :param font: 字体路径或者名称，或者使用默认的,SimHei
+    :return:
+    """
+    import seaborn
+    import matplotlib.pyplot as plt
+    import matplotlib.font_manager as fm
+    font_file = font if os.path.isfile(font) else FONT_TABLES[font]
+    font = os.path.basename(font_file).split(".")[0]
+    fm.fontManager.addfont(font_file)
+    plt.rcParams['font.sans-serif'] = [font]
+    plt.rcParams['axes.unicode_minus'] = False
+    seaborn.set_theme(font=font)  # 增加一行代码, 设置 seaborn 使用的字体文件
+    return font_file
+
+
 def draw_font_example():
     size = 512
     string = "我"

pybaseutils-2.2.22/pybaseutils/http_utils.py

@@ -0,0 +1,230 @@
+# -*- coding: utf-8 -*-
+"""
+# --------------------------------------------------------
+# @Author : Pan
+# @E-mail : 390737991@qq.com
+# @Date   : 2019-12-31 09:11:25
+# --------------------------------------------------------
+"""
+import os
+import time
+import requests
+import json
+from pybaseutils import log_utils, text_utils, json_utils, thread_utils
+
+logger = log_utils.get_logger()
+
+
+def post(url, params=None, timeout=6, max_retries=1, **kwargs):
+    """
+    :param url: 请求URL
+    :param params: 请求参数
+    :param timeout: 设置超时
+    :return:
+    """
+    result = None
+    counts = 0
+    while counts < max_retries:
+        try:
+            t1 = time.time()
+            # r = requests.get(url, params=params, timeout=timeout,**kwargs)
+            r = requests.post(url, json=params, timeout=timeout, **kwargs)
+            t2 = time.time()
+            elapsed = (t2 - t1) * 1000
+            code = r.status_code
+            if code == 200:
+                logger.info(f'code={code}, url={url}, elapsed:{elapsed:3.3f}ms')
+                result = r.json()
+                break
+            else:
+                r.raise_for_status()  # 如果响应状态码不是200，抛出异常
+        except Exception as e:
+            counts += 1
+            logger.error(f'Error msg:{e}, url={url}, try to retry times={counts}/{max_retries}')
+            time.sleep(0.1)
+    return result
+
+
+def get(url, params=None, timeout=6, max_retries=1, **kwargs):
+    """
+    :param url: 请求URL
+    :param params: 请求参数
+    :param timeout: 设置超时
+    :return:
+    """
+    result = None
+    counts = 0
+    log = kwargs.pop("log", True)
+    while counts < max_retries:
+        try:
+            t1 = time.time()
+            r = requests.get(url, params=params, timeout=timeout, **kwargs)
+            # r = requests.post(url, json=params, timeout=timeout, **kwargs)
+            t2 = time.time()
+            elapsed = (t2 - t1) * 1000
+            code = r.status_code
+            if code == 200:
+                if log: logger.info(f'code={code}, url={url}, elapsed:{elapsed:3.3f}ms')
+                result = r.json()
+                break
+            else:
+                r.raise_for_status()  # 如果响应状态码不是200，抛出异常
+        except Exception as e:
+            counts += 1
+            logger.error(f'Error msg:{e}, url={url}, try to retry times={counts}/{max_retries}')
+            time.sleep(0.1)
+    return result
+
+
+def get_url_files(url, prefix="", postfix=None, basename=False, timeout=30):
+    """
+    获得文件服务器中文件列表
+    :param url: http URL地址等
+    :param prefix: http URL地址等
+    :param postfix: None或者[]表示所有文件
+    :param basename: 返回的列表是文件名（True），还是文件的完整路径(False)
+    :return:
+    """
+    from bs4 import BeautifulSoup
+    try:
+        if url.endswith("/"): url = url[:-1]
+        # 设置超时和重试
+        r = requests.get(url, timeout=timeout, auth=("admin", "admin"))
+        r.raise_for_status()
+        # 解析目录列表
+        soup = BeautifulSoup(r.text, 'html.parser')
+        file_list = [a['href'] for a in soup.select('a[href]')]
+        # 过滤并拼接完整URL
+        file_list = text_utils.find_match_texts(file_list, pattern=[prefix], org=True) if prefix else file_list
+        file_list = text_utils.find_match_texts(file_list, pattern=postfix, org=True) if postfix else file_list
+        file_list.sort()
+        file_list = file_list if basename else [f"{url}/{f}" for f in file_list]
+        return file_list
+    except Exception as e:
+        print(f"请求失败: {str(e)}")
+        return []
+
+
+def download_file(url, out, timeout=30, max_retries=1, log=False):
+    """
+    根据url下载文件
+    :param url: 文件URL
+    :param out: 输出保存目录
+    :param log: 是否打印LOG信息
+    :return:
+    """
+    try:
+        os.makedirs(out, exist_ok=True)
+        name = url.split('/')[-1]
+        path = os.path.join(out, name)
+        data = get(url, params=None, timeout=timeout, max_retries=max_retries, log=False)
+        assert data
+        if name.endswith("json"):
+            json_utils.save_json(path, data)
+        else:
+            with open(path, 'w', encoding='utf-8') as f:
+                f.write(data)
+        if log: print(f"下载成功: {url} --> {path}")
+        return path
+    except Exception as e:
+        if log: print(f"下载失败: {url}")
+        return None
+
+
+def download_files(urls: list, out, max_workers=8, timeout=30, max_retries=1):
+    """
+    :param urls: 文件URL列表
+    :param out: 下载保持目录
+    :param max_workers:
+    :param timeout:
+    :param max_retries:
+    :return: 返回下载成功的本地文件列表file_list
+    """
+    pool = thread_utils.ThreadPool(max_workers=max_workers)
+    inputs = [(url, out, timeout, max_retries) for url in urls]
+    print("----" * 10)
+    print(f"启动{max_workers}个线程下载{len(urls)}个文件,请等待....")
+    t0 = time.time()
+    file_list = pool.task_maps(func=download_file, inputs=inputs)
+    t1 = time.time()
+    dt = (t1 - t0) * 1000
+    # 下载失败的url列表loss_list
+    loss_list = [urls[i] for i in range(len(file_list)) if not file_list[i]]
+    print(f"成功:{len(file_list) - len(loss_list)},失败:{len(loss_list)},耗时:{dt:.2f}ms")
+    return file_list, loss_list
+
+
+def read_url_file(url, timeout=30, max_retries=1):
+    """
+    根据url读取文件
+    :param url: 文件URL
+    :param timeout:
+    :param max_retries:
+    :return:
+    """
+    try:
+        data = get(url, params=None, timeout=timeout, max_retries=max_retries, log=False)
+        return data
+    except Exception as e:
+        print(f"读取失败: {url}")
+        return None
+
+
+def read_url_files(urls: list, max_workers=8, timeout=30, max_retries=1):
+    """
+    根据url读取文件
+    :param urls: 文件urls列表
+    :param max_workers: 开启线程数目
+    :param timeout:
+    :param max_retries:
+    :return: 返回读取成功的文件数据file_list
+    """
+    pool = thread_utils.ThreadPool(max_workers=max_workers)
+    inputs = [(url, timeout, max_retries) for url in urls]
+    print("----" * 10)
+    print(f"启动{max_workers}个线程读取{len(urls)}个文件,请等待....")
+    t0 = time.time()
+    file_list = pool.task_maps(func=read_url_file, inputs=inputs)
+    t1 = time.time()
+    dt = (t1 - t0) * 1000
+    # 读取失败的url列表loss_list
+    loss_list = [urls[i] for i in range(len(file_list)) if not file_list[i]]
+    print(f"成功:{len(file_list) - len(loss_list)},失败:{len(loss_list)},耗时:{dt:.2f}ms")
+    return file_list, loss_list
+
+
+def tojson(data: dict, keys: list):
+    """
+    指定需要反序列的数据
+    :param data:
+    :param keys:
+    :return:
+    """
+    if isinstance(data, dict):
+        for k, v in data.items():
+            if k in keys:
+                try:
+                    data[k] = json.loads(v)
+                except Exception as e:
+                    print(e)
+            else:
+                data[k] = tojson(v, keys=keys)
+    elif isinstance(data, list):
+        for i in range(len(data)):
+            data[i] = tojson(data[i], keys=keys)
+    return data
+
+
+if __name__ == '__main__':
+    urls = [
+        "http://aije-mvp-nginx.partner.dm-ai.com/req-resp/aije-job-m8ch701s-awbg/nlp-12.json",
+        "http://aije-mvp-nginx.partner.dm-ai.com/req-resp/aije-job-m8ch701s-awbg/nlp-13.json",
+        "http://aije-mvp-nginx.partner.dm-ai.com/req-resp/aije-job-m8ch701s-awbg/nlp-17.json",
+    ]
+    out = "/home/PKing/Downloads/tmp"
+    url = 'https://aije-mvp-nginx.partner.dm-ai.com/req-resp/aije-job-m911dzow-17cu'
+    urls = get_url_files(url, postfix=["*.json"])
+    file_list1, loss_list1 = download_files(urls, out=out, max_retries=1)
+    file_list2, loss_list2 = read_url_files(urls, max_retries=1)
+    print(loss_list1)
+    print(loss_list2)

{pybaseutils-2.2.20 → pybaseutils-2.2.22}/pybaseutils/image_utils.py

@@ -2862,6 +2862,32 @@ def get_image_block(image, grid=[3, 3], same=False):
     return image_block
 
 
+def apply_mosaic(image, boxes, radius=8, scale=[1.0, 1.0]):
+    """
+    马赛克
+    :param image: BGR image
+    :param boxes: 人脸框(xmin,ymin,xmax,ymax)
+    :param radius: 马赛克强度
+    :param scale: 人脸框扩大范围
+    :return:
+    """
+    ar = (5, 20)
+    h, w = image.shape[:2]
+    boxes = np.asarray(boxes, dtype=np.int32)
+    boxes = extend_xyxy(boxes, scale=scale, valid_range=(0, 0, w, h))
+    for box in boxes:
+        x1, y1, x2, y2 = box
+        roi = image[y1:y2, x1:x2]
+        size = [(x2 - x1) // radius, (y2 - y1) // radius]
+        size[0] = max(min(size[0], ar[1]), ar[0])
+        size[1] = max(min(size[1], ar[1]), ar[0])
+        img = cv2.resize(roi, tuple(size))
+        # img = cv2.resize(roi, (radius, radius))
+        img = cv2.resize(img, (x2 - x1, y2 - y1), interpolation=cv2.INTER_NEAREST)
+        image[y1:y2, x1:x2] = img
+    return image
+
+
 def image_composite(image: np.ndarray, alpha: np.ndarray, bg_img=(219, 142, 67)):
     """
     图像融合：合成图 = 前景*alpha+背景*(1-alpha)