docs: 添加涉密文件自检工具实施计划

UmiOCR-data/plugins/win7_x64_RapidOCR-json/__init__.py

@@ -0,0 +1,14 @@
+from . import api_rapidocr
+from . import rapidocr_config
+
+# 插件信息
+PluginInfo = {
+    # 插件组别
+    "group": "ocr",
+    # 全局配置
+    "global_options": rapidocr_config.globalOptions,
+    # 局部配置
+    "local_options": rapidocr_config.localOptions,
+    # 接口类
+    "api_class": api_rapidocr.Api,
+}

UmiOCR-data/plugins/win7_x64_RapidOCR-json/api_rapidocr.py

@@ -0,0 +1,79 @@
+# 调用 RapidOCR-json.exe 的 Python Api
+# 项目主页：
+# https://github.com/hiroi-sora/RapidOCR-json
+
+import os
+from .rapidocr import Rapid_pipe
+from .rapidocr_config import LangDict
+
+# exe路径
+ExePath = os.path.dirname(os.path.abspath(__file__)) + "/RapidOCR-json.exe"
+
+
+class Api:  # 公开接口
+    def __init__(self, globalArgd):
+        # 测试路径是否存在
+        if not os.path.exists(ExePath):
+            raise ValueError(f'[Error] Exe path "{ExePath}" does not exist.')
+        # 初始化参数
+        self.api = None  # api对象
+        self.exeConfigs = {  # exe启动参数字典
+            "models": "models",
+            "ensureAscii": 1,
+            "det": None,
+            "cls": None,
+            "rec": None,
+            "keys": None,
+            "doAngle": 0,
+            "mostAngle": 0,
+            "maxSideLen": None,
+            "numThread": globalArgd["numThread"],
+        }
+
+    def start(self, argd):  # 启动引擎。返回： "" 成功，"[Error] xxx" 失败
+        # 加载局部参数
+        tempConfigs = self.exeConfigs.copy()
+        try:
+            lang = LangDict[argd["language"]]
+            tempConfigs.update(lang)
+            if argd["angle"]:
+                tempConfigs["doAngle"] = tempConfigs["mostAngle"] = 1
+            else:
+                tempConfigs["doAngle"] = tempConfigs["mostAngle"] = 0
+            tempConfigs["maxSideLen"] = argd["maxSideLen"]
+        except Exception as e:
+            self.api = None
+            return f"[Error] OCR start fail. Argd: {argd}\n{e}"
+
+        # 若引擎已启动，且局部参数与传入参数一致，则无需重启
+        if not self.api == None:
+            if set(tempConfigs.items()) == set(self.exeConfigs.items()):
+                return ""
+            # 若引擎已启动但需要更改参数，则停止旧引擎
+            self.stop()
+        # 启动新引擎
+        self.exeConfigs = tempConfigs
+        try:
+            self.api = Rapid_pipe(ExePath, tempConfigs)
+        except Exception as e:
+            self.api = None
+            return f"[Error] OCR init fail. Argd: {tempConfigs}\n{e}"
+        return ""
+
+    def stop(self):  # 停止引擎
+        if self.api == None:
+            return
+        self.api.exit()
+        self.api = None
+
+    def runPath(self, imgPath: str):  # 路径识图
+        res = self.api.run(imgPath)
+        return res
+
+    def runBytes(self, imageBytes):  # 字节流
+        res = self.api.runBytes(imageBytes)
+        return res
+
+    def runBase64(self, imageBase64):  # base64字符串
+        res = self.api.runBase64(imageBase64)
+        return res

UmiOCR-data/plugins/win7_x64_RapidOCR-json/i18n.csv

@@ -0,0 +1,11 @@
+key,en_US,zh_TW,ja_JP
+RapidOCR（本地）,RapidOCR (local),RapidOCR（本地）,RapidOCR（ローカル）
+线程数,Number of threads,線程數,スレッド数
+文字识别（RapidOCR）,Text Recognition (RapidOCR),文字識別（RapidOCR）,文字認識（RapidOCR）
+语言/模型库,Language/Model Library,語言/模型庫,言語/モデルライブラリ
+纠正文本方向,Correct Text Direction,糾正文字方向,テキスト方向の修正
+启用方向分类，识别倾斜或倒置的文本。可能降低识别速度。,Enable directional classification to recognize tilted or inverted text. May reduce recognition speed.,啟用方向分類，識別傾斜或倒置的文字。 可能降低識別速度。,方向分類を有効にして、傾斜または反転したテキストを識別します。認識速度が低下する可能性があります。
+限制图像边长,Limit image edge length,限制影像邊長,画像の辺の長さを制限する
+（默认）,(Default),（默認）,(デフォルト)
+无限制,Unlimited,無限制,制限なし
+将边长大于该值的图片进行压缩，可以提高识别速度。可能降低识别精度。,Compressing images with side lengths greater than this value can improve recognition speed. May reduce recognition accuracy.,將邊長大於該值的圖片進行壓縮，可以提高識別速度。 可能降低識別精度。,辺の長さがこの値より大きい画像を圧縮することで、認識速度を高めることができます。認識精度が低下する可能性があります。

UmiOCR-data/plugins/win7_x64_RapidOCR-json/models/configs.txt

@@ -0,0 +1,35 @@
+简体中文
+ch_PP-OCRv3_det_infer.onnx
+ch_ppocr_mobile_v2.0_cls_infer.onnx
+rec_ch_PP-OCRv4_infer.onnx
+dict_chinese.txt
+
+English
+ch_PP-OCRv3_det_infer.onnx
+ch_ppocr_mobile_v2.0_cls_infer.onnx
+rec_en_PP-OCRv3_infer.onnx
+dict_en.txt
+
+繁體中文
+ch_PP-OCRv3_det_infer.onnx
+ch_ppocr_mobile_v2.0_cls_infer.onnx
+rec_chinese_cht_PP-OCRv3_infer.onnx
+dict_chinese_cht.txt
+
+日本語
+ch_PP-OCRv3_det_infer.onnx
+ch_ppocr_mobile_v2.0_cls_infer.onnx
+rec_japan_PP-OCRv3_infer.onnx
+dict_japan.txt
+
+한국어
+ch_PP-OCRv3_det_infer.onnx
+ch_ppocr_mobile_v2.0_cls_infer.onnx
+rec_korean_PP-OCRv3_infer.onnx
+dict_korean.txt
+
+Русский
+ch_PP-OCRv3_det_infer.onnx
+ch_ppocr_mobile_v2.0_cls_infer.onnx
+rec_cyrillic_PP-OCRv3_infer.onnx
+dict_cyrillic.txt

UmiOCR-data/plugins/win7_x64_RapidOCR-json/models/dict_cyrillic.txt

@@ -0,0 +1,163 @@
+ 
+!
+#
+$
+%
+&
+'
+(
++
+,
+-
+.
+/
+0
+1
+2
+3
+4
+5
+6
+7
+8
+9
+:
+?
+@
+A
+B
+C
+D
+E
+F
+G
+H
+I
+J
+K
+L
+M
+N
+O
+P
+Q
+R
+S
+T
+U
+V
+W
+X
+Y
+Z
+_
+a
+b
+c
+d
+e
+f
+g
+h
+i
+j
+k
+l
+m
+n
+o
+p
+q
+r
+s
+t
+u
+v
+w
+x
+y
+z
+É
+é
+Ё
+Є
+І
+Ј
+Љ
+Ў
+А
+Б
+В
+Г
+Д
+Е
+Ж
+З
+И
+Й
+К
+Л
+М
+Н
+О
+П
+Р
+С
+Т
+У
+Ф
+Х
+Ц
+Ч
+Ш
+Щ
+Ъ
+Ы
+Ь
+Э
+Ю
+Я
+а
+б
+в
+г
+д
+е
+ж
+з
+и
+й
+к
+л
+м
+н
+о
+п
+р
+с
+т
+у
+ф
+х
+ц
+ч
+ш
+щ
+ъ
+ы
+ь
+э
+ю
+я
+ё
+ђ
+є
+і
+ј
+љ
+њ
+ћ
+ў
+џ
+Ґ
+ґ

UmiOCR-data/plugins/win7_x64_RapidOCR-json/models/dict_en.txt

@@ -0,0 +1,95 @@
+0
+1
+2
+3
+4
+5
+6
+7
+8
+9
+:
+;
+<
+=
+>
+?
+@
+A
+B
+C
+D
+E
+F
+G
+H
+I
+J
+K
+L
+M
+N
+O
+P
+Q
+R
+S
+T
+U
+V
+W
+X
+Y
+Z
+[
+\
+]
+^
+_
+`
+a
+b
+c
+d
+e
+f
+g
+h
+i
+j
+k
+l
+m
+n
+o
+p
+q
+r
+s
+t
+u
+v
+w
+x
+y
+z
+{
+|
+}
+~
+!
+"
+#
+$
+%
+&
+'
+(
+)
+*
++
+,
+-
+.
+/
+ 

UmiOCR-data/plugins/win7_x64_RapidOCR-json/models/dict_french.txt

@@ -0,0 +1,136 @@
+f
+e
+n
+c
+h
+_
+i
+m
+g
+/
+r
+v
+a
+l
+t
+w
+o
+d
+6
+1
+.
+p
+B
+u
+2
+à
+3
+R
+y
+4
+U
+E
+A
+5
+P
+O
+S
+T
+D
+7
+Z
+8
+I
+N
+L
+G
+M
+H
+0
+J
+K
+-
+9
+F
+C
+V
+é
+X
+'
+s
+Q
+:
+è
+x
+b
+Y
+Œ
+É
+z
+W
+Ç
+È
+k
+Ô
+ô
+€
+À
+Ê
+q
+ù
+°
+ê
+î
+*
+Â
+j
+"
+,
+â
+%
+û
+ç
+ü
+?
+!
+;
+ö
+(
+)
+ï
+º
+ó
+ø
+å
++
+™
+á
+Ë
+<
+²
+Á
+Î
+&
+@
+œ
+ε
+Ü
+ë
+[
+]
+í
+ò
+Ö
+ä
+ß
+«
+»
+ú
+ñ
+æ
+µ
+³
+Å
+$
+#
+ 

UmiOCR-data/plugins/win7_x64_RapidOCR-json/models/dict_german.txt

@@ -0,0 +1,143 @@
+ 
+!
+"
+#
+$
+%
+&
+'
+(
+)
+*
++
+,
+-
+.
+/
+0
+1
+2
+3
+4
+5
+6
+7
+8
+9
+:
+;
+=
+>
+?
+@
+A
+B
+C
+D
+E
+F
+G
+H
+I
+J
+K
+L
+M
+N
+O
+P
+Q
+R
+S
+T
+U
+V
+W
+X
+Y
+Z
+[
+]
+_
+a
+b
+c
+d
+e
+f
+g
+h
+i
+j
+k
+l
+m
+n
+o
+p	
+q
+r
+s
+t
+u
+v
+w
+x
+y
+z
+£
+§
+­
+°
+´
+µ
+·
+º
+¿
+Á
+Ä
+Å
+É
+Ï
+Ô
+Ö
+Ü
+ß
+à
+á
+â
+ã
+ä
+å
+æ
+ç
+è
+é
+ê
+ë
+í
+ï
+ñ
+ò
+ó
+ô
+ö
+ø
+ù
+ú
+û
+ü
+ō
+Š
+Ÿ
+ʒ
+β
+δ
+з
+Ṡ
+‘
+€
+©
+ª
+«
+¬

UmiOCR-data/plugins/win7_x64_RapidOCR-json/rapidocr.py

@@ -0,0 +1,124 @@
+import os
+import atexit  # 退出处理
+import subprocess  # 进程，管道
+from json import loads as jsonLoads, dumps as jsonDumps
+from sys import platform as sysPlatform  # popen静默模式
+from base64 import b64encode  # base64 编码
+
+
+class Rapid_pipe:  # 调用OCR（管道模式）
+    def __init__(self, exePath: str, argument: dict = None):
+        """初始化识别器（管道模式）。\n
+        `exePath`: 识别器`PaddleOCR_json.exe`的路径。\n
+        `argument`: 启动参数，字典`{"键":值}`。参数说明见 https://github.com/hiroi-sora/PaddleOCR-json
+        """
+        cwd = os.path.abspath(os.path.join(exePath, os.pardir))  # 获取exe父文件夹
+        # 处理启动参数
+        if not argument == None:
+            for key, value in argument.items():
+                if isinstance(value, str):  # 字符串类型的值加双引号
+                    exePath += f' --{key}="{value}"'
+                else:
+                    exePath += f" --{key}={value}"
+        if "ensureAscii" not in exePath:
+            exePath += f" --ensureAscii=1"
+        # 设置子进程启用静默模式，不显示控制台窗口
+        self.ret = None
+        startupinfo = None
+        if "win32" in str(sysPlatform).lower():
+            startupinfo = subprocess.STARTUPINFO()
+            startupinfo.dwFlags = (
+                subprocess.CREATE_NEW_CONSOLE | subprocess.STARTF_USESHOWWINDOW
+            )
+            startupinfo.wShowWindow = subprocess.SW_HIDE
+        self.ret = subprocess.Popen(  # 打开管道
+            exePath,
+            cwd=cwd,
+            stdin=subprocess.PIPE,
+            stdout=subprocess.PIPE,
+            stderr=subprocess.DEVNULL,  # 丢弃stderr的内容
+            startupinfo=startupinfo,  # 开启静默模式
+        )
+        # 启动子进程
+        while True:
+            if not self.ret.poll() == None:  # 子进程已退出，初始化失败
+                raise Exception(f"OCR init fail.")
+            initStr = self.ret.stdout.readline().decode("utf-8", errors="ignore")
+            if "OCR init completed." in initStr:  # 初始化成功
+                break
+        atexit.register(self.exit)  # 注册程序终止时执行强制停止子进程
+
+    def runDict(self, writeDict: dict):
+        """传入指令字典，发送给引擎进程。\n
+        `writeDict`: 指令字典。\n
+        `return`:  {"code": 识别码, "data": 内容列表或错误信息字符串}\n"""
+        # 检查子进程
+        if not self.ret:
+            return {"code": 901, "data": f"引擎实例不存在。"}
+        if not self.ret.poll() == None:
+            return {"code": 902, "data": f"子进程已崩溃。"}
+        # 输入信息
+        writeStr = jsonDumps(writeDict, ensure_ascii=True, indent=None) + "\n"
+        try:
+            self.ret.stdin.write(writeStr.encode("utf-8"))
+            self.ret.stdin.flush()
+        except Exception as e:
+            return {"code": 902, "data": f"向识别器进程传入指令失败，疑似子进程已崩溃。{e}"}
+        # 获取返回值
+        try:
+            getStr = self.ret.stdout.readline().decode("utf-8", errors="ignore")
+        except Exception as e:
+            return {"code": 903, "data": f"读取识别器进程输出值失败。异常信息：[{e}]"}
+        try:
+            return jsonLoads(getStr)
+        except Exception as e:
+            return {"code": 904, "data": f"识别器输出值反序列化JSON失败。异常信息：[{e}]。原始内容：[{getStr}]"}
+
+    def run(self, imgPath: str):
+        """对一张本地图片进行文字识别。\n
+        `exePath`: 图片路径。\n
+        `return`:  {"code": 识别码, "data": 内容列表或错误信息字符串}\n"""
+        writeDict = {"image_path": imgPath}
+        return self.runDict(writeDict)
+
+    def runBase64(self, imageBase64: str):
+        """对一张编码为base64字符串的图片进行文字识别。\n
+        `imageBase64`: 图片base64字符串。\n
+        `return`:  {"code": 识别码, "data": 内容列表或错误信息字符串}\n"""
+        writeDict = {"image_base64": imageBase64}
+        return self.runDict(writeDict)
+
+    def runBytes(self, imageBytes):
+        """对一张图片的字节流信息进行文字识别。\n
+        `imageBytes`: 图片字节流。\n
+        `return`:  {"code": 识别码, "data": 内容列表或错误信息字符串}\n"""
+        imageBase64 = b64encode(imageBytes).decode("utf-8")
+        return self.runBase64(imageBase64)
+
+    def exit(self):
+        """关闭引擎子进程"""
+        if not self.ret:
+            return
+        self.ret.kill()  # 关闭子进程
+        self.ret = None
+        atexit.unregister(self.exit)  # 移除退出处理
+        print("### RapidOCR引擎子进程关闭！")
+
+    @staticmethod
+    def printResult(res: dict):
+        """用于调试，格式化打印识别结果。\n
+        `res`: OCR识别结果。"""
+
+        # 识别成功
+        if res["code"] == 100:
+            index = 1
+            for line in res["data"]:
+                print(f"{index}-置信度：{round(line['score'], 2)}，文本：{line['text']}")
+                index += 1
+        elif res["code"] == 100:
+            print("图片中未识别出文字。")
+        else:
+            print(f"图片识别失败。错误码：{res['code']}，错误信息：{res['data']}")
+
+    def __del__(self):
+        self.exit()

UmiOCR-data/plugins/win7_x64_RapidOCR-json/rapidocr_config.py

@@ -0,0 +1,109 @@
+import os
+import psutil
+from plugin_i18n import Translator
+
+tr = Translator(__file__, "i18n.csv")
+
+# 模块配置路径
+MODELS_CONFIGS = "/models/configs.txt"
+# 保存语言字典
+LangDict = {}
+
+
+# 动态获取模型库列表
+def _getlanguageList():
+    global LangDict
+    """configs.txt 格式示例：
+    简体中文(V4)
+    ch_PP-OCRv4_det_infer.onnx
+    ch_ppocr_mobile_v2.0_cls_infer.onnx
+    rec_ch_PP-OCRv4_infer.onnx
+    dict_chinese.txt
+
+    """
+    optionsList = []
+    configsPath = os.path.dirname(os.path.abspath(__file__)) + MODELS_CONFIGS
+    try:
+        with open(configsPath, "r", encoding="utf-8") as file:
+            content = file.read()
+            parts = content.split("\n\n")
+            for part in parts:
+                items = part.split("\n")
+                if len(items) == 5:
+                    title, det, cls, rec, keys = items
+                    LangDict[title] = {
+                        "det": det,
+                        "cls": cls,
+                        "rec": rec,
+                        "keys": keys,
+                    }
+                    optionsList.append([title, title])
+        return optionsList
+    except FileNotFoundError:
+        print("[Error] RapidOCR配置文件configs不存在，请检查文件路径是否正确。", configsPath)
+    except IOError:
+        print("[Error] RapidOCR配置文件configs无法打开或读取。")
+    return []
+
+
+_LanguageList = _getlanguageList()
+
+
+# 获取最佳线程数
+def _getThreads():
+    try:
+        phyCore = psutil.cpu_count(logical=False)  # 物理核心数
+        lgiCore = psutil.cpu_count(logical=True)  # 逻辑核心数
+        if (
+            not isinstance(phyCore, int)
+            or not isinstance(lgiCore, int)
+            or lgiCore < phyCore
+        ):
+            raise ValueError("核心数计算异常")
+        # 物理核数=逻辑核数，返回逻辑核数
+        if phyCore * 2 == lgiCore or phyCore == lgiCore:
+            return lgiCore
+        # 大小核处理器，返回大核线程数
+        big = lgiCore - phyCore
+        return big * 2
+    except Exception as e:
+        print("[Warning] 无法获取CPU核心数！", e)
+        return 4
+
+
+_threads = _getThreads()
+
+globalOptions = {
+    "title": tr("RapidOCR（本地）"),
+    "type": "group",
+    "numThread": {
+        "title": tr("线程数"),
+        "default": _threads,
+        "min": 1,
+        "isInt": True,
+    },
+}
+
+localOptions = {
+    "title": tr("文字识别（RapidOCR）"),
+    "type": "group",
+    "language": {
+        "title": tr("语言/模型库"),
+        "optionsList": _LanguageList,
+    },
+    "angle": {
+        "title": tr("纠正文本方向"),
+        "default": False,
+        "toolTip": tr("启用方向分类，识别倾斜或倒置的文本。可能降低识别速度。"),
+    },
+    "maxSideLen": {
+        "title": tr("限制图像边长"),
+        "optionsList": [
+            [1024, "1024 " + tr("（默认）")],
+            [2048, "2048"],
+            [4096, "4096"],
+            [999999, tr("无限制")],
+        ],
+        "toolTip": tr("将边长大于该值的图片进行压缩，可以提高识别速度。可能降低识别精度。"),
+    },
+}