[ocr] v2 (#49)

dont update , last version works fine

ocr/__init__.py

@@ -1,9 +1,11 @@
-from discord.utils import maybe_coroutine
+from redbot.core.errors import CogLoadError
 
 from .ocr import OCR
 
 __red_end_user_data_statement__ = "This cog does not persistently store any data or metadata about users."
 
 
 async def setup(bot):
-    await maybe_coroutine(bot.add_cog, OCR())
+    if not getattr(bot, "session", None):
+        raise CogLoadError("This cog requires bot.session attr to be set.")
+    await bot.add_cog(OCR(bot))

ocr/converter.py

@@ -1,63 +1,75 @@
-# This is a slightly modified version of converter originally made by TrustyJAID for his NotSoBot cog
-# Attribution at: https://github.com/TrustyJAID/Trusty-cogs/blob/master/notsobot/converter.py
-# I have included original LICENSE notice bundled with this cog to adhere to the license terms.
-# I am forever indebted to and wholeheartedly thank TrustyJAID for providing this converter.
+# originally made by TrustyJAID for his NotSoBot cog
+# https://github.com/TrustyJAID/Trusty-cogs/blob/master/notsobot/converter.py
 import re
 
-from typing import Pattern, List
-
 import discord
-from redbot.core import commands
+from redbot.core.commands import BadArgument, Context, Converter
+
+IMAGE_LINKS: re.Pattern[str] = re.compile(
+    r"(https?:\/\/[^\"\'\s]*\.(?:png|jpg|jpeg|webp)(\?size=[0-9]{1,4})?)", flags=re.I
+)
 
-IMAGE_LINKS: Pattern = re.compile(
-    r"(https?:\/\/[^\"\'\s]*\.(?:png|jpg|jpeg|webp|gif)(\?size=[0-9]*)?)", flags=re.I
+DISCORD_CDN: tuple[str, str] = (
+    "https://cdn.discordapp.com/attachments",
+    "https://media.discordapp.net/attachments",
 )
 
 
-class ImageFinder(commands.Converter):
+class ImageFinder(Converter):
     """
     This is a class to convert NotSoBot's image searching
     capabilities into a more general converter class
     """
 
-    async def convert(self, ctx: commands.Context, argument: str) -> List[str]:
-        attachments = ctx.message.attachments
-        matches = IMAGE_LINKS.finditer(argument)
-        urls = []
-        if matches:
+    async def convert(self, ctx: Context, argument: str) -> list[str]:
+        urls: list[str] = []
+        if argument.startswith(DISCORD_CDN):
+            urls.append(argument.split()[0])
+        if matches := IMAGE_LINKS.finditer(argument):
             urls.extend(match.group(1) for match in matches)
-        if attachments:
-            urls.extend(
-                match.group(1)
-                for attachment in attachments
-                if (match := IMAGE_LINKS.match(attachment.url))
-            )
-
+        if attachments := ctx.message.attachments:
+            urls.extend(img.url for img in attachments if img.content_type and img.content_type.startswith("image"))
+        if (e := ctx.message.embeds) and e[0].image:
+            urls.append(e[0].image.url)
+        if not urls:
+            if ctx.message.reference and (message := ctx.message.reference.resolved):
+                urls = await find_images_in_replies(message)
+            else:
+                urls = await search_for_images(ctx)
+        if not urls:
+            raise BadArgument("No images or image links found in chat bro 🥸")
         return urls
 
-    async def find_images_in_replies(self, reference: discord.Message) -> List[str]:
-        urls = []
-        if match := IMAGE_LINKS.search(reference.content):
-            urls.append(match.group(1))
-        if reference.attachments:
-            if match := IMAGE_LINKS.match(reference.attachments[0].url):
-                urls.append(match.group(1))
-        if reference.embeds and reference.embeds[0].image:
-            urls.append(reference.embeds[0].image.url)
-        return urls
 
-    async def search_for_images(self, ctx: commands.Context) -> List[str]:
-        urls = []
-        async for message in ctx.channel.history(limit=20):
-            if message.embeds and message.embeds[0].image:
-                urls.append(message.embeds[0].image.url)
-            if message.attachments:
-                urls.extend(
-                    match.group(1)
-                    for attachment in message.attachments
-                    if (match := IMAGE_LINKS.match(attachment.url))
-                )
-
-            if match := IMAGE_LINKS.search(message.content):
-                urls.append(match.group(1))
-        return urls
+async def find_images_in_replies(reference: discord.DeletedReferencedMessage | discord.Message | None) -> list[str]:
+    if not reference or not isinstance(reference, discord.Message):
+        return []
+    urls = []
+    argument = reference.system_content
+    if argument.startswith(DISCORD_CDN):
+        urls.append(argument.split()[0])
+    if match := IMAGE_LINKS.search(argument):
+        urls.append(match.group(1))
+    if reference.attachments:
+        urls.extend(
+            img.url for img in reference.attachments if img.content_type and img.content_type.startswith("image")
+        )
+    if reference.embeds and reference.embeds[0].image:
+        urls.append(reference.embeds[0].image.url)
+    return urls
+
+
+async def search_for_images(ctx: Context) -> list[str]:
+    urls = []
+    async for message in ctx.channel.history(limit=20):
+        if message.embeds and message.embeds[0].image:
+            urls.append(message.embeds[0].image.url)
+        if message.attachments:
+            urls.extend(
+                img.url for img in message.attachments if img.content_type and img.content_type.startswith("image")
+            )
+        if message.system_content.startswith(DISCORD_CDN):
+            urls.append(message.system_content.split()[0])
+        if match := IMAGE_LINKS.search(message.system_content):
+            urls.append(match.group(1))
+    return urls

ocr/info.json

@@ -1,12 +1,12 @@
 {
     "name": "OCR",
-    "author": ["ow0x", "TrustyJAID"],
+    "author": ["owocado", "TrustyJAID"],
     "short": "Detect text in images through OCR.",
     "description": "Detect text in images through OCR.",
     "install_msg": "**NOTE FOR BOT OWNER:**\n\nThis cog uses free ocr.space API which may give subpar results.\n\n**`[OPTIONAL]`**\nThere is optional support for Google Cloud Vision OCR API for improved text detection.\n(requires you to have a Google cloud project with active enabled billing account).\nYou may read more on that over at:\n<https://gist.github.com/ow0x/a7f17deaea4612a7dba4ba707210f7d8>",
     "end_user_data_statement": "This cog does not persistently store any data or metadata about users.",
     "tags": ["ocr", "image to text"],
-    "min_bot_version": "3.4.12",
+    "min_bot_version": "3.6.0",
     "hidden": false,
     "disabled": false,
     "type": "COG"

ocr/iso639.py

@@ -0,0 +1,147 @@
+from typing import Mapping
+
+
+ISO639_MAP: Mapping[str, str] = {
+    "af": "Afrikaans",
+    "ak": "Twi (Akan)",
+    "am": "Amharic",
+    "ar": "Arabic",
+    "as": "Assamese",
+    "ay": "Aymara",
+    "az": "Azerbaijani",
+    "be": "Belarusian",
+    "bg": "Bulgarian",
+    "bho": "Bhojpuri",
+    "bm": "Bambara",
+    "bn": "Bengali",
+    "bs": "Bosnian",
+    "ca": "Catalan",
+    "ceb": "Cebuano",
+    "ckb": "Kurdish (Sorani)",
+    "co": "Corsican",
+    "cs": "Czech",
+    "cy": "Welsh",
+    "da": "Danish",
+    "de": "German",
+    "doi": "Dogri",
+    "dv": "Divehi",
+    "ee": "Ewe",
+    "el": "Greek",
+    "en": "English",
+    "eo": "Esperanto",
+    "es": "Spanish",
+    "et": "Estonian",
+    "eu": "Basque",
+    "fa": "Persian",
+    "fi": "Finnish",
+    "fo": "Faroese",
+    "fr": "French",
+    "fy": "Frisian",
+    "ga": "Irish Gaelic",
+    "gd": "Scottish Gaelic",
+    "gl": "Galician",
+    "gn": "Guarani",
+    "gom": "Konkani",
+    "gu": "Gujarati",
+    "ha": "Hausa",
+    "haw": "Hawaiian",
+    "he": "Hebrew",
+    "hi": "Hindi",
+    "hmn": "Hmong",
+    "hr": "Croatian",
+    "ht": "Haitian Creole",
+    "hu": "Hungarian",
+    "hy": "Armenian",
+    "id": "Indonesian",
+    "ig": "Igbo",
+    "ilo": "Ilocano",
+    "is": "Icelandic",
+    "it": "Italian",
+    "iw": "Hebrew",
+    "ja": "Japanese",
+    "jv": "Javanese",
+    "jw": "Javanese",
+    "ka": "Georgian",
+    "kk": "Kazakh",
+    "kl": "Kalaallisut",
+    "km": "Khmer",
+    "kn": "Kannada",
+    "ko": "Korean",
+    "kri": "Krio",
+    "ku": "Kurdish (Kurmanji)",
+    "ky": "Kyrgyz",
+    "la": "Latin",
+    "lb": "Luxembourgish",
+    "lg": "Luganda",
+    "ln": "Lingala",
+    "lo": "Lao",
+    "lt": "Lithuanian",
+    "lus": "Mizo",
+    "lv": "Latvian",
+    "mai": "Maithili",
+    "mg": "Malagasy",
+    "mi": "Māori",
+    "mk": "Macedonian",
+    "ml": "Malayalam",
+    "mn": "Mongolian",
+    "mni-Mtei": "Meitei (Manipuri)",
+    "mr": "Marathi",
+    "ms": "Malay",
+    "mt": "Maltese",
+    "my": "Burmese",
+    "ne": "Nepali",
+    "nl": "Dutch",
+    "no": "Norwegian",
+    "nso": "Sepedi (Northern Sotho)",
+    "ny": "Chichewa",
+    "om": "Oromo",
+    "or": "Odia (Oriya)",
+    "pa": "Punjabi",
+    "pl": "Polish",
+    "ps": "Pashto",
+    "pt": "Portuguese",
+    "qu": "Quechua",
+    "ro": "Romanian",
+    "ru": "Russian",
+    "rw": "Kinyarwanda",
+    "sa": "Sanskrit",
+    "sd": "Sindhi",
+    "si": "Sinhalese",
+    "sk": "Slovak",
+    "sl": "Slovenian",
+    "sm": "Samoan",
+    "sn": "Shona",
+    "so": "Somali",
+    "sq": "Albanian",
+    "sr": "Serbian",
+    "st": "Sesotho",
+    "su": "Sundanese",
+    "sv": "Swedish",
+    "sw": "Swahili",
+    "ta": "Tamil",
+    "te": "Telugu",
+    "tg": "Tajik",
+    "th": "Thai",
+    "ti": "Tigrinya",
+    "tk": "Turkmen",
+    "tl": "Filipino",
+    "tr": "Turkish",
+    "ts": "Tsonga",
+    "tt": "Tatar",
+    "ug": "Uyghur",
+    "uk": "Ukrainian",
+    "ur": "Urdu",
+    "uz": "Uzbek",
+    "vi": "Vietnamese",
+    "xh": "Xhosa",
+    "yi": "Yiddish",
+    "yo": "Yoruba",
+    "zu": "Zulu",
+    "zh": "Chinese Simplified",
+    "zh-cn": "Chinese Simplified",
+    "zh-tw": "Chinese Traditional",
+    "zh-CN": "Chinese Simplified",
+    "zh-TW": "Chinese Traditional",
+    "auto": "Autodetect",
+}
+

ocr/models.py

@@ -0,0 +1,68 @@
+from dataclasses import dataclass, field
+from typing import List
+
+
+@dataclass(slots=True)
+class TextAnnotation:
+    locale: str = "en"
+    description: str = ""
+
+
+@dataclass(slots=True)
+class DetectedLanguage:
+    languageCode: str
+    confidence: float
+
+
+@dataclass(slots=True)
+class Property:
+    detectedLanguages: List[DetectedLanguage]
+
+
+@dataclass(slots=True)
+class Page:
+    width: int
+    height: int
+    confidence: float
+    property: Property | None = None
+
+
+@dataclass(slots=True)
+class FullTextAnnotation:
+    pages: List[Page] = field(default_factory=list)
+    text: str = ""
+
+    @property
+    def language_code(self) -> str:
+        if not self.pages:
+            return "auto"
+        if not self.pages[0].property:
+            return "auto"
+        if not self.pages[0].property.detectedLanguages:
+            return "auto"
+        return self.pages[0].property.detectedLanguages[0].languageCode
+
+
+@dataclass(slots=True)
+class VisionError:
+    code: int
+    message: str
+    status: str | None
+
+    def __str__(self) -> str:
+        return f"Error code: {self.code} ({self.message})"
+
+
+@dataclass(slots=True)
+class VisionPayload:
+    fullTextAnnotation: FullTextAnnotation | None
+    error: VisionError | None
+    textAnnotations: List[TextAnnotation] = field(default_factory=list)
+
+    @property
+    def text_value(self) -> str | None:
+        if not self.fullTextAnnotation:
+            if self.error:
+                return self.error.message
+            return None
+        return self.fullTextAnnotation.text or self.textAnnotations[0].description