chore: split cli and functions

Split the functions out to __init__.py and introduce a class for the
application, improving separation of concerns.

This allows for users to simply import our module and call
the functions as they desire, with the CLI being optional.

Additionally, if someone doesn't like the way that we handle
configuration, they can import our module instead of forking the
project.

This also makes the code slightly more readable, as well as reducing the
number of variables that are globally-scoped in the file.

hydrus_ocr/__init__.py

@@ -0,0 +1,106 @@
+from io import BytesIO
+
+# manipulate images
+from PIL import Image, ImageFilter
+import pyocr
+
+from hydrus_api import Client
+
+from hydrus_ocr.errors import MissingToolError
+
+# This is a list of all filetypes supported by both PIL and Hydrus
+valid_file_types = [
+        "image/jpeg",
+        "image/png",
+        "image/webp",
+        "image/tiff",
+        "image/qoi",
+        "image/x-icon",
+        "image/bmp",
+        "image/heif",
+        "image/heic",
+        "image/avif"
+        ]
+
+
+class HydrusOCR():
+    def __init__(self, api_url, access_key, tag_service_key, ocr_language) -> None:
+        self.api_url: str = api_url
+        self.access_key: str = access_key
+        self.hydrus_client: Client = Client(access_key)
+        self.tag_service_key: str = tag_service_key
+        self.ocr_language: str = ocr_language
+
+    def remove_ocr_service_tag(self, file_id: int, tag: str) -> None:
+        """
+        Removes a tag from an item in the program's tag service.
+        """
+        self.hydrus_client.add_tags(
+            file_ids=[file_id],
+            service_keys_to_actions_to_tags={self.tag_service_key: {1: [tag]}}
+            )
+
+    def add_ocr_service_tag(self, file_id: int, tag: str) -> None:
+        """
+        Adds a tag to an item inside of the program's tag service.
+        """
+        self.hydrus_client.add_tags(
+                file_ids=[file_id],
+                service_keys_to_tags={self.tag_service_key: [tag]}
+                )
+
+    def find_images(self):
+        """
+        Finds all images with the 'ocr wanted' tag set in our tag service.
+        """
+        tags = ["ocr wanted", "-ocr rejected"]
+        return self.hydrus_client.search_files(tags,
+                                               tag_service_key=self.tag_service_key)
+
+    def get_image(self, file_id: int) -> Image:
+        """
+        Retrieves an image from Hydrus and converts it to an Image object.
+        """
+        file = self.hydrus_client.get_file(file_id=file_id)
+        if file.headers["Content-Type"] in valid_file_types:
+            return Image.open(BytesIO(file.content))
+        else:
+            self.remove_ocr_service_tag(file_id, "ocr wanted")
+            self.add_ocr_service_tag(file_id, "ocr rejected")
+            return None
+
+    def ocr_image(self, image: Image) -> str:
+        """
+        OCR the image file using Tesseract.
+        """
+        image.filter(ImageFilter.SHARPEN)
+        tool = pyocr.get_available_tools()[0]
+        if not tool:
+            raise MissingToolError(MissingToolError.__doc__)
+
+        return tool.image_to_string(
+                image,
+                lang=self.ocr_language,
+                builder=pyocr.builders.TextBuilder()
+                )
+
+    def write_ocr_to_hydrus(self, file_id: int, text: str) -> None:
+        """
+        Sets the 'ocr' note in a Hydrus file to the 'text' parameter,
+        and mark OCR as completed.
+        """
+        notes = {'ocr': text}
+        self.hydrus_client.set_notes(notes, file_id=file_id)
+        self.add_ocr_service_tag(file_id, "ocr completed")
+        self.remove_ocr_service_tag(file_id, "ocr wanted")
+
+    def process_image(self, file_id: int) -> None:
+        """
+        Run all of the processing steps on a specific image.
+        """
+        image = self.get_image(file_id)
+        if image:
+            text = self.ocr_image(image)
+            self.write_ocr_to_hydrus(file_id, text)
+        else:
+            pass

hydrus_ocr/__main__.py

@@ -5,22 +5,11 @@
 """
 import os
 import sys
-import io
 import threading
 from time import sleep
 
-# interact with the Hydrus client API
-import hydrus_api as hydrus
-
-# manipulate images
-from PIL import Image, ImageFilter
-import pyocr
-
-from hydrus_ocr.errors import (
-        MissingTokenError,
-        NoSubcommandError,
-        MissingToolError
-        )
+from . import HydrusOCR
+from .errors import MissingTokenError, NoSubcommandError
 
 # Configuration
 hydrus_api_url: str = os.getenv("HYDRUS_OCR_API_URL", "http://localhost:45869")
@@ -45,112 +34,16 @@
 else:
     raise NoSubcommandError(NoSubcommandError.__doc__)
 
-# This is a list of all filetypes supported by both PIL and Hydrus
-valid_file_types = [
-        "image/jpeg",
-        "image/png",
-        "image/webp",
-        "image/tiff",
-        "image/qoi",
-        "image/x-icon",
-        "image/bmp",
-        "image/heif",
-        "image/heic",
-        "image/avif"
-        ]
-
-# Initialize the client with values from configuration
-hydrus_client: hydrus.Client = hydrus.Client(access_key, hydrus_api_url)
-
-
-def remove_ocr_service_tag(file_id: int, tag: str) -> None:
-    """
-    Removes a tag from an item in the program's tag service.
-    """
-    hydrus_client.add_tags(
-            file_ids=[file_id],
-            service_keys_to_actions_to_tags={tag_service_key: {1: [tag]}}
-            )
-
-
-def add_ocr_service_tag(file_id: int, tag: str) -> None:
-    """
-    Adds a tag to an item inside of the program's tag service.
-    """
-    hydrus_client.add_tags(
-            file_ids=[file_id],
-            service_keys_to_tags={tag_service_key: [tag]}
-            )
-
-
-def find_images():
-    """
-    Finds all images with the 'ocr wanted' tag set in our tag service.
-    """
-    tags = ["ocr wanted", "-ocr rejected"]
-    return hydrus_client.search_files(tags, tag_service_key=tag_service_key)
-
-
-def get_image(file_id: int) -> Image:
-    """
-    Retrieves an image from Hydrus and converts it to an Image object.
-    """
-    file = hydrus_client.get_file(file_id=file_id)
-    if file.headers["Content-Type"] in valid_file_types:
-        return Image.open(io.BytesIO(file.content))
-    else:
-        remove_ocr_service_tag(file_id, "ocr wanted")
-        add_ocr_service_tag(file_id, "ocr rejected")
-        return None
-
-
-def ocr_image(image: Image) -> str:
-    """
-    OCR the image file using Tesseract.
-    """
-    image.filter(ImageFilter.SHARPEN)
-    tool = pyocr.get_available_tools()[0]
-    if not tool:
-        raise MissingToolError(MissingToolError.__doc__)
-
-    return tool.image_to_string(
-            image,
-            lang=ocr_language,
-            builder=pyocr.builders.TextBuilder()
-            )
-
-
-def write_ocr_to_hydrus(file_id: int, text: str) -> None:
-    """
-    Sets the 'ocr' note in a Hydrus file to the 'text' parameter,
-    and mark OCR as completed.
-    """
-    notes = {'ocr': text}
-    hydrus_client.set_notes(notes, file_id=file_id)
-    add_ocr_service_tag(file_id, "ocr completed")
-    remove_ocr_service_tag(file_id, "ocr wanted")
-
-
-def process_image(file_id: int) -> None:
-    """
-    Run all of the processing steps on a specific image.
-    """
-    image = get_image(file_id)
-    if image:
-        text = ocr_image(image)
-        write_ocr_to_hydrus(file_id, text)
-    else:
-        pass
-
+ocr_client = HydrusOCR(hydrus_api_url, access_key, tag_service_key, ocr_language)
 
 def mainloop():
     """
-    Run the program using the functions defined above.
+    Run the program using the HydrusOCR class.
     """
     run = True
     while run:
-        for i in find_images()["file_ids"]:
-            threading.Thread(target=process_image, args=(i,)).start()
+        for i in ocr_client.find_images()["file_ids"]:
+            threading.Thread(target=ocr_client.process_image, args=(i,)).start()
 
         # If we're not in daemon mode, exit immediately after first run
         if not daemon: