mkv-episode-matcher 0.3.3__py3-none-any.whl → 1.0.0__py3-none-any.whl

mkv_episode_matcher/libraries/pgs2srt/Libraries/SubZero/post_processing.py

@@ -1,260 +0,0 @@
-# MIT License
-#
-# Copyright (c) 2018 Hannes Tismer
-#
-# Permission is hereby granted, free of charge, to any person obtaining a copy
-# of this software and associated documentation files (the "Software"), to deal
-# in the Software without restriction, including without limitation the rights
-# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
-# copies of the Software, and to permit persons to whom the Software is
-# furnished to do so, subject to the following conditions:
-#
-# The above copyright notice and this permission notice shall be included in all
-# copies or substantial portions of the Software.
-#
-# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
-# SOFTWARE.
-#
-#
-# Copyright for portions of project Sub-Zero are held by Bram Walet, 2014 as part of project Subliminal.bundle.
-# The original license is supplied below.
-#
-# The MIT License (MIT)
-#
-# Copyright (c) 2014 Bram Walet
-#
-# Permission is hereby granted, free of charge, to any person obtaining a copy
-# of this software and associated documentation files (the "Software"), to deal
-# in the Software without restriction, including without limitation the rights
-# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
-# copies of the Software, and to permit persons to whom the Software is
-# furnished to do so, subject to the following conditions:
-#
-# The above copyright notice and this permission notice shall be included in all
-# copies or substantial portions of the Software.
-#
-# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
-# SOFTWARE.
-
-
-import re
-
-from Libraries.SubZero.dictionaries.data import data
-from Libraries.SubZero.SubZero import (
-    MultipleLineProcessor,
-    MultipleWordReProcessor,
-    NReProcessor,
-    ReProcessor,
-    SubtitleTextModification,
-    WholeLineProcessor,
-)
-from tld import get_tld
-
-
-class CommonFixes(SubtitleTextModification):
-    identifier = "common"
-    description = "Basic common fixes"
-    exclusive = True
-    order = 40
-
-    long_description = "Fix common and whitespace/punctuation issues in subtitles"
-
-    processors = [
-        # normalize hyphens
-        NReProcessor(re.compile(r"(?u)([‑‐﹘﹣])"), "-", name="CM_hyphens"),
-        # -- = em dash
-        NReProcessor(
-            re.compile(r"(?u)(\w|\b|\s|^)(-\s?-{1,2})"), r"\1—", name="CM_multidash"
-        ),
-        # line = _/-/\s
-        NReProcessor(
-            re.compile(r'(?u)(^\W*[-_.:<>~"\']+\W*$)'), "", name="CM_non_word_only"
-        ),
-        # remove >>
-        NReProcessor(re.compile(r"(?u)^\s?>>\s*"), "", name="CM_leading_crocodiles"),
-        # line = : text
-        NReProcessor(
-            re.compile(r"(?u)(^\W*:\s*(?=\w+))"), "", name="CM_empty_colon_start"
-        ),
-        # fix music symbols
-        NReProcessor(
-            re.compile(r"(?u)(^[-\s>~]*[*#¶]+\s+)|(\s*[*#¶]+\s*$)"),
-            lambda x: "♪ " if x.group(1) else " ♪",
-            name="CM_music_symbols",
-        ),
-        # '' = "
-        NReProcessor(
-            re.compile(r"(?u)([\'’ʼ❜‘‛][\'’ʼ❜‘‛]+)"), '"', name="CM_double_apostrophe"
-        ),
-        # double quotes instead of single quotes inside words
-        NReProcessor(
-            re.compile(r'(?u)([A-zÀ-ž])"([A-zÀ-ž])'),
-            r"\1'\2",
-            name="CM_double_as_single",
-        ),
-        # normalize quotes
-        NReProcessor(
-            re.compile(r'(?u)(\s*["”“‟„])\s*(["”“‟„]["”“‟„\s]*)'),
-            lambda match: '"' + (" " if match.group(2).endswith(" ") else ""),
-            name="CM_normalize_quotes",
-        ),
-        # normalize single quotes
-        NReProcessor(re.compile(r"(?u)([\'’ʼ❜‘‛])"), "'", name="CM_normalize_squotes"),
-        # remove leading ...
-        NReProcessor(re.compile(r"(?u)^\.\.\.[\s]*"), "", name="CM_leading_ellipsis"),
-        # remove "downloaded from" tags
-        NReProcessor(re.compile(r"(?ui).+downloaded\s+from.+"), "", name="CM_crap"),
-        # no space after ellipsis
-        NReProcessor(
-            re.compile(r'(?u)\.\.\.(?![\s.,!?\'"])(?!$)'),
-            "... ",
-            name="CM_ellipsis_no_space",
-        ),
-        # no space before spaced ellipsis
-        NReProcessor(
-            re.compile(r"(?u)(?<=[^\s])(?<!\s)\. \. \."),
-            " . . .",
-            name="CM_ellipsis_no_space2",
-        ),
-        # multiple spaces
-        NReProcessor(re.compile(r"(?u)[\s]{2,}"), " ", name="CM_multiple_spaces"),
-        # more than 3 dots
-        NReProcessor(re.compile(r"(?u)\.{3,}"), "...", name="CM_dots"),
-        # no space after starting dash
-        NReProcessor(re.compile(r"(?u)^-(?![\s-])"), "- ", name="CM_dash_space"),
-        # remove starting spaced dots (not matching ellipses)
-        NReProcessor(
-            re.compile(r"(?u)^(?!\s?(\.\s\.\s\.)|(\s?\.{3}))(?=\.+\s+)[\s.]*"),
-            "",
-            name="CM_starting_spacedots",
-        ),
-        # space missing before doublequote
-        ReProcessor(
-            re.compile(r'(?u)(?<!^)(?<![\s(\["])("[^"]+")'),
-            r" \1",
-            name="CM_space_before_dblquote",
-        ),
-        # space missing after doublequote
-        ReProcessor(
-            re.compile(r'(?u)("[^"\s][^"]+")([^\s.,!?)\]]+)'),
-            r"\1 \2",
-            name="CM_space_after_dblquote",
-        ),
-        # space before ending doublequote?
-        # replace uppercase I with lowercase L in words
-        NReProcessor(
-            re.compile(r"(?u)([a-zà-ž]+)(I+)"),
-            lambda match: r"{}{}".format(match.group(1), "l" * len(match.group(2))),
-            name="CM_uppercase_i_in_word",
-        ),
-        # fix spaces in numbers (allows for punctuation: ,.:' (comma/dot only fixed if after space, those may be
-        # countdowns otherwise); don't break up ellipses
-        NReProcessor(
-            re.compile(
-                r"(?u)(\b[0-9]+[0-9:\']*(?<!\.\.)\s+(?!\.\.)[0-9,.:\'\s]*(?=[0-9]+)[0-9,.:\'])"
-            ),
-            lambda match: match.group(1).replace(" ", "")
-            if match.group(1).count(" ") == 1
-            else match.group(1),
-            name="CM_spaces_in_numbers",
-        ),
-        # uppercase after dot
-        # NReProcessor(re.compile(r'(?u)((?<!(?=\s*[A-ZÀ-Ž-_0-9.]\s*))(?:[^.\s])+\.\s+)([a-zà-ž])'),
-        #              lambda match: r'%s%s' % (match.group(1), match.group(2).upper()), name="CM_uppercase_after_dot"),
-        # remove double interpunction
-        NReProcessor(
-            re.compile(r"(?u)(\s*[,!?])\s*([,.!?][,.!?\s]*)"),
-            lambda match: match.group(1).strip()
-            + (" " if match.group(2).endswith(" ") else ""),
-            name="CM_double_interpunct",
-        ),
-        # remove spaces before punctuation; don't break spaced ellipses
-        NReProcessor(
-            re.compile(r"(?u)(?:(?<=^)|(?<=\w)) +([!?.,](?![!?.,]| \.))"),
-            r"\1",
-            name="CM_punctuation_space",
-        ),
-        # add space after punctuation
-        NReProcessor(
-            re.compile(r"(?u)(([^\s]*)([!?.,:])([A-zÀ-ž]{2,}))"),
-            lambda match: f"{match.group(2)}{match.group(3)} {match.group(4)}"
-            if not get_tld(match.group(1), fail_silently=True, fix_protocol=True)
-            else match.group(1),
-            name="CM_punctuation_space2",
-        ),
-        # fix lowercase I in english
-        NReProcessor(
-            re.compile(r"(?u)(\b)i(\b)"),
-            r"\1I\2",
-            name="CM_EN_lowercase_i",
-            # supported=lambda p: p.language == ENGLISH),
-        ),
-    ]
-
-
-class FixOCR(SubtitleTextModification):
-    identifier = "OCR_fixes"
-    description = "Fix common OCR issues"
-    exclusive = True
-    order = 10
-    data_dict = None
-
-    long_description = "Fix issues that happen when a subtitle gets converted from bitmap to text through OCR"
-
-    def __init__(self, language):
-        super(FixOCR, self).__init__()
-        data_dict = data.get(language)
-        if not data_dict:
-            # logger.debug("No SnR-data available for language %s", parent.language)
-            return
-
-        self.data_dict = data_dict
-        self.processors = self.get_processors()
-
-    def get_processors(self):
-        if not self.data_dict:
-            return []
-
-        return [
-            # remove broken HI tag colons (ANNOUNCER'., ". instead of :) after at least 3 uppercase chars
-            # don't modify stuff inside quotes
-            NReProcessor(
-                re.compile(
-                    r'(?u)(^[^"\'’ʼ❜‘‛”“‟„]*(?<=[A-ZÀ-Ž]{3})[A-ZÀ-Ž-_\s0-9]+)'
-                    r'(["\'’ʼ❜‘‛”“‟„]*[.,‚،⹁、;]+)(\s*)(?!["\'’ʼ❜‘‛”“‟„])'
-                ),
-                r"\1:\3",
-                name="OCR_fix_HI_colons",
-            ),
-            # fix F'bla
-            NReProcessor(
-                re.compile(r"(?u)(\bF)(\')([A-zÀ-ž]*\b)"), r"\1\3", name="OCR_fix_F"
-            ),
-            WholeLineProcessor(self.data_dict["WholeLines"], name="OCR_replace_line"),
-            MultipleWordReProcessor(
-                self.data_dict["WholeWords"], name="OCR_replace_word"
-            ),
-            MultipleWordReProcessor(
-                self.data_dict["BeginLines"], name="OCR_replace_beginline"
-            ),
-            MultipleWordReProcessor(
-                self.data_dict["EndLines"], name="OCR_replace_endline"
-            ),
-            MultipleWordReProcessor(
-                self.data_dict["PartialLines"], name="OCR_replace_partialline"
-            ),
-            MultipleLineProcessor(
-                self.data_dict["PartialWordsAlways"],
-                name="OCR_replace_partialwordsalways",
-            ),
-        ]

mkv_episode_matcher/libraries/pgs2srt/README.md

@@ -1,26 +0,0 @@
-# pgs2srt
-
-Uses [pgsreader](https://github.com/EzraBC/pgsreader) and [pyteseract](https://pypi.org/project/pytesseract/) to convert image based pgs subtitles files (.sup) to text based subrip (.srt) files.
-
-## Requirements
-Python3, pip3, and Tesseract
-
-## Installation
-* Run ```git clone https://github.com/PimvanderLoos/pgs2srt.git```
-* Inside the repo folder, run ```pip3 install -r requirements.txt```
-* In your .bashrc or .zshrc add ```alias pgs2srt='<absolute path to repo>/pgs2srt.py'```
-
-## How to run
-
-    pgs2srt <pgs filename>.sup
-
-## Improving accuracy
-On Debian and Ubuntu, the default trained models files for Tesseract are from the [fast](https://github.com/tesseract-ocr/tessdata_fast) set. While these are a bit faster than other options, this comes at the cost of accuracy. If you want higher accuracy, I'd recommend using either the [legacy](https://github.com/tesseract-ocr/tessdata) or the [best](https://github.com/tesseract-ocr/tessdata_best) trained models. Note that the fast and best options only support the LSTM OCR Engine Mode (oem 1).
-
-## Caveats
-
-This is in no way a perfect converter, and tesseract will make incorrect interpretations of characters. Extremely alpha, issues, pull requests and suggestions welcome!
-
-
-## Credits
-This project uses the common + OCR fixes developed by [Sub-Zero.bundle](https://github.com/pannal/Sub-Zero.bundle).

mkv_episode_matcher/libraries/pgs2srt/imagemaker.py

@@ -1,89 +0,0 @@
-import numpy as np
-from PIL import Image
-
-
-def read_rle_bytes(ods_bytes):
-    pixels = []
-    line_builder = []
-
-    i = 0
-    while i < len(ods_bytes):
-        if ods_bytes[i]:
-            incr = 1
-            color = ods_bytes[i]
-            length = 1
-        else:
-            check = ods_bytes[i + 1]
-            if check == 0:
-                incr = 2
-                color = 0
-                length = 0
-                pixels.append(line_builder)
-                line_builder = []
-            elif check < 64:
-                incr = 2
-                color = 0
-                length = check
-            elif check < 128:
-                incr = 3
-                color = 0
-                length = ((check - 64) << 8) + ods_bytes[i + 2]
-            elif check < 192:
-                incr = 3
-                color = ods_bytes[i + 2]
-                length = check - 128
-            else:
-                incr = 4
-                color = ods_bytes[i + 3]
-                length = ((check - 192) << 8) + ods_bytes[i + 2]
-        line_builder.extend([color] * length)
-        i += incr
-
-    if line_builder:
-        print(f"Probably an error; hanging pixels: {line_builder}")
-
-    return pixels
-
-
-def ycbcr2rgb(ar):
-    xform = np.array([[1, 0, 1.402], [1, -0.34414, -0.71414], [1, 1.772, 0]])
-    rgb = ar.astype(float)
-    # Subtracting by 128 the R and G channels
-    rgb[:, [1, 2]] -= 128
-    # .dot is multiplication of the matrices and xform.T is a transpose of the array axes
-    rgb = rgb.dot(xform.T)
-    # Makes any pixel value greater than 255 just be 255 (Max for RGB colorspace)
-    np.putmask(rgb, rgb > 255, 255)
-    # Sets any pixel value less than 0 to 0 (Min for RGB colorspace)
-    np.putmask(rgb, rgb < 0, 0)
-    return np.uint8(rgb)
-
-
-def px_rgb_a(ods, pds, swap):
-    px = read_rle_bytes(ods.img_data)
-    px = np.array([[255] * (ods.width - len(l)) + l for l in px], dtype=np.uint8)
-
-    # Extract the YCbCrA palette data, swapping channels if requested.
-    if swap:
-        ycbcr = np.array([(entry.Y, entry.Cb, entry.Cr) for entry in pds.palette])
-    else:
-        ycbcr = np.array([(entry.Y, entry.Cr, entry.Cb) for entry in pds.palette])
-    try:
-        rgb = ycbcr2rgb(ycbcr)
-    except AttributeError:
-        print("Error: The image is not in YCbCr format.")
-        exit(1)
-    # Separate the Alpha channel from the YCbCr palette data
-    a = [entry.Alpha for entry in pds.palette]
-    a = np.array([[a[x] for x in l] for l in px], dtype=np.uint8)
-
-    return px, rgb, a
-
-
-def make_image(ods, pds, swap=False):
-    px, rgb, a = px_rgb_a(ods, pds, swap)
-    alpha = Image.fromarray(a, mode="L")
-    img = Image.fromarray(px, mode="P")
-    img.putalpha(alpha)
-    img.putpalette(rgb)
-    return img

mkv_episode_matcher/libraries/pgs2srt/pgs2srt.py

@@ -1,150 +0,0 @@
-#!/usr/bin/env python3
-
-import argparse
-import re
-from datetime import datetime, timedelta
-
-import pytesseract
-from imagemaker import make_image
-from Libraries.SubZero.post_processing import CommonFixes, FixOCR
-from pgsreader import PGSReader
-from PIL import Image, ImageOps
-
-parser = argparse.ArgumentParser(description="Convert PGS subtitles to SubRip format.")
-
-parser.add_argument("input", type=str, help="The input file (a .sup file).")
-parser.add_argument("--output", type=str, help="The output file (a .srt file).")
-parser.add_argument(
-    "--oem",
-    type=int,
-    help="The OCR Engine Mode to use (Default: 1).",
-    default=1,
-    choices=range(4),
-)
-parser.add_argument(
-    "--language", type=str, help="The language to use (Default: eng).", default="eng"
-)
-parser.add_argument(
-    "--fix_common",
-    help="Fixes common whitespace/punctuation issues.",
-    dest="fix_common",
-    action="store_true",
-)
-parser.add_argument(
-    "--fix_common_ocr",
-    help="Fixes common OCR issues for supported languages.",
-    dest="fix_ocr",
-    action="store_true",
-)
-
-args = parser.parse_args()
-
-assert args.input is not None
-
-# Unescape escaped spaces
-file = args.input.replace("\\ ", " ")
-
-print(f"Parsing: {file}")
-
-# Load a PGS/SUP file.
-pgs = PGSReader(file)
-
-# Set index
-i = 0
-
-# Complete subtitle track index
-si = 0
-
-tesseract_lang = args.language
-tesseract_config = f"-c tessedit_char_blacklist=[] --psm 6 --oem {args.oem}"
-
-# If an output file for the subrip output is provided, use that.
-# Otherwise remove the ".sup" extension from the input and append
-# ".srt".
-output_file = (
-    args.output
-    if args.output is not None
-    else (args.input.replace(".sup", "") + ".srt")
-)
-
-# SubRip output
-output = ""
-
-fix_common = CommonFixes() if args.fix_common else None
-fix_ocr = FixOCR(args.language) if args.fix_ocr else None
-
-# Iterate the pgs generator
-for ds in pgs.iter_displaysets():
-    try:
-        # If set has image, parse the image
-        if ds.has_image:
-            # Get Palette Display Segment
-            pds = ds.pds[0]
-            # Get Object Display Segment
-            ods = ds.ods[0]
-
-            if pds and ods:
-                # Create and show the bitmap image and convert it to RGBA
-                src = make_image(ods, pds).convert("RGBA")
-
-                # Create grayscale image with black background
-                img = Image.new("L", src.size, "BLACK")
-                # Paste the subtitle bitmap
-                img.paste(src, (0, 0), src)
-                # Invert images so the text is readable by Tesseract
-                img = ImageOps.invert(img)
-
-                # Parse the image with tesesract
-                text = pytesseract.image_to_string(
-                    img, lang=tesseract_lang, config=tesseract_config
-                ).strip()
-
-                # Replace "|" with "I"
-                # Works better than blacklisting "|" in Tesseract,
-                # which results in I becoming "!" "i" and "1"
-                text = re.sub(r"[|/\\]", "I", text)
-                text = re.sub(r"[_]", "L", text)
-
-                if args.fix_common:
-                    text = fix_common.process(text)
-                if args.fix_ocr:
-                    text = fix_ocr.modify(text)
-
-                start = datetime.fromtimestamp(ods.presentation_timestamp / 1000)
-                start = start + timedelta(hours=-1)
-
-        else:
-            # Get Presentation Composition Segment
-            pcs = ds.pcs[0]
-
-            if pcs:
-                end = datetime.fromtimestamp(pcs.presentation_timestamp / 1000)
-                end = end + timedelta(hours=-1)
-
-                if (
-                    isinstance(start, datetime)
-                    and isinstance(end, datetime)
-                    and len(text)
-                ):
-                    si = si + 1
-                    sub_output = str(si) + "\n"
-                    sub_output += (
-                        start.strftime("%H:%M:%S,%f")[0:12]
-                        + " --> "
-                        + end.strftime("%H:%M:%S,%f")[0:12]
-                        + "\n"
-                    )
-                    sub_output += text + "\n\n"
-
-                    output += sub_output
-                    start = end = text = None
-        i = i + 1
-
-    except Exception as e:
-        print(e)
-        exit(1)
-
-f = open(output_file, "w")
-f.write(output)
-f.close()
-print(f"Saved to: {output_file}")