pycaption 2.3.10__tar.gz → 2.3.13__tar.gz

{pycaption-2.3.10/pycaption.egg-info → pycaption-2.3.13}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pycaption
-Version: 2.3.10
+Version: 2.3.13
 Summary: Closed caption converter
 Author: Joe Norton
 Author-email: joey@nortoncrew.com

{pycaption-2.3.10 → pycaption-2.3.13}/pycaption/__init__.py

@@ -21,6 +21,7 @@ from .srt import SRTReader, SRTWriter
 from .transcript import TranscriptWriter
 from .webvtt import WebVTTReader, WebVTTWriter
 
+
 __all__ = [
     "CaptionConverter",
     "DFXPReader",

{pycaption-2.3.10 → pycaption-2.3.13}/pycaption/base.py

@@ -7,11 +7,13 @@ CaptionConverter orchestrator and base classes for readers/writers.
 
 import logging
 import os
+
 from datetime import timedelta
 from numbers import Number
 
 from .exceptions import CaptionReadError, CaptionReadTimingError, InvalidInputError
 
+
 logger = logging.getLogger(__name__)
 
 # `und` a special identifier for an undetermined language according to ISO 639-2
@@ -40,8 +42,7 @@ class CaptionConverter:
         """
         if not hasattr(caption_reader, "read"):
             raise InvalidInputError(
-                "The caption_reader must be a BaseReader instance "
-                "with a read() method."
+                "The caption_reader must be a BaseReader instance with a read() method."
             )
         self.captions = caption_reader.read(content)
         return self

{pycaption-2.3.10 → pycaption-2.3.13}/pycaption/dfxp/constants.py

@@ -8,6 +8,7 @@ import re
 
 from ..geometry import Alignment, HorizontalAlignmentEnum, Layout, VerticalAlignmentEnum
 
+
 DFXP_BASE_MARKUP = """
 <tt xmlns="http://www.w3.org/ns/ttml"
     xmlns:tts="http://www.w3.org/ns/ttml#styling">

{pycaption-2.3.10 → pycaption-2.3.13}/pycaption/dfxp/extras.py

@@ -10,9 +10,11 @@ from xml.sax.saxutils import escape
 from bs4 import BeautifulSoup
 
 from ..base import BaseWriter, CaptionNode, merge_concurrent_captions
+from ..utils import encode_leading_nbsp
 from .constants import DFXP_DEFAULT_REGION
 from .writer import DFXPWriter
 
+
 LEGACY_DFXP_BASE_MARKUP = """
 <tt xmlns="http://www.w3.org/ns/ttml"
     xmlns:tts="http://www.w3.org/ns/ttml#styling">
@@ -203,7 +205,7 @@ class LegacyDFXPWriter(BaseWriter):
         start = caption.format_start()
         end = caption.format_end()
         p = dfxp.new_tag("p", begin=start, end=end)
-        p.string = self._recreate_text(caption, dfxp)
+        p.string = encode_leading_nbsp(self._recreate_text(caption, dfxp))
 
         if dfxp.find("style", {"xml:id": "p"}):
             p["style"] = "p"
@@ -255,7 +257,7 @@ class LegacyDFXPWriter(BaseWriter):
 
         return line
 
-    def _recreate_style(self, content, dfxp):
+    def _recreate_style(self, content, dfxp):  # noqa: C901
         """Convert an internal style dict to DFXP/TTS attributes.
 
         :type content: dict

{pycaption-2.3.10 → pycaption-2.3.13}/pycaption/dfxp/reader.py

@@ -47,7 +47,9 @@ from .constants import (
     VERTICAL_ALIGNMENT_TO_DFXP,
 )
 
-_LEADING_WHITESPACE_RE = re.compile("^(?:[\n\r]+\\s*)?(.+)")
+
+# Pretty-print indentation only: \s would also eat U+00A0 line indentation.
+_LEADING_WHITESPACE_RE = re.compile("^(?:[\n\r]+[ \t\n\r]*)?(.+)")
 
 _DFXP_WRITING_MODE_MAP = {
     "tbrl": WritingDirectionEnum.VERTICAL_RL,
@@ -147,7 +149,8 @@ class DFXPReader(BaseReader):
                     style_dict[id_] = self._convert_style(style)
 
         caption_set = CaptionSet(
-            caption_dict, styles=style_dict,
+            caption_dict,
+            styles=style_dict,
             visual_alignment_default=HorizontalAlignmentEnum.START,
         )
 
@@ -179,8 +182,7 @@ class DFXPReader(BaseReader):
             tickrate = float(tt_attrs["ttp:tickrate"])
         except ValueError:
             raise CaptionReadSyntaxError(
-                f"ttp:tickRate must be a number, "
-                f"got '{tt_attrs['ttp:tickrate']}'"
+                f"ttp:tickRate must be a number, got '{tt_attrs['ttp:tickrate']}'"
             )
         if tickrate <= 0:
             raise CaptionReadSyntaxError(
@@ -203,8 +205,7 @@ class DFXPReader(BaseReader):
             framerate_int = int(framerate_str)
         except ValueError:
             raise CaptionReadSyntaxError(
-                f"ttp:frameRate must be a positive integer, "
-                f"got '{framerate_str}'"
+                f"ttp:frameRate must be a positive integer, got '{framerate_str}'"
             )
         self.tickrate = float(framerate_int * sub_framerate)
 
@@ -396,7 +397,7 @@ class DFXPReader(BaseReader):
                 self._convert_tag_to_node(a)
 
     @staticmethod
-    def _convert_style(tag):
+    def _convert_style(tag):  # noqa: C901
         """Convert DFXP/TTS style attributes on a tag to an internal style dict.
 
         Maps tts:fontStyle, tts:fontWeight, tts:textDecoration, tts:textAlign,

{pycaption-2.3.10 → pycaption-2.3.13}/pycaption/dfxp/writer.py

@@ -5,6 +5,7 @@ including style elements, region-based positioning, and writing direction.
 """
 
 import re
+
 from copy import deepcopy
 from xml.sax.saxutils import escape
 
@@ -12,6 +13,7 @@ from bs4 import BeautifulSoup
 
 from ..base import BaseWriter, CaptionNode
 from ..geometry import HorizontalAlignmentEnum, LineAlignmentEnum, WritingDirectionEnum
+from ..utils import encode_leading_nbsp
 from .constants import (
     DFXP_ATTR_XML_ID,
     DFXP_ATTR_XML_LANG,
@@ -26,6 +28,7 @@ from .constants import (
     _create_external_alignment,
 )
 
+
 _WRITING_DIRECTION_TO_DFXP = {
     WritingDirectionEnum.VERTICAL_RL: "tbrl",
     WritingDirectionEnum.VERTICAL_LR: "tblr",
@@ -204,7 +207,9 @@ class DFXPWriter(BaseWriter):
         start = caption.format_start()
         end = caption.format_end()
         p = dfxp.new_tag("p", begin=start, end=end)
-        p.string = self._recreate_text(caption, dfxp, caption_set, lang)
+        p.string = encode_leading_nbsp(
+            self._recreate_text(caption, dfxp, caption_set, lang)
+        )
 
         if dfxp.find("style", {DFXP_ATTR_XML_ID: "p"}):
             p["style"] = "p"
@@ -213,7 +218,7 @@ class DFXPWriter(BaseWriter):
 
         return p
 
-    def _recreate_text(self, caption, dfxp, caption_set=None, lang=None):
+    def _recreate_text(self, caption, dfxp, caption_set=None, lang=None):  # noqa: C901
         """Serialize all nodes of a caption into DFXP inline markup.
 
         Handles text nodes, line breaks (<br/>), and style spans.  Nodes are
@@ -756,7 +761,7 @@ def _text_shadow_to_outline(value):
     return thickness
 
 
-def _recreate_style(content, dfxp):
+def _recreate_style(content, dfxp):  # noqa: C901
     """Convert an internal style dict to DFXP/TTS style attributes.
 
     Maps pycaption's internal keys (class, italics, bold, underline, color,
@@ -807,7 +812,7 @@ def _recreate_style(content, dfxp):
     return dfxp_style
 
 
-def _convert_layout_to_attributes(layout, fallback_alignment=None):
+def _convert_layout_to_attributes(layout, fallback_alignment=None):  # noqa: C901
     """Convert a Layout object to a dict of DFXP region attributes.
 
     Maps origin, extent, padding, alignment, and writing_direction to their

{pycaption-2.3.10 → pycaption-2.3.13}/pycaption/geometry.py

@@ -7,12 +7,15 @@ CONVENTIONS:
   responsible for the recalculation should return a new object with the
   necessary modifications.
 """
+
 import re
+
 from enum import Enum
 from functools import total_ordering
 
 from .exceptions import CaptionReadSyntaxError, RelativizationError
 
+
 _UNIT_MISMATCH_MSG = "The sizes should have the same measure units."
 
 

{pycaption-2.3.10 → pycaption-2.3.13}/pycaption/microdvd.py

@@ -5,6 +5,7 @@ A special line {0}{0}fps_value sets the framerate (default 25 fps).
 """
 
 import re
+
 from copy import deepcopy
 
 from .base import (

{pycaption-2.3.10 → pycaption-2.3.13}/pycaption/sami/constants.py

@@ -10,6 +10,7 @@ from cssutils import log
 
 from ..geometry import HorizontalAlignmentEnum
 
+
 log.setLevel(FATAL)
 
 SAMI_BASE_MARKUP = """

{pycaption-2.3.10 → pycaption-2.3.13}/pycaption/sami/reader.py

@@ -4,6 +4,8 @@ Parses SAMI documents into pycaption CaptionSet objects, handling CSS
 styles, margin-based positioning, and multi-language content.
 """
 
+import re
+
 from bs4 import BeautifulSoup, NavigableString
 
 from ..base import BaseReader, Caption, CaptionList, CaptionNode, CaptionSet
@@ -11,8 +13,12 @@ from ..exceptions import CaptionReadNoCaptions, CaptionReadTimingError
 from ..geometry import Alignment, HorizontalAlignmentEnum, Layout, Padding, Size
 from .parser import SAMIParser
 
+
 _TAG_TO_STYLE = {"i": "italics", "b": "bold", "u": "underline"}
 
+# A line break plus the pretty-print indentation of the closing tag.
+_TRAILING_WHITESPACE_RE = re.compile("[\n\r][ \t\n\r]*$")
+
 
 class SAMIReader(BaseReader):
     """Reads SAMI caption files into a CaptionSet.
@@ -202,8 +208,9 @@ class SAMIReader(BaseReader):
         if isinstance(tag, NavigableString):
             tag_text = str(tag)
             if tag_text and tag_text[0] in "\n\r":
-                tag_text = tag_text.lstrip()
-            tag_text = tag_text.rstrip("\n\r")
+                # Not a bare lstrip(): U+00A0 is line indentation, not markup.
+                tag_text = tag_text.lstrip(" \t\n\r")
+            tag_text = _TRAILING_WHITESPACE_RE.sub("", tag_text)
             if not tag_text:
                 return
             self.line.append(CaptionNode.create_text(tag_text, inherit_from))

{pycaption-2.3.10 → pycaption-2.3.13}/pycaption/sami/writer.py

@@ -11,8 +11,10 @@ from bs4 import BeautifulSoup
 
 from ..base import BaseWriter, CaptionNode
 from ..geometry import HorizontalAlignmentEnum
+from ..utils import encode_leading_nbsp
 from .constants import HORIZONTAL_ALIGNMENT_MAP, SAMI_BASE_MARKUP
 
+
 _NON_CSS_KEYS = frozenset(
     {
         "classes",
@@ -103,7 +105,7 @@ class SAMIWriter(BaseWriter):
             p["style"] = p_style
 
         p["class"] = self._recreate_p_lang(caption, lang, captions)
-        p.string = self._recreate_text(caption.nodes)
+        p.string = encode_leading_nbsp(self._recreate_text(caption.nodes))
 
         sync.append(p)
 

{pycaption-2.3.10 → pycaption-2.3.13}/pycaption/scc/__init__.py

@@ -3,4 +3,5 @@
 from .reader import SCCReader
 from .writer import SCC_TOKENS_PER_CAPTION_MAX, SCCWriter
 
+
 __all__ = ["SCCReader", "SCCWriter", "SCC_TOKENS_PER_CAPTION_MAX"]

{pycaption-2.3.10 → pycaption-2.3.13}/pycaption/scc/constants.py

@@ -7,8 +7,10 @@ background color codes, and derived inverse lookup tables.
 """
 
 import re as _re
+
 from itertools import product
 
+
 COMMANDS = {
     "9420": "",
     "9429": "",
@@ -1317,6 +1319,9 @@ SPECIAL_OR_EXTENDED_CHAR_TO_CODE = {
 SPECIAL_OR_EXTENDED_CHAR_TO_CODE.update(
     {character: code for code, character in SPECIAL_CHARS.items()}
 )
+# A non-breaking space is what CEA-608's transparent space is: a blank column
+# that is not a break opportunity. Unmapped, it fell back to the pound sign.
+SPECIAL_OR_EXTENDED_CHAR_TO_CODE["\xa0"] = "91b9"
 
 # Time to transmit a single codeword = 1 second / 29.97
 MICROSECONDS_PER_CODEWORD = 1000.0 * 1000.0 / (30.0 * 1000.0 / 1001.0)

{pycaption-2.3.10 → pycaption-2.3.13}/pycaption/scc/reader.py

@@ -78,6 +78,7 @@ http://www.theneitherworld.com/mcpoodle/SCC_TOOLS/DOCS/SCC_FORMAT.HTML
 """
 
 import re
+
 from collections import deque
 from copy import deepcopy
 
@@ -108,8 +109,13 @@ from .specialized_collections import (
 )
 from .state_machines import DefaultProvidingPositionTracker
 
+
 _TIMECODE_RE = re.compile(r"\d{2}:\d{2}:\d{2}[:;](\d{1,2})")
 
+# Appended to a too-long line whose text kept going on the row with no PAC or
+# row change of its own. Matched downstream, so the wording is part of the API.
+ROW_CONTINUATION_SUFFIX = " (continues on the row without a row change)"
+
 
 class SCCReader(BaseReader):
     """Converts a given unicode string to a CaptionSet.
@@ -241,18 +247,43 @@ class SCCReader(BaseReader):
 
     def _validate_line_lengths(self):
         """Raise CaptionLineLengthError if any line exceeds 32 characters."""
-        violations = [
-            f"around {cap.format_start()} - {line} - Length {len(line)}"
-            for cap in (c.to_real_caption() for c in self.caption_stash._collection)
-            for line in cap.get_text().split("\n")
-            if len(line) > 32
-        ]
+        violations = []
+        for precap in self.caption_stash._collection:
+            cap = precap.to_real_caption()
+            raw = "".join(cap.get_text_nodes())
+            # get_text() strips the caption, dropping any lines it opens with
+            skipped = raw[: len(raw) - len(raw.lstrip())].count("\n")
+            for index, line in enumerate(cap.get_text().split("\n")):
+                if len(line) > 32:
+                    description = self._describe_line_length(
+                        line, (index + skipped) in precap.continued_rows
+                    )
+                    violations.append(f"around {cap.format_start()} - {description}")
         if violations:
             raise CaptionLineLengthError(
                 "32 character limit for caption cue in scc file.\n"
                 "Lines longer than 32:\n" + "\n".join(violations)
             )
 
+    @staticmethod
+    def _describe_line_length(line, continues_row=False):
+        """Describe a line too long for the row, separating the text from the
+        indentation that places it, which can be what pushes it off the row,
+        and naming a missing PAC or row change when that is what joined it.
+        """
+        text = line.lstrip("\xa0")
+        indent = len(line) - len(text)
+        if not indent:
+            description = f"{line} - Length {len(line)}"
+        else:
+            description = (
+                f"{text} - Length {len(text)} plus a {indent}-column indent "
+                f"= {len(line)}"
+            )
+        if continues_row:
+            description += ROW_CONTINUATION_SUFFIX
+        return description
+
     @staticmethod
     def _fix_last_captions_without_ending(caption_list):
         """Set end = start + 4s for trailing captions that were never ended."""
@@ -297,6 +328,7 @@ class SCCReader(BaseReader):
         parts = r.findall(line.lower())
 
         self.time_translator.start_at(parts[0][0])
+        self.position_tracker.start_burst()
         word_list = parts[0][2].split(" ")
 
         for idx, word in enumerate(word_list):
@@ -364,7 +396,11 @@ class SCCReader(BaseReader):
         if self._is_doubled_type(word) and word == self.last_command:
             if word in CUE_STARTING_COMMAND:
                 self.double_starter = True
-            self.last_command = ""
+            # A doubled PAC stays on record so that a Tab Offset following it is
+            # still recognised as belonging to a PAC and applied; clearing it
+            # would leave the offset looking like a stray duplicate and drop the
+            # indentation it carries.
+            self.last_command = word if self._is_pac_command(word) else ""
             return True
 
         if self._is_pac_command(word) and word in self.last_command: