ttsforge 0.1.0__tar.gz → 0.1.2__tar.gz

{ttsforge-0.1.0 → ttsforge-0.1.2}/.codecrate.toml

@@ -1,8 +1,9 @@
 [codecrate]
 output = "context_ttsforge.md"
 keep_docstrings = true
-dedupe = true
+dedupe = false
 metadata = false
+manifest = false
 respect_gitignore = true
 exclude = ["*/.venv/*"]
 include = ["**/*.py", "**/*.toml", "**/*.rst", "**/*.md"]

{ttsforge-0.1.0 → ttsforge-0.1.2}/.github/workflows/codecov.yml

@@ -12,7 +12,7 @@ jobs:
     - name: Install espeak-ng
       run: |
         sudo apt-get update
-        sudo apt-get install -y espeak-ng
+        sudo apt-get install -y espeak-ng ffmpeg
     - name: 'generate report'
       run: |
         pip install coverage click pytest pytest-cov

{ttsforge-0.1.0 → ttsforge-0.1.2}/.gitignore

@@ -216,7 +216,6 @@ onnx/
 *.m4a
 
 # Test/demo scripts at project root
-test_*.py
 demo_*.py
 
 # Binary data files

{ttsforge-0.1.0 → ttsforge-0.1.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ttsforge
-Version: 0.1.0
+Version: 0.1.2
 Summary: Generate audiobooks from EPUB files using Kokoro ONNX TTS.
 Author-email: Holger Nahrstaedt <nahrstaedt@gmail.com>
 License: MIT License
@@ -396,14 +396,14 @@ SSMD files use a simple markdown-like syntax:
 **Custom Phonemes**:
 
 ```
-[Hermione](ph: /hɝmˈIni/)    # Override pronunciation
-[API](ph: /ˌeɪpiˈaɪ/)        # Technical terms
+[Hermione]{ph="hɝmˈIni"}    # Override pronunciation
+[API]{ph="ˌeɪpiˈaɪ"}        # Technical terms
 ```
 
 **Language Switching** (planned):
 
 ```
-[Bonjour](fr)    # Mark text as French
+[Bonjour]{lang="fr"}    # Mark text as French
 ```
 
 #### Example SSMD File
@@ -411,7 +411,7 @@ SSMD files use a simple markdown-like syntax:
 ```ssmd
 Chapter One ...p
 
-[Harry](ph: /hæɹi/) Potter was a *highly unusual* boy in many ways. ...s
+[Harry]{ph="hæɹi"} Potter was a *highly unusual* boy in many ways. ...s
 For one thing, he **hated** the summer holidays more than any other
 time of year. ...s For another, he really wanted to do his homework,
 but was forced to do it in secret, in the dead of the night. ...p
@@ -498,12 +498,12 @@ Edit `custom_phonemes.json` to fix any incorrect phonemes. The file format is:
   },
   "entries": {
     "Hermione": {
-      "phoneme": "/hɝmˈIni/",
+      "phoneme": "hɝmˈIni",
       "occurrences": 847,
       "verified": false
     },
     "Kubernetes": {
-      "phoneme": "/kubɚnˈɛtɪs/",
+      "phoneme": "kubɚnˈɛtɪs",
       "occurrences": 12,
       "verified": false
     }
@@ -515,8 +515,8 @@ Or use the simple format:
 
 ```json
 {
-  "Hermione": "/hɝmˈIni/",
-  "Kubernetes": "/kubɚnˈɛtɪs/"
+  "Hermione": "hɝmˈIni",
+  "Kubernetes": "kubɚnˈɛtɪs"
 }
 ```
 
@@ -548,9 +548,9 @@ You can create a dictionary manually without extraction:
 
 ```json
 {
-  "Katniss": "/kætnɪs/",
-  "Peeta": "/pitə/",
-  "Panem": "/pænəm/"
+  "Katniss": "kætnɪs",
+  "Peeta": "pitə",
+  "Panem": "pænəm"
 }
 ```
 
@@ -617,6 +617,7 @@ ttsforge convert book.epub --gpu
 | `pause_paragraph`           | `0.9`          | Paragraph pause (seconds)            |
 | `pause_variance`            | `0.05`         | Pause variance (seconds)             |
 | `pause_mode`                | `auto`         | Pause mode (`tts`, `manual`, `auto`) |
+| `enable_short_sentence`     | `None`         | Handle short sentences               |
 | `announce_chapters`         | `true`         | Speak chapter titles                 |
 | `chapter_pause_after_title` | `2.0`          | Pause after chapter title            |
 | `phonemization_lang`        | `None`         | Override phonemization language      |

{ttsforge-0.1.0 → ttsforge-0.1.2}/README.md

@@ -333,14 +333,14 @@ SSMD files use a simple markdown-like syntax:
 **Custom Phonemes**:
 
 ```
-[Hermione](ph: /hɝmˈIni/)    # Override pronunciation
-[API](ph: /ˌeɪpiˈaɪ/)        # Technical terms
+[Hermione]{ph="hɝmˈIni"}    # Override pronunciation
+[API]{ph="ˌeɪpiˈaɪ"}        # Technical terms
 ```
 
 **Language Switching** (planned):
 
 ```
-[Bonjour](fr)    # Mark text as French
+[Bonjour]{lang="fr"}    # Mark text as French
 ```
 
 #### Example SSMD File
@@ -348,7 +348,7 @@ SSMD files use a simple markdown-like syntax:
 ```ssmd
 Chapter One ...p
 
-[Harry](ph: /hæɹi/) Potter was a *highly unusual* boy in many ways. ...s
+[Harry]{ph="hæɹi"} Potter was a *highly unusual* boy in many ways. ...s
 For one thing, he **hated** the summer holidays more than any other
 time of year. ...s For another, he really wanted to do his homework,
 but was forced to do it in secret, in the dead of the night. ...p
@@ -435,12 +435,12 @@ Edit `custom_phonemes.json` to fix any incorrect phonemes. The file format is:
   },
   "entries": {
     "Hermione": {
-      "phoneme": "/hɝmˈIni/",
+      "phoneme": "hɝmˈIni",
       "occurrences": 847,
       "verified": false
     },
     "Kubernetes": {
-      "phoneme": "/kubɚnˈɛtɪs/",
+      "phoneme": "kubɚnˈɛtɪs",
       "occurrences": 12,
       "verified": false
     }
@@ -452,8 +452,8 @@ Or use the simple format:
 
 ```json
 {
-  "Hermione": "/hɝmˈIni/",
-  "Kubernetes": "/kubɚnˈɛtɪs/"
+  "Hermione": "hɝmˈIni",
+  "Kubernetes": "kubɚnˈɛtɪs"
 }
 ```
 
@@ -485,9 +485,9 @@ You can create a dictionary manually without extraction:
 
 ```json
 {
-  "Katniss": "/kætnɪs/",
-  "Peeta": "/pitə/",
-  "Panem": "/pænəm/"
+  "Katniss": "kætnɪs",
+  "Peeta": "pitə",
+  "Panem": "pænəm"
 }
 ```
 
@@ -554,6 +554,7 @@ ttsforge convert book.epub --gpu
 | `pause_paragraph`           | `0.9`          | Paragraph pause (seconds)            |
 | `pause_variance`            | `0.05`         | Pause variance (seconds)             |
 | `pause_mode`                | `auto`         | Pause mode (`tts`, `manual`, `auto`) |
+| `enable_short_sentence`     | `None`         | Handle short sentences               |
 | `announce_chapters`         | `true`         | Speak chapter titles                 |
 | `chapter_pause_after_title` | `2.0`          | Pause after chapter title            |
 | `phonemization_lang`        | `None`         | Override phonemization language      |

{ttsforge-0.1.0 → ttsforge-0.1.2}/docs/api/index.rst

@@ -60,10 +60,6 @@ Utilities
 **ttsforge.vocab**
    Vocabulary utilities and metadata.
 
-**ttsforge.trim**
-   Audio trimming utilities for silence removal.
-
-
 Quick API Examples
 ------------------
 
@@ -80,9 +76,9 @@ Basic Text-to-Speech
        voice="af_heart",
        speed=1.0,
        use_gpu=False,
-       pause_clause=0.25,
-       pause_sentence=0.2,
-       pause_paragraph=0.75,
+       pause_clause=0.3,
+       pause_sentence=0.5,
+       pause_paragraph=0.9,
        pause_variance=0.05,
    )
    runner = KokoroRunner(opts, log=print)
@@ -242,8 +238,3 @@ Auto-generated API Documentation
    :members:
    :undoc-members:
    :show-inheritance:
-
-.. automodule:: ttsforge.trim
-   :members:
-   :undoc-members:
-   :show-inheritance:

{ttsforge-0.1.0 → ttsforge-0.1.2}/docs/cli.rst

@@ -85,6 +85,9 @@ Options
 ``--pause-mode MODE``
    Pause mode: ``tts``, ``manual``, or ``auto``. Default: ``auto``.
 
+``--enable-short-sentence``
+   Enable special handling for short sentences (less than 5 words).
+
 ``--announce-chapters / --no-announce-chapters``
    Read chapter titles aloud before chapter content. Default: enabled.
 

{ttsforge-0.1.0 → ttsforge-0.1.2}/docs/configuration.rst

@@ -452,6 +452,7 @@ Here's an example ``config.json`` with custom settings:
      "pause_paragraph": 0.9,
      "pause_variance": 0.05,
      "pause_mode": "auto",
+     "enable_short_sentence": None,
      "announce_chapters": true,
      "chapter_pause_after_title": 2.0,
       "save_chapters_separately": false,

{ttsforge-0.1.0 → ttsforge-0.1.2}/docs/ssmd.rst

@@ -101,15 +101,15 @@ Override pronunciation using IPA phonemes:
 
 .. code-block:: ssmd
 
-   [word](ph: /phoneme/)
+   [word]{ph="phoneme"}
 
 Examples:
 
 .. code-block:: ssmd
 
-   [Hermione](ph: /hɝmˈIni/) Granger was Harry's best friend. ...s
-   The [API](ph: /ˌeɪpiˈaɪ/) supports [JSON](ph: /dʒˈeɪsɑn/). ...s
-   [Kubernetes](ph: /kubɚnˈɛtɪs/) is a container orchestrator. ...s
+   [Hermione]{ph="hɝmˈIni"} Granger was Harry's best friend. ...s
+   The [API]{ph="ˌeɪpiˈaɪ"} supports [JSON]{ph="dʒˈeɪsɑn"}. ...s
+   [Kubernetes]{ph="kubɚnˈɛtɪs"} is a container orchestrator. ...s
 
 
 Language Switching (Planned)
@@ -119,8 +119,8 @@ Mark text as a different language (placeholder for future):
 
 .. code-block:: ssmd
 
-   [Bonjour](fr)      # French text
-   [Hola](es)         # Spanish text
+   [Bonjour]{lang="fr"}      # French text
+   [Hola]{lang="es"}         # Spanish text
 
 
 Complete Example
@@ -132,14 +132,14 @@ Here's a complete SSMD file example:
 
    Chapter One ...p
 
-   [Harry](ph: /hæɹi/) Potter was a *highly unusual* boy in many ways. ...s
+   [Harry]{ph="hæɹi"} Potter was a *highly unusual* boy in many ways. ...s
    For one thing, he **hated** the summer holidays more than any other
    time of year. ...s For another, he really wanted to do his homework,
    but was forced to do it in secret, in the dead of the night. ...p
 
    And he also happened to be a wizard. ...p
 
-   The [Dursleys](ph: /dɝzliz/) had everything they wanted, but they
+   The [Dursleys]{ph="dɝzliz"} had everything they wanted, but they
    also had a secret. ...s And their greatest fear was that somebody
    would discover it. ...p
 
@@ -163,7 +163,7 @@ The generated SSMD will include:
 
 .. code-block:: ssmd
 
-   [Hermione](ph: /hɝmˈIni/) loved reading books. ...s
+   [Hermione]{ph="hɝmˈIni"} loved reading books. ...s
 
 
 HTML Emphasis Detection

ttsforge-0.1.2/tests/test_chapter_marker_leading_space.py

@@ -0,0 +1,88 @@
+"""Test chapter marker removal with leading whitespace."""
+
+import re
+
+
+class TestChapterMarkerLeadingWhitespace:
+    """Test that chapter markers are removed even with leading whitespace."""
+
+    def test_marker_with_no_leading_space(self):
+        """Test normal case - marker at start of line."""
+        text = "<<CHAPTER: Test Chapter>>\n\nThis is the content."
+        pattern = r"^<<CHAPTER:[^>]*>>\s*\n*"
+        result = re.sub(pattern, "", text, count=1, flags=re.MULTILINE)
+        assert result == "This is the content."
+
+    def test_marker_with_leading_space(self):
+        """Test marker with a leading space - should be removed with new pattern."""
+        text = " <<CHAPTER: Test Chapter>>\n\nThis is the content."
+        # New pattern handles leading whitespace
+        pattern = r"^\s*<<CHAPTER:[^>]*>>\s*\n*"
+        result = re.sub(pattern, "", text, count=1, flags=re.MULTILINE)
+        assert result == "This is the content."
+
+    def test_marker_with_leading_tabs(self):
+        """Test marker with leading tabs - should be removed with new pattern."""
+        text = "\t<<CHAPTER: Test Chapter>>\n\nThis is the content."
+        pattern = r"^\s*<<CHAPTER:[^>]*>>\s*\n*"
+        result = re.sub(pattern, "", text, count=1, flags=re.MULTILINE)
+        assert result == "This is the content."
+
+    def test_marker_with_multiple_spaces(self):
+        """Test marker with multiple leading spaces -
+        should be removed with new pattern."""
+        text = "   <<CHAPTER: Test Chapter>>\n\nThis is the content."
+        pattern = r"^\s*<<CHAPTER:[^>]*>>\s*\n*"
+        result = re.sub(pattern, "", text, count=1, flags=re.MULTILINE)
+        assert result == "This is the content."
+
+    def test_improved_pattern_handles_leading_whitespace(self):
+        """Test that improved pattern handles all leading whitespace cases."""
+        # Improved pattern that handles leading whitespace
+        improved_pattern = r"^\s*<<CHAPTER:[^>]*>>\s*\n*"
+
+        test_cases = [
+            ("<<CHAPTER: Test>>\n\nContent", "Content"),
+            (" <<CHAPTER: Test>>\n\nContent", "Content"),
+            ("\t<<CHAPTER: Test>>\n\nContent", "Content"),
+            ("   <<CHAPTER: Test>>\n\nContent", "Content"),
+            (" \t <<CHAPTER: Test>>\n\nContent", "Content"),
+        ]
+
+        for text, expected in test_cases:
+            result = re.sub(improved_pattern, "", text, count=1, flags=re.MULTILINE)
+            assert result == expected, f"Failed for input: {repr(text)}"
+
+    def test_marker_not_at_line_start_still_removed_with_multiline(self):
+        """Test that marker after newline is removed (MULTILINE mode)."""
+        text = "Some text\n<<CHAPTER: Test>>\n\nContent"
+        improved_pattern = r"^\s*<<CHAPTER:[^>]*>>\s*\n*"
+        result = re.sub(improved_pattern, "", text, count=1, flags=re.MULTILINE)
+        assert result == "Some text\nContent"
+
+    def test_only_first_marker_removed(self):
+        """Test that only the first marker is removed (count=1)."""
+        text = "<<CHAPTER: One>>\n\nSome text <<CHAPTER: Two>> inside it."
+        improved_pattern = r"^\s*<<CHAPTER:[^>]*>>\s*\n*"
+        result = re.sub(improved_pattern, "", text, count=1, flags=re.MULTILINE)
+        assert result == "Some text <<CHAPTER: Two>> inside it."
+
+    def test_real_world_epub_scenario(self):
+        """Test realistic epub2text output with potential whitespace issues."""
+        # Simulate what epub2text might return with whitespace quirks
+        epub_content = " <<CHAPTER: THE STORY SO FAR>>\n\nIn the shadow of the Apt..."
+
+        # Old pattern (fails)
+        old_pattern = r"^<<CHAPTER:[^>]*>>\s*\n*"
+        old_result = re.sub(old_pattern, "", epub_content, count=1, flags=re.MULTILINE)
+
+        # New pattern (works)
+        new_pattern = r"^\s*<<CHAPTER:[^>]*>>\s*\n*"
+        new_result = re.sub(new_pattern, "", epub_content, count=1, flags=re.MULTILINE)
+
+        # Verify old pattern fails to remove marker
+        assert "<<CHAPTER:" in old_result, "Old pattern should fail with leading space"
+
+        # Verify new pattern successfully removes marker
+        assert "<<CHAPTER:" not in new_result, "New pattern should remove marker"
+        assert new_result == "In the shadow of the Apt..."

ttsforge-0.1.2/tests/test_chapter_selection.py

@@ -0,0 +1,20 @@
+import pytest
+
+from ttsforge.chapter_selection import parse_chapter_selection
+
+
+def test_parse_all() -> None:
+    assert parse_chapter_selection("all", 5) == [0, 1, 2, 3, 4]
+
+
+def test_parse_ranges_and_commas() -> None:
+    assert parse_chapter_selection("1-3,5", 6) == [0, 1, 2, 4]
+
+
+def test_parse_open_ended_range() -> None:
+    assert parse_chapter_selection("3-", 5) == [2, 3, 4]
+
+
+def test_parse_invalid_range() -> None:
+    with pytest.raises(ValueError):
+        parse_chapter_selection("5-2", 6)

ttsforge-0.1.2/tests/test_cli_smoke.py

@@ -0,0 +1,27 @@
+from pathlib import Path
+
+from click.testing import CliRunner
+
+from ttsforge.cli import main
+
+
+def test_info_and_list_smoke(tmp_path: Path) -> None:
+    text = """Title: Sample Book
+Author: Jane Doe
+Language: English
+
+CHAPTER I
+This is the first chapter.
+
+CHAPTER II
+This is the second chapter.
+"""
+    input_file = tmp_path / "sample.txt"
+    input_file.write_text(text, encoding="utf-8")
+
+    runner = CliRunner()
+    info_result = runner.invoke(main, ["info", str(input_file)])
+    assert info_result.exit_code == 0
+
+    list_result = runner.invoke(main, ["list", str(input_file)])
+    assert list_result.exit_code == 0

{ttsforge-0.1.0 → ttsforge-0.1.2}/tests/test_constants.py

@@ -20,7 +20,7 @@ class TestLanguageDescriptions:
 
     def test_all_language_codes_have_descriptions(self):
         """All language codes should have descriptions."""
-        expected_codes = {"a", "b", "e", "f", "h", "i", "j", "p", "z"}
+        expected_codes = {"a", "b", "d", "e", "f", "h", "i", "j", "p", "z"}
         assert set(LANGUAGE_DESCRIPTIONS.keys()) == expected_codes
 
     def test_english_variants(self):
@@ -132,11 +132,6 @@ class TestDefaultVoiceForLang:
                 lang in DEFAULT_VOICE_FOR_LANG
             ), f"Language {lang} needs default voice"
 
-    def test_default_voices_exist_in_voices_list(self):
-        """All default voices should exist in VOICES list."""
-        for lang, voice in DEFAULT_VOICE_FOR_LANG.items():
-            assert voice in VOICES, f"Default voice {voice} for {lang} not in VOICES"
-
     def test_default_voices_match_language(self):
         """Default voices should match their language."""
         for lang, voice in DEFAULT_VOICE_FOR_LANG.items():

ttsforge-0.1.2/tests/test_conversion_state.py

@@ -0,0 +1,84 @@
+import json
+from pathlib import Path
+
+from ttsforge.conversion import ChapterState, ConversionState
+
+
+def test_conversion_state_roundtrip(tmp_path: Path) -> None:
+    state = ConversionState(
+        source_file="book.epub",
+        source_hash="abc123",
+        output_file="book.m4b",
+        work_dir=str(tmp_path),
+        voice="af_heart",
+        language="a",
+        speed=1.0,
+        split_mode="auto",
+        output_format="m4b",
+        chapters=[
+            ChapterState(
+                index=0,
+                title="Chapter 1",
+                content_hash="hash",
+                completed=True,
+                audio_file="chapter_001.wav",
+                duration=1.2,
+                char_count=100,
+                ssmd_file="chapter_001.ssmd",
+                ssmd_hash="ssmdhash",
+            )
+        ],
+        started_at="2024-01-01 00:00:00",
+    )
+    state_file = tmp_path / "state.json"
+    state.save(state_file)
+
+    loaded = ConversionState.load(state_file)
+    assert loaded is not None
+    assert loaded.voice == "af_heart"
+    assert loaded.chapters[0].audio_file == "chapter_001.wav"
+    assert loaded.chapters[0].completed is True
+
+    assert not (tmp_path / "state.json.tmp").exists()
+
+
+def test_conversion_state_backward_compat(tmp_path: Path) -> None:
+    data = {
+        "version": 1,
+        "source_file": "book.epub",
+        "source_hash": "hash",
+        "output_file": "book.m4b",
+        "work_dir": str(tmp_path),
+        "voice": "af_heart",
+        "language": "a",
+        "speed": 1.0,
+        "split_mode": "auto",
+        "output_format": "m4b",
+        "chapters": [
+            {
+                "index": 0,
+                "title": "Chapter 1",
+                "content_hash": "hash",
+                "completed": False,
+                "audio_file": None,
+                "duration": 0.0,
+                "char_count": 10,
+                "ssmd_file": None,
+                "ssmd_hash": None,
+            }
+        ],
+        "segment_pause_min": 0.1,
+        "segment_pause_max": 0.3,
+        "paragraph_pause_min": 0.5,
+        "paragraph_pause_max": 1.0,
+    }
+
+    state_file = tmp_path / "legacy_state.json"
+    state_file.write_text(json.dumps(data), encoding="utf-8")
+
+    loaded = ConversionState.load(state_file)
+    assert loaded is not None
+    assert loaded.pause_sentence == 0.2
+    assert loaded.pause_paragraph == 0.75
+    assert loaded.pause_clause == 0.3
+    assert loaded.pause_variance >= 0.01

{ttsforge-0.1.0 → ttsforge-0.1.2}/tests/test_phoneme_dictionary.py

@@ -13,8 +13,8 @@ class TestPhonemeDictionary:
     def test_load_simple_dictionary(self):
         """Test loading a simple phoneme dictionary."""
         test_dict = {
-            "Misaki": "/misˈɑki/",
-            "Kubernetes": "/kubɚnˈɛtɪs/",
+            "Misaki": "misˈɑki",
+            "Kubernetes": "kubɚnˈɛtɪs",
         }
 
         with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as f:
@@ -40,8 +40,8 @@ class TestPhonemeDictionary:
                 "language": "en-us",
             },
             "entries": {
-                "Misaki": {"phoneme": "/misˈɑki/", "occurrences": 42},
-                "nginx": {"phoneme": "/ˈɛnʤɪnˈɛks/", "occurrences": 8},
+                "Misaki": {"phoneme": "misˈɑki", "occurrences": 42},
+                "nginx": {"phoneme": "ˈɛnʤɪnˈɛks", "occurrences": 8},
             },
         }
 
@@ -67,8 +67,8 @@ class TestPhonemeDictionary:
         """Test loading dictionary with metadata format but simple string values."""
         test_dict = {
             "entries": {
-                "Misaki": "/misˈɑki/",
-                "nginx": "/ˈɛnʤɪnˈɛks/",
+                "Misaki": "misˈɑki",
+                "nginx": "ˈɛnʤɪnˈɛks",
             }
         }
 
@@ -112,8 +112,8 @@ class TestPhonemeDictionary:
     def test_phonemize_with_dictionary(self):
         """Test phonemization with custom dictionary - through SSMD notation."""
         test_dict = {
-            "Misaki": "/misˈɑki/",
-            "Kubernetes": "/kubɚnˈɛtɪs/",
+            "Misaki": "misˈɑki",
+            "Kubernetes": "kubɚnˈɛtɪs",
         }
 
         with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as f:
@@ -130,8 +130,8 @@ class TestPhonemeDictionary:
             ssmd_text = tokenizer._phoneme_dictionary_obj.apply(text)
 
             # Verify SSMD notation is applied
-            assert "[Misaki]{ph=" in ssmd_text or "[Misaki](ph:" in ssmd_text
-            assert "[Kubernetes]{ph=" in ssmd_text or "[Kubernetes](ph:" in ssmd_text
+            assert "[Misaki]{ph=" in ssmd_text
+            assert "[Kubernetes]{ph=" in ssmd_text
         finally:
             Path(temp_path).unlink()
 
@@ -162,7 +162,7 @@ class TestPhonemeDictionary:
 
     def test_case_sensitive_matching(self):
         """Test case-sensitive dictionary matching."""
-        test_dict = {"Misaki": "/misˈɑki/"}
+        test_dict = {"Misaki": "misˈɑki"}
 
         with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as f:
             json.dump(test_dict, f)
@@ -184,13 +184,13 @@ class TestPhonemeDictionary:
             assert phoneme_count == 1, f"Expected 1 match, got {phoneme_count}"
 
             # Verify it's "Misaki" that matched
-            assert "[Misaki]{ph=" in ssmd_text or "[Misaki](ph:" in ssmd_text
+            assert "[Misaki]{ph=" in ssmd_text
         finally:
             Path(temp_path).unlink()
 
     def test_word_boundaries(self):
         """Test that word boundaries are respected."""
-        test_dict = {"test": "/tˈɛst/"}
+        test_dict = {"test": "tˈɛst"}
 
         with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as f:
             json.dump(test_dict, f)
@@ -246,7 +246,7 @@ class TestPhonemeDictionary:
     def test_special_characters_in_words(self):
         """Test dictionary words with special regex characters (periods, etc.)."""
         # Use a simple word that can be phonemized
-        test_dict = {"Misaki": "/misˈɑki/"}
+        test_dict = {"Misaki": "misˈɑki"}
 
         with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as f:
             json.dump(test_dict, f)
@@ -262,13 +262,13 @@ class TestPhonemeDictionary:
             ssmd_text = tokenizer._phoneme_dictionary_obj.apply(text)
 
             # Should use custom phoneme
-            assert "[Misaki]{ph=" in ssmd_text or "[Misaki](ph:" in ssmd_text
+            assert "[Misaki]{ph=" in ssmd_text
         finally:
             Path(temp_path).unlink()
 
     def test_multiple_occurrences(self):
         """Test that all occurrences of a word are replaced."""
-        test_dict = {"test": "/tˈɛst/"}
+        test_dict = {"test": "tˈɛst"}
 
         with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as f:
             json.dump(test_dict, f)
@@ -293,8 +293,8 @@ class TestPhonemeDictionary:
         # Note: Multi-word phoneme annotations have limitations in kokorog2p's
         # markdown processing. Testing with overlapping single words instead.
         test_dict = {
-            "testing": "/tˈɛstɪŋ/",
-            "test": "/tˈɛst/",  # Shorter word, different pronunciation
+            "testing": "tˈɛstɪŋ",
+            "test": "tˈɛst",  # Shorter word, different pronunciation
         }
 
         with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as f: