Merge pull request #152 from pettarin/devel

readbeyond · web-flow · commit 702eb2b0a383 · 2016-12-19T17:06:38.000+01:00
Fix for issue #151
diff --git a/README.md b/README.md
@@ -1,16 +1,16 @@
-# aeneas 
+# aeneas
 
 **aeneas** is a Python/C library and a set of tools to automagically synchronize audio and text (aka forced alignment).
 
-* Version: 1.7.0
-* Date: 2016-12-07
+* Version: 1.7.1
+* Date: 2016-12-21
 * Developed by: [ReadBeyond](http://www.readbeyond.it/)
 * Lead Developer: [Alberto Pettarin](http://www.albertopettarin.it/)
 * License: the GNU Affero General Public License Version 3 (AGPL v3)
 * Contact: [aeneas@readbeyond.it](mailto:aeneas@readbeyond.it)
 * Quick Links: [Home](http://www.readbeyond.it/aeneas/) - [GitHub](https://github.com/readbeyond/aeneas/) - [PyPI](https://pypi.python.org/pypi/aeneas/) - [Docs](http://www.readbeyond.it/aeneas/docs/) - [Tutorial](http://www.readbeyond.it/aeneas/docs/clitutorial.html) - [Benchmark](https://readbeyond.github.io/aeneas-benchmark/) - [Mailing List](https://groups.google.com/d/forum/aeneas-forced-alignment) - [Web App](http://aeneasweb.org)
 
- 
+
 ## Goal
 
 **aeneas** automatically generates a **synchronization map**
@@ -109,7 +109,7 @@ The generic OS-independent procedure is simple:
    `espeak`, `ffmpeg`, `ffprobe`, `pip`, and `python`
 
 3. First install `numpy` with `pip` and then `aeneas` (this order is important):
-    
+
     ```bash
     pip install numpy
     pip install aeneas
@@ -185,7 +185,7 @@ The generic OS-independent procedure is simple:
     ```bash
     python -m aeneas.tools.execute_job job.zip output_directory
     ```
-    
+
    File `job.zip` should contain a `config.txt` or `config.xml`
    configuration file, providing **aeneas**
    with all the information needed to parse the input assets
@@ -251,7 +251,7 @@ which explains how to use the built-in command line tools.
 * Extensive test suite including 1,200+ unit/integration/performance tests, that run and must pass before each release
 
 
-## Limitations and Missing Features 
+## Limitations and Missing Features
 
 * Audio should match the text: large portions of spurious text or audio might produce a wrong sync map
 * Audio is assumed to be spoken: not suitable for song captioning, YMMV for CC applications
@@ -304,7 +304,7 @@ No copy rights were harmed in the making of this project.
 
 ## Supporting and Contributing
 
-### Sponsors 
+### Sponsors
 
 * **July 2015**: [Michele Gianella](https://plus.google.com/+michelegianella/about) generously supported the development of the boundary adjustment code (v1.0.4)
 
diff --git a/README.rst b/README.rst
@@ -4,8 +4,8 @@ aeneas
 **aeneas** is a Python/C library and a set of tools to automagically
 synchronize audio and text (aka forced alignment).
 
--  Version: 1.7.0
--  Date: 2016-12-07
+-  Version: 1.7.1
+-  Date: 2016-12-21
 -  Developed by: `ReadBeyond <http://www.readbeyond.it/>`__
 -  Lead Developer: `Alberto Pettarin <http://www.albertopettarin.it/>`__
 -  License: the GNU Affero General Public License Version 3 (AGPL v3)
diff --git a/VERSION b/VERSION
@@ -1 +1 @@
-1.7.0
+1.7.1
diff --git a/aeneas/__init__.py b/aeneas/__init__.py
@@ -35,4 +35,4 @@
     """
 __license__ = "GNU AGPL v3"
 __status__ = "Production"
-__version__ = "1.7.0"
+__version__ = "1.7.1"
diff --git a/aeneas/cdtw/cdtw_setup.py b/aeneas/cdtw/cdtw_setup.py
@@ -49,7 +49,7 @@
 
 setup(
     name="cdtw",
-    version="1.7.0",
+    version="1.7.1",
     description="Python C Extension for computing the DTW as fast as your bare metal allows.",
     ext_modules=[CMODULE],
     include_dirs=[misc_util.get_numpy_include_dirs()]
diff --git a/aeneas/cew/cew_setup.py b/aeneas/cew/cew_setup.py
@@ -47,7 +47,7 @@
 
 setup(
     name="cew",
-    version="1.7.0",
+    version="1.7.1",
     description="Python C Extension for synthesizing text with eSpeak.",
     ext_modules=[CMODULE]
 )
diff --git a/aeneas/cfw/cfw_setup.py b/aeneas/cfw/cfw_setup.py
@@ -54,7 +54,7 @@
 
 setup(
     name="cfw",
-    version="1.7.0",
+    version="1.7.1",
     description="Python C Extension for synthesizing text with Festival.",
     ext_modules=[CMODULE]
 )
diff --git a/aeneas/cmfcc/cmfcc_setup.py b/aeneas/cmfcc/cmfcc_setup.py
@@ -50,7 +50,7 @@
 
 setup(
     name="cmfcc",
-    version="1.7.0",
+    version="1.7.1",
     description="Python C Extension for computing the MFCCs as fast as your bare metal allows.",
     ext_modules=[CMODULE],
     include_dirs=[misc_util.get_numpy_include_dirs()]
diff --git a/aeneas/cwave/cwave_setup.py b/aeneas/cwave/cwave_setup.py
@@ -49,7 +49,7 @@
 
 setup(
     name="cwave",
-    version="1.7.0",
+    version="1.7.1",
     description="Python C Extension for for reading WAVE files.",
     ext_modules=[CMODULE],
     include_dirs=[misc_util.get_numpy_include_dirs()]
diff --git a/aeneas/globalfunctions.py b/aeneas/globalfunctions.py
@@ -1005,18 +1005,27 @@ def delete_directory(path):
             pass
 
 
-def delete_file(handler, path):
+def close_file_handler(handler):
     """
-    Safely delete file.
+    Safely close the given file handler.
 
     :param object handler: the file handler (as returned by tempfile)
-    :param string path: the file path
     """
     if handler is not None:
         try:
             os.close(handler)
         except:
             pass
+
+
+def delete_file(handler, path):
+    """
+    Safely delete file.
+
+    :param object handler: the file handler (as returned by tempfile)
+    :param string path: the file path
+    """
+    close_file_handler(handler)
     if path is not None:
         try:
             os.remove(path)
diff --git a/aeneas/tests/test_globalfunctions.py b/aeneas/tests/test_globalfunctions.py
@@ -562,6 +562,14 @@ def test_delete_directory_not_existing(self):
         gf.delete_directory(orig)
         self.assertFalse(gf.directory_exists(orig))
 
+    def test_close_file_handler(self):
+        handler, path = gf.tmp_file()
+        self.assertTrue(gf.file_exists(path))
+        gf.close_file_handler(handler)
+        self.assertTrue(gf.file_exists(path))
+        gf.delete_file(handler, path)
+        self.assertFalse(gf.file_exists(path))
+
     def test_delete_file_existing(self):
         handler, path = gf.tmp_file()
         self.assertTrue(gf.file_exists(path))
diff --git a/aeneas/tools/abstract_cli_program.py b/aeneas/tools/abstract_cli_program.py
@@ -366,6 +366,7 @@ def run(self, arguments, show_help=True):
 
         # create logger
         self.logger = Logger(tee=self.verbose, tee_show_datetime=self.very_verbose)
+        self.log([u"Running aeneas %s", aeneas_version])
         self.log([u"Formal arguments: %s", self.formal_arguments])
         self.log([u"Actual arguments: %s", self.actual_arguments])
         self.log([u"Runtime configuration: '%s'", self.rconf.config_string])
diff --git a/aeneas/tools/execute_task.py b/aeneas/tools/execute_task.py
@@ -133,7 +133,7 @@ class ExecuteTaskCLI(AbstractCLIProgram):
             u"description": u"input: plain text, output: TSV, tts engine: Festival",
             u"audio": AUDIO_FILE,
             u"text": gf.relative_path("res/plain.txt", __file__),
-            u"config": u"task_language=eng|is_text_type=plain|os_task_file_format=tsv",
+            u"config": u"task_language=eng-USA|is_text_type=plain|os_task_file_format=tsv",
             u"syncmap": "output/sonnet.festival.tsv",
             u"options": u"-r=\"tts=festival\"",
             u"show": False
@@ -205,7 +205,7 @@ class ExecuteTaskCLI(AbstractCLIProgram):
             u"description": u"input: multilevel plain text (mplain), different TTS engines, output: JSON",
             u"audio": AUDIO_FILE,
             u"text": gf.relative_path("res/mplain.txt", __file__),
-            u"config": u"task_language=eng|is_text_type=mplain|os_task_file_format=json",
+            u"config": u"task_language=eng-USA|is_text_type=mplain|os_task_file_format=json",
             u"syncmap": "output/sonnet.mplain.json",
             u"options": u"-r=\"tts_l1=festival|tts_l2=festival|tts_l3=espeak\"",
             u"show": False
@@ -376,7 +376,7 @@ class ExecuteTaskCLI(AbstractCLIProgram):
             u"description": u"input: single word granularity plain text, output: AUD, tts engine: Festival, TTS cache on",
             u"audio": AUDIO_FILE,
             u"text": gf.relative_path("res/words.txt", __file__),
-            u"config": u"task_language=eng|is_text_type=plain|os_task_file_format=aud",
+            u"config": u"task_language=eng-USA|is_text_type=plain|os_task_file_format=aud",
             u"syncmap": "output/sonnet.words.aud",
             u"options": u"-r=\"tts=festival|tts_cache=True\"",
             u"show": False
diff --git a/aeneas/ttswrappers/basettswrapper.py b/aeneas/ttswrappers/basettswrapper.py
@@ -470,7 +470,7 @@ def _synthesize_single_python_helper(self, text, voice_code, output_file_path=No
         return the audio data at the end of the function call;
         if ``False``, just return ``(True, None)`` in case of success.
 
-        :rtype: tuple (result, (duration, sample_rate, encoding, data)) or (result, None)
+        :rtype: tuple (result, (duration, sample_rate, codec, data)) or (result, None)
         """
         raise NotImplementedError(u"This function must be implemented in concrete subclasses supporting Python call")
 
@@ -489,7 +489,7 @@ def _synthesize_single_c_extension_helper(self, text, voice_code, output_file_pa
         If ``output_file_path`` is ``None``,
         the audio data will not persist to file at the end of the method.
 
-        :rtype: tuple (result, (duration, sample_rate, encoding, data))
+        :rtype: tuple (result, (duration, sample_rate, codec, data))
         """
         raise NotImplementedError(u"This function might be implemented in concrete subclasses supporting C extension call")
 
@@ -521,12 +521,12 @@ def _synthesize_single_subprocess_helper(self, text, voice_code, output_file_pat
         return the audio data at the end of the function call;
         if ``False``, just return ``(True, None)`` in case of success.
 
-        :rtype: tuple (result, (duration, sample_rate, encoding, data)) or (result, None)
+        :rtype: tuple (result, (duration, sample_rate, codec, data)) or (result, None)
         """
         # return zero if text is the empty string
         if len(text) == 0:
             #
-            # NOTE sample_rate, encoding, data do not matter
+            # NOTE sample_rate, codec, data do not matter
             #      if the duration is 0.000 => set them to None
             #
             self.log(u"len(text) is zero: returning 0.000")
@@ -640,7 +640,7 @@ def _read_audio_data(self, file_path):
         """
         Read audio data from file.
 
-        :rtype: tuple (True, (duration, sample_rate, encoding, data)) or (False, None) on exception
+        :rtype: tuple (True, (duration, sample_rate, codec, data)) or (False, None) on exception
         """
         try:
             self.log(u"Reading audio data...")
@@ -680,22 +680,30 @@ def _synthesize_multiple_generic(self, helper_function, text_file, output_file_p
         """
         self.log(u"Calling TTS engine using multiple generic function...")
 
-        # get sample rate and encoding
-        self.log(u"Determining codec and sample rate with dummy text...")
-        succeeded, data = helper_function(
-            text=u"Dummy text to get sample_rate",
-            voice_code=self._language_to_voice_code(self.DEFAULT_LANGUAGE),
-            output_file_path=None
-        )
-        if not succeeded:
-            self.log_crit(u"An unexpected error occurred in helper_function")
-            return (False, None)
-        du_nu, sample_rate, encoding, da_nu = data
-        self.log(u"Determining codec and sample rate with dummy text... done")
+        # get sample rate and codec
+        self.log(u"Determining codec and sample rate...")
+        if (self.OUTPUT_AUDIO_FORMAT is None) or (len(self.OUTPUT_AUDIO_FORMAT) != 3):
+            self.log(u"Determining codec and sample rate with dummy text...")
+            succeeded, data = helper_function(
+                text=u"Dummy text to get sample_rate",
+                voice_code=self._language_to_voice_code(self.DEFAULT_LANGUAGE),
+                output_file_path=None
+            )
+            if not succeeded:
+                self.log_crit(u"An unexpected error occurred in helper_function")
+                return (False, None)
+            du_nu, sample_rate, codec, da_nu = data
+            self.log(u"Determining codec and sample rate with dummy text... done")
+        else:
+            self.log(u"Reading codec and sample rate from OUTPUT_AUDIO_FORMAT")
+            codec, channels_nu, sample_rate = self.OUTPUT_AUDIO_FORMAT
+        self.log(u"Determining codec and sample rate... done")
+        self.log([u"  codec:       %s", codec])
+        self.log([u"  sample rate: %d", sample_rate])
 
         # open output file
         output_file = AudioFile(rconf=self.rconf, logger=self.logger)
-        output_file.audio_format = encoding
+        output_file.audio_format = codec
         output_file.audio_channels = 1
         output_file.audio_sample_rate = sample_rate
 
@@ -821,5 +829,7 @@ def _loop_use_cache(self, helper_function, num, fragment):
                 self.log(u"Added fragment to cache")
             else:
                 self.log(u"Fragment has zero duration, not adding it to cache")
+            self.log([u"Closing file handler for cached output file path '%s'", file_path])
+            gf.close_file_handler(file_handler)
         self.log([u"Examining fragment %d (cache)... done", num])
         return (True, data)
diff --git a/aeneas/ttswrappers/festivalttswrapper.py b/aeneas/ttswrappers/festivalttswrapper.py
@@ -118,7 +118,7 @@ class FESTIVALTTSWrapper(BaseTTSWrapper):
         ENG_SCT: ENG_SCT,
         ENG_USA: ENG_USA,
     }
-    DEFAULT_LANGUAGE = ENG
+    DEFAULT_LANGUAGE = ENG_USA
 
     CODE_TO_HUMAN = {
         CES: u"Czech",
diff --git a/aeneas_check_setup.py b/aeneas_check_setup.py
@@ -44,7 +44,7 @@
 """
 __license__ = "GNU AGPL 3"
 __status__ = "Production"
-__version__ = "1.7.0"
+__version__ = "1.7.1"
 
 ANSI_ERROR = u"\033[91m"
 ANSI_OK = u"\033[92m"
diff --git a/bin/aeneas_check_setup.py b/bin/aeneas_check_setup.py
@@ -44,7 +44,7 @@
 """
 __license__ = "GNU AGPL 3"
 __status__ = "Production"
-__version__ = "1.7.0"
+__version__ = "1.7.1"
 
 ANSI_ERROR = u"\033[91m"
 ANSI_OK = u"\033[92m"
diff --git a/bin/aeneas_convert_syncmap.py b/bin/aeneas_convert_syncmap.py
@@ -40,7 +40,7 @@
 """
 __license__ = "GNU AGPL 3"
 __status__ = "Production"
-__version__ = "1.7.0"
+__version__ = "1.7.1"
 
 
 def main():
diff --git a/bin/aeneas_download.py b/bin/aeneas_download.py
@@ -40,7 +40,7 @@
 """
 __license__ = "GNU AGPL 3"
 __status__ = "Production"
-__version__ = "1.7.0"
+__version__ = "1.7.1"
 
 
 def main():
diff --git a/bin/aeneas_execute_job.py b/bin/aeneas_execute_job.py
@@ -42,7 +42,7 @@
 """
 __license__ = "GNU AGPL 3"
 __status__ = "Production"
-__version__ = "1.7.0"
+__version__ = "1.7.1"
 
 
 def main():
diff --git a/bin/aeneas_execute_task.py b/bin/aeneas_execute_task.py
@@ -41,7 +41,7 @@
 """
 __license__ = "GNU AGPL 3"
 __status__ = "Production"
-__version__ = "1.7.0"
+__version__ = "1.7.1"
 
 
 def main():
diff --git a/bin/aeneas_plot_waveform.py b/bin/aeneas_plot_waveform.py
@@ -40,7 +40,7 @@
 """
 __license__ = "GNU AGPL 3"
 __status__ = "Production"
-__version__ = "1.7.0"
+__version__ = "1.7.1"
 
 
 def main():
diff --git a/bin/aeneas_synthesize_text.py b/bin/aeneas_synthesize_text.py
@@ -41,7 +41,7 @@
 """
 __license__ = "GNU AGPL 3"
 __status__ = "Production"
-__version__ = "1.7.0"
+__version__ = "1.7.1"
 
 
 def main():
diff --git a/bin/aeneas_validate.py b/bin/aeneas_validate.py
@@ -47,7 +47,7 @@
 """
 __license__ = "GNU AGPL 3"
 __status__ = "Production"
-__version__ = "1.7.0"
+__version__ = "1.7.1"
 
 
 def main():
diff --git a/check_dependencies.py b/check_dependencies.py
@@ -44,7 +44,7 @@
 """
 __license__ = "GNU AGPL 3"
 __status__ = "Production"
-__version__ = "1.7.0"
+__version__ = "1.7.1"
 
 ANSI_ERROR = u"\033[91m"
 ANSI_OK = u"\033[92m"
diff --git a/debian/changelog b/debian/changelog
diff --git a/docs/source/changelog.rst b/docs/source/changelog.rst
diff --git a/docs/source/conf.py b/docs/source/conf.py
diff --git a/docs/source/libtutorial.rst b/docs/source/libtutorial.rst
diff --git a/install_dependencies.sh b/install_dependencies.sh
diff --git a/pyinstaller-aeneas-cli.py b/pyinstaller-aeneas-cli.py
diff --git a/pyinstaller-onedir.spec b/pyinstaller-onedir.spec
diff --git a/pyinstaller-onefile.spec b/pyinstaller-onefile.spec
diff --git a/requirements.txt b/requirements.txt
diff --git a/run_all_unit_tests.py b/run_all_unit_tests.py
diff --git a/setup.py b/setup.py
diff --git a/setupmeta.py b/setupmeta.py

Original file line number	Diff line number	Diff line change
`@@ -47,7 +47,7 @@`
`47`	`47`
`48`	`48`	`setup(`
`49`	`49`	`name="cew",`
`50`		`- version="1.7.0",`
	`50`	`+ version="1.7.1",`
`51`	`51`	`description="Python C Extension for synthesizing text with eSpeak.",`
`52`	`52`	`ext_modules=[CMODULE]`
`53`	`53`	`)`
Original file line number	Diff line number	Diff line change
`@@ -54,7 +54,7 @@`
`54`	`54`
`55`	`55`	`setup(`
`56`	`56`	`name="cfw",`
`57`		`- version="1.7.0",`
	`57`	`+ version="1.7.1",`
`58`	`58`	`description="Python C Extension for synthesizing text with Festival.",`
`59`	`59`	`ext_modules=[CMODULE]`
`60`	`60`	`)`