Version 0.1.2

* Change default speed to ``default``
* Remove duplicated breaks after paragraphs
* Add metadata to resulting MP3 (title, author, track, year,
  genre)

README.rst

@@ -67,7 +67,7 @@ Setting
      * Emphasis level of section titles
      * ``"strong"``, ``"moderate"``, ``"none"``, ``"reduced"``
    - * ``ssml_paragraph_speed``
-     * ``"medium"``
+     * ``"default"``
      * Speech speed
      * ``"x-slow"``, ``"slow"``, ``"medium"``, ``"fast"``, ``"x-fast"``, ``"default"``
    - * ``ssml_break_after_paragraph``
@@ -108,3 +108,21 @@ Repository
 -----------
 
 https://github.com/shibukawa/sphinxcontrib-ssmlbuilder
+
+History
+-------
+
+* 0.1.2 Jan 24 2017
+
+  * Change default speed to ``default``
+  * Remove duplicated breaks after paragraphs
+  * Add metadata to resulting MP3 (title, author, track, year, genre)
+
+* 0.1.1 Jan 23 2017
+
+  * Add skipping block option
+
+* 0.1 Jan 23 2017
+
+  * Initial Version
+

setup.py

@@ -16,7 +16,7 @@
 
 setup(
     name='sphinxcontrib-ssmlbuilder',
-    version='0.1.1',
+    version='0.1.2',
     url='https://github.com/shibukawa/sphinxcontrib-ssmlbuilder',
     download_url='http://pypi.python.org/pypi/sphinxcontrib-ssmlbuilder',
     license='BSD',

sphinxcontrib/builder.py

@@ -15,6 +15,7 @@
 from .writer import SSMLWriter
 from sphinx.util.osutil import SEP, os_path, relative_uri, ensuredir, \
     movefile, copyfile
+from sphinx import addnodes
 import os
 from os import path
 import codecs
@@ -23,6 +24,8 @@
 from fnmatch import fnmatch
 from boto3 import Session
 from botocore.exceptions import BotoCoreError, ClientError
+import datetime
+import re
 
 
 class SSMLBuilder(Builder):
@@ -37,9 +40,8 @@ class SSMLBuilder(Builder):
     ssml_break_around_section_title = [2000, 1600, 1000, 1000, 1000, 1000]
     ssml_break_after_paragraph = 1000
     ssml_emphasis_section_title = ['none', 'none', 'none', 'none', 'none', 'none']
-    ssml_paragraph_speed = 'medium'
+    ssml_paragraph_speed = 'default'
     ssml_polly_audio_output_folder = 'polly'
-    ssml_polly_audio_format = 'mp3'
     ssml_polly_aws_profile = ''
     ssml_polly_aws_voiceid = 'Joanna'
     ssml_polly_apply_docnames = ''
@@ -97,7 +99,7 @@ def prepare_writing(self, docnames):
         self.writer = SSMLWriter(self)
 
     def write_doc(self, docname, doctree):
-        destination = {"hashes": {}, "sequence": []}
+        destination = {"hashes": {}, "sequence": [], "title": ""}
         outfilename = path.join(self.outdir, self.file_transform(docname))
         ensuredir(path.dirname(outfilename))
         self.writer.write(doctree, destination, docname, path.join(self.outdir, docname))
@@ -110,6 +112,23 @@ def write_doc(self, docname, doctree):
         except (IOError, OSError) as err:
             self.warn("error writing file %s: %s" % (outfilename, err))
 
+    def sort_docnames(self):
+        result = [self.config.master_doc]
+        self._sort_docnames(self.config.master_doc, result)
+        return result
+
+    def _sort_docnames(self, docname, traversed):
+        tree = self.env.get_doctree(docname)
+        for toctreenode in tree.traverse(addnodes.toctree):
+            includefiles = map(str, toctreenode['includefiles'])
+            for includefile in includefiles:
+                 if includefile not in traversed:
+                     try:
+                         traversed.append(includefile)
+                         self._sort_docnames(includefile, traversed)
+                     except Exception:
+                         pass
+
     def exec_polly(self):
         print("ssml_polly_aws_profile: ", self.ssml_polly_aws_profile)
         print("ssml_polly_apply_docnames: ", self.config.ssml_polly_apply_docnames)
@@ -137,10 +156,9 @@ def exec_polly(self):
                 allhash.add(hashname)
             hash2path.update(d["hashes"])
             if fnmatch(docname, apply_docname):
-                targets.append({"docname": docname, "sequence": d["sequence"]})
+                targets.append({"docname": docname, "sequence": d["sequence"], "title": d["title"]})
                 for hashname in d["hashes"]:
                     allneededhash.add(hashname)
-
         # read existing path
         existinghash = set()
         for mp3file in os.listdir(workdirpath):
@@ -152,7 +170,7 @@ def exec_polly(self):
         #print("targets", targets)
         #print("must_remove:", must_remove)
         #print("must_convert:", must_convert)
-      
+
         # exec polly
         session = Session(profile_name=self.ssml_polly_aws_profile)
         polly = session.client("polly")
@@ -177,16 +195,37 @@ def exec_polly(self):
             mp3file.write(response.get("AudioStream").read())
             mp3file.close()
 
+        # metadata
+        album = self.config.project
+        year = datetime.datetime.now().year
+        author = ''
+        match = re.match(r'(\d{4}), (.*)', self.config.copyright)
+        if match:
+            year = match.group(1)
+            author = match.group(2)
+        document_order = self.sort_docnames()
+
         # concat mp3 fragments
         task = 1 
         for target in targets:
             docname = target['docname']
+            title = target['title']
+            track = document_order.index(docname) + 1
+
             print(f"concatinating MP3 fragments: {docname}.mp3 ({task}/{len(targets)})")
             task+=1
             sources = [hashkey + '.mp3' for hashkey in target['sequence']]
             outfilename = path.join(outputpath, f"{docname}.mp3")
             ensuredir(path.dirname(outfilename))
-            args = ['ffmpeg', "-i", "concat:" + "|".join(sources), "-c", "copy", outfilename]
+
+            args = ['ffmpeg', "-y", "-i", "concat:" + "|".join(sources), "-c", "copy",
+                    '-metadata', f'album="{album}"',
+                    '-metadata', f'author="{author}"',
+                    '-metadata', f'title="{title}"',
+                    '-metadata', f'track="{track}"',
+                    '-metadata', 'genre="Audio Book"',
+                    '-metadata', f'year="{year}"',
+                    outfilename]
             #print(args, workdirpath)
             p = subprocess.Popen(args, shell=False, cwd=workdirpath)
             p.wait()

sphinxcontrib/ssmlbuilder.py

@@ -22,7 +22,7 @@ def setup(app):
     app.add_config_value('ssml_break_around_section_title', [2000, 1600, 1000, 1000, 1000, 1000], True)
     app.add_config_value('ssml_emphasis_section_title', ['none', 'none', 'none', 'none', 'none', 'none'], True)
     app.add_config_value('ssml_skip_block', {'comment': True, 'table': True, 'codeblock': True}, True)
-    app.add_config_value('ssml_paragraph_speed', 'medium', True)
+    app.add_config_value('ssml_paragraph_speed', 'default', True)
     app.add_config_value('ssml_break_after_paragraph', 1000, True)
     app.add_config_value('ssml_polly_audio_output_folder', "polly", True)
     app.add_config_value('ssml_polly_aws_profile', "", False)

sphinxcontrib/writer.py

@@ -61,13 +61,13 @@ def __init__(self, document, builder, destination, docname, basepath):
         self.basepath = basepath
         self.state = ['regular']
 
+    def active(self):
+        laststate = self.state[-1]
+        return not self.builder.ssml_skip_block.get(laststate, False)
+
     def add_text(self, text):
         # type: (unicode) -> None
-        laststate = self.state[-1]
-        if self.builder.ssml_skip_block.get(laststate, False):
-            # this code block is configured to be skip
-            pass
-        else:
+        if self.active():
             self.contents.append([len(text), REGULAR, escape(text)])
 
     def reset_content(self):
@@ -107,7 +107,7 @@ def next_node_character_count(index):
                 middle += "-" + str(i + 1)
             filepath = self.basepath + middle + ".ssml"
             filename = self.docname + middle + ".ssml"
-            sha = hashlib.sha256(output.encode('utf-8')).hexdigest()
+            sha = hashlib.sha1(output.encode('utf-8')).hexdigest()
             self.destination["hashes"][sha] = filename
             self.destination["sequence"].append(sha)
             f = open(filepath, "w")
@@ -168,6 +168,8 @@ def depart_glossary(self, node):
         pass
 
     def visit_title(self, node):
+        if not self.destination['title']:
+            self.destination['title'] = node.astext()
         level = self.sectionlevel-1
         breaklength = self.builder.ssml_break_around_section_title[level]
         emphasis = self.builder.ssml_emphasis_section_title[level]
@@ -389,7 +391,8 @@ def depart_compact_paragraph(self, node):
 
     def visit_paragraph(self, node):
         # type: (nodes.Node) -> None
-        self.contents.append([0, JOIN_BEFORE, '<break time="%dms" />' % self.builder.ssml_break_after_paragraph])
+        if self.active():
+            self.contents.append([0, JOIN_BEFORE, '<break time="%dms" />' % self.builder.ssml_break_after_paragraph])
         info("visit", node)
 
     def depart_paragraph(self, node):