gradint-build/samples/utils/README.txt → samples/utils/README.txt

@@ -10,26 +10,20 @@ C:\Windows\system32 or somewhere.
 
 All require Python (from www.python.org).  All systems
 except Windows have that anyway if you can run gradint.
+However, some of these utilities are still Python2-only,
+while the main Gradint is now compatible with both 2 and 3.
 
 SPLITTING SOUND FILES
 ---------------------
 
-splitter.py - lets you split a long recording into
-individual files while you listen to the recording.  Needs
-fast reaction times.  Works best with recordings that are
-not TOO big (max 5 to 10 minutes at a time).
-
 autosplit.py - splits a long recording into individual files
 completely automatically, but only if the recording has been
 made in near-broadcasting-studio conditions.
 
-strip0.py - strips absolute silence off the beginning and
-end of audio files (only useful if you're dealing with files
-from a textbook CD or something)
-
 manual-splitter.py (Unix only but not too difficult to
 modify for other systems) - a helper script so that you can
-use Audacity (or another sound editor) to split the file.
+use Audacity (or another sound editor) to split the file
+in non-realtime.  (Realtime splitting can be done in Gradint.)
 Use the "export selection as wav" command (you can assign a
 hot-key to it), and you don't have to type in a different
 filename each time because this script can run in the
@@ -39,23 +33,11 @@ Useful if the recording is so messy that nothing else works.
 OTHER PROCESSING
 ----------------
 
-equalise.py - adjusts the volume of all files to a similar
-level (usually increasing it also).  Use this if the volume
-of your recordings varies too much.
-
 filemove.sh - a Unix script that can help you to re-organise
 your directories while automatically reflecting those
 changes in the gradint progress database, so the
 re-organisation does not interfere with your progress so far
 
-make-smaller - some brief notes on what to do in Unix if you
-find that your collection of words is taking up too much
-hard disk space (or too much space on the backup device)
-
-email-lesson* - scripts that can help you to
-automatically distribute daily lessons to students
-using a web server with reminder emails
-
 cache-synth.py and cleanup-cache.py - cache all words that
 can be synthesized, and cleanup the cache (if you later
 remove some words from the vocabulary).  Useful if you have
@@ -65,13 +47,11 @@ you run your gradint on.
 list-synth.py and list2cache.py - aids in adding words from
 online synthesizers (or real people) to the synth cache
 
-vocab2html.py - make an HTML index for the synth
-cache, with the help of vocab.txt
-(you can also use it with espeak.cgi)
-
 transliterate.py - make a transliterated vocab report
 (for use with grep or on PDAs or whatever)
 
+player.py - play 
+
 diagram.py - make a diagram of a gradint lesson
 
-log2opl.py - translate a lesson to speaker's notes on EPOC PDA
+trace.py - make a raytraced animation of a lesson

gradint-build/samples/utils/autosplit.py → samples/utils/autosplit.py

@@ -4,13 +4,12 @@ import os, struct, sndhdr, sys
 try: import winsound
 except: winsound=None
 macsound = (sys.platform.find("mac")>=0 or sys.platform.find("darwin")>=0)
-if macsound: sys.stderr.write("Warning: You need to have qtplay (from gradint or wherever) in your PATH for this to work\n")
 
 # python 3+:
 try: input
 except: input=lambda x:eval(raw_input(x))
 
-startCount = 1   # or 0, or 485 or whatever
+startCount = 0   # or however many WAVs already exist (should be even)
 
 threshold = 10 # 3 is too low for recorded sound, but if using speech synth you might want to set it to 1
 shortestSilence = 0.3
@@ -18,17 +17,29 @@ shortestSound = 0.4
 
 if len(sys.argv)>1: exec(" ".join(sys.argv[1:])) # so you can override the above on the command line
 
+sox_8bit, sox_16bit, sox_32bit, sox_signed, sox_unsigned = "-b", "-w", "-l", "-s", "-u"
+if not winsound: # adapted from gradint (see comments there)
+  sox_formats=os.popen("sox --help 2>&1").read()
+  sf2 = ' '.join(sox_formats.lower().split())
+  if sf2.startswith("sox: sox v"):
+    if sf2[10]==' ': soxMaj=15
+    else: soxMaj = int(sf2[10:sf2.index('.')])
+  else: soxMaj=0
+  if soxMaj>=14:
+    if soxMaj==14 and sf2[13]<'4': sox_8bit, sox_16bit, sox_32bit = "-1", "-2", "-4"
+    else: sox_8bit, sox_16bit, sox_32bit, sox_signed, sox_unsigned = "-b 8", "-b 16", "-b 32", "-e signed-integer", "-e unsigned-integer" # TODO: check the last one
+
 def autosplit(filename,lang1,lang2,threshold):
     (wtype,rate,channels,wframes,bits) = sndhdr.what(filename)
     if bits==8:
-        soxBits="-b -u"
+        soxBits=sox_8bit+" "+sox_unsigned
         structBits="B"
     elif bits==16:
-        soxBits="-w -s"
+        soxBits=sox_16bit+" "+sox_signed
         structBits="h"
         threshold *= 256
     elif bits==32:
-        soxBits="-l -s"
+        soxBits=sox_32bit+" "+sox_signed
         structBits="i"
         threshold *= (256 * 256 * 256)
     else: raise Exception("Unsupported bits per sample")
@@ -47,7 +58,8 @@ def autosplit(filename,lang1,lang2,threshold):
         (sounding, bytes) = nextSample()
         if inSilence and not sounding and bytes: continue
         elif bytes:
-            dataToWriteout.append(bytes)
+            if dataToWriteout or sounding:
+                dataToWriteout.append(bytes)
             if sounding: numSilences = inSilence = 0
             else: numSilences += 1
         if numSilences >= int(shortestSilence*rate) or not bytes:
@@ -64,8 +76,9 @@ def autosplit(filename,lang1,lang2,threshold):
                 open(fname, "wb").write(''.join(dataToWriteout))
                 os.system("sox %s \"%s\" \"%s.wav\"" % (soxParams,fname,fname))
                 os.unlink(fname)
+                print fname+".wav"
                 if winsound: winsound.PlaySound(fname+".wav",winsound.SND_FILENAME)
-                elif macsound: os.system("qtplay "+fname+".wav")
+                elif macsound: os.system("afplay "+fname+".wav")
                 else: os.system("play "+fname+".wav")
             # Anyway, clear the output buffer
             dataToWriteout = []

gradint-build/samples/utils/cache-synth.py → samples/utils/cache-synth.py

-#!/usr/bin/env python
+#!/usr/bin/env python2
 
 # cache-synth.py [--test] language [language ...]
 
@@ -9,6 +9,9 @@
 # the same directory as gradint.py with all the
 # settings.
 
+# For a more advanced version of this, see the
+# actually_generate option in synth-batchconvert-helper.py
+
 import sys,os,time
 langs = sys.argv[1:] ; testMode = False
 if langs and langs[0]=='--test':

gradint-build/samples/utils/cleanup-cache.py → samples/utils/cleanup-cache.py

-#!/usr/bin/env python
+#!/usr/bin/env python2
 
 # delete cached synthesized words that are not used
 # (i.e. not mentioned in vocab.txt or samples).

gradint-build/samples/utils/diagram.py → samples/utils/diagram.py

+#!/usr/bin/env python2
 
 # diagram.py: script to generate diagrams of gradint lessons
 # (C) 2008 Silas S. Brown.  License: GPL
 
-# gradint is run normally (passing any command-line arguments on)
+# Gradint is run normally (passing any command-line arguments on)
 # and then a diagram of the lesson it made is written to diagram.svg
 
 # you can get .ps by doing: inkscape -p '> diagram.ps' diagram.svg
@@ -59,7 +60,7 @@ def CompositeEvent_draw(self,startTime,pixelsPerSec,topY,height):
 gradint.CompositeEvent.draw=CompositeEvent_draw
 
 def Event_colour(self,language):
-    if hasattr(self,"wordToCancel"):
+    if self.makesSenseToLog():
       if language==gradint.firstLanguage: return "yellow" # TODO: 2nd to 3rd lang etc?
       else: return "green"
     else: return "grey" # prompts

samples/utils/filemove.sh

+#!/bin/bash
+
+SamplesDir="samples/" # Must include trailing /
+ProgressFile="progress.txt"
+if ! [ -e $SamplesDir ]; then echo "Error: $SamplesDir does not exist (are you in the right directory?)"; exit 1; fi
+if ! [ -e $ProgressFile ]; then echo "Error: $ProgressFile does not exist (are you in the right directory?)";exit 1;fi
+
+if test "a$1" == a; then
+  echo "Usage: $0 oldname newname"
+  echo "oldname and newname are relative to $SamplesDir, and can be prefixes of several files/directories"
+  echo "Moves files from one samples directory to another, keeping $ProgressFile adjusted.  Make sure gradint is not running (including waiting for start) when in use."
+  exit 1
+fi
+
+Src=$1
+Dest=$2
+
+find "$SamplesDir" -follow -type f | grep "^$SamplesDir$Src" | \
+while true; do read || break;
+  SrcFile=$REPLY
+  DestFile=$(echo "$SrcFile"|sed -e "s|^$SamplesDir$Src|$SamplesDir$Dest|")
+  mkdir -p "$DestFile" ; rmdir "$DestFile" # ensure parent dirs exist before moving file across
+  mv -b "$SrcFile" "$DestFile"
+  SrcFile=$(echo "$SrcFile"|sed -e "s|$SamplesDir||")
+  DestFile=$(echo "$DestFile"|sed -e "s|$SamplesDir||")
+  gzip -fdc "$ProgressFile" | sed -e "s|$SrcFile|$DestFile|g" > /tmp/newprog ; mv /tmp/newprog "$ProgressFile" # (ideally should re-write to batch these changes, but leave like this for now in case need to recover from unfinished operation)
+done
+
+rmdir "$SamplesDir$Src" 2>/dev/null >/dev/null # IF it's a directory

gradint-build/samples/utils/list-synth.py → samples/utils/list-synth.py

-#!/usr/bin/env python
+#!/usr/bin/env python2
 
 # list-synth.py language [language ...]
 # list all words that can be synthesized

gradint-build/samples/utils/list2cache.py → samples/utils/list2cache.py

-#!/usr/bin/env python
+#!/usr/bin/env python2
 
 # list2cache.py language