Posts Tagged Mac OS X

Another Haiku Detector Update, and Some Observations on Mac Speech Synthesis

Screenshot of Haiku Detector

I subjected Haiku Detector to some serious stress-testing with a 29MB text file (that’s 671481 sentences, containing 16810 haiku, of which some are intentional) a few days ago, and kept finding more things that needed fixing or could do with improvement. A few days in a nerdsniped daze later, I have a new version, and some interesting tidbits about the way Mac speech synthesis pronounces things. Here’s some of what I did:

  • Tweaked the user interface a bit, partly to improve responsiveness after 10000 or so haiku have been found.
  • Made the list of haiku stay scrolled to the bottom so you can see the new ones as they’re found.
  • Added a progress bar instead of the spinner that was there before.
  • Fixed a memory issue.
  • Changed a setting so it should work in Mac OS X 10.6, as I said here it would, but I didn’t have a 10.6 system to test it on, and it turns out it does not run on one. I think 10.7 (Lion) is the lowest version it will run on.
  • Added some example text on startup so that it’s easier to know what to do.
  • Made it a Developer ID signed application, because now that I have a bit more time to do Mac development (since I don’t have a day job; would you like to hire me?), it was worth signing up to the paid Mac Developer Program again. Once I get an icon for Haiku Detector, I’ll put it on the app store.
  • Fixed a few bugs and made a few other changes relating to how syllables are counted, which lines certain punctuation goes on, and which things are counted as haiku.

That last item is more difficult than you’d think, because the Mac speech synthesis engine (which I use to count syllables for Haiku Detector) is very clever, and pronounces words differently depending on context and punctuation. Going through words until the right number of syllables for a given line of the haiku are reached can produce different results depending on which punctuation you keep, and a sentence or group of sentences which is pronounced with 17 syllables as a whole might not have words in it which add up to 17 syllables, or it might, but only if you keep a given punctuation mark at the start of one line or the end of the previous. There are therefore many cases where the speech synthesis says the syllable count of each line is wrong but the sum of the words is correct, or vice versa, and I had to make some decisions on which of those to keep. I’ve made better decisions in this version than the last one, but I may well change things in the next version if it gives better results.

Here are some interesting examples of words which are pronounced differently depending on punctuation or context:

ooohh Pronounced with one syllable, as you would expect
ooohh. Pronounced with one syllable, as you would expect
ooohh.. Spelled out (Oh oh oh aitch aitch)
ooohh… Pronounced with one syllable, as you would expect
H H Pronounced aitch aitch
H H H Pronounced aitch aitch aitch
H H H H H H H H Pronounced aitch aitch aitch
Da-da-de-de-da Pronounced with five syllables, roughly as you would expect
Da-da-de-de-da- Pronounced dee-ay-dash-di-dash-di-dash-di-dash-di-dash. The dashes are pronounced for anything with hyphens in it that also ends in a hyphen, despite the fact that when splitting Da-da-de-de-da-de-da-de-da-de-da-de-da-da-de-da-da into a haiku, it’s correct punctuation to leave the hyphen at the end of the line:


Though in a different context, where – is a minus sign, and meant to be pronounced, it might need to go at the start of the next line. Greater-than and less-than signs have the same ambiguity, as they are not pronounced when they surround a single word as in an html tag, but are if they are unmatched or surround multiple words separated by spaces. Incidentally, surrounding da-da in angle brackets causes the dash to be pronounced where it otherwise wouldn’t be.

U.S or u.s Pronounced you dot es (this way, domain names such as are pronounced correctly.)
U.S. or u.s. Pronounced you es
US Pronounced you es, unless in a capitalised sentence such as ‘TAKE US AWAY’, where it’s pronounced ‘us’

I also discovered what I’m pretty sure is a bug, and I’ve reported it to Apple. If two carriage returns (not newlines) are followed by any integer, then a dot, then a space, the number is pronounced ‘zero’ no matter what it is. You can try it with this file; download the file, open it in TextEdit, select the entire text of the file, then go to the Edit menu, Speech submenu, and choose ‘Start Speaking’. Quite a few haiku were missed or spuriously found due to that bug, but I happened to find it when trimming out harmless whitespace.

Apart from that bug, it’s all very clever. Note how even without the correct punctuation, it pronounces the ‘dr’s and ‘st’s in this sentence correctly:

the dr who lives on rodeo dr who is better than the dr I met on the st john’s st turnpike

However, it pronounces the second ‘st’ as ‘saint’ in the following:

the dr who lives on rodeo dr who is better than the dr I met in the st john’s st john

This is not just because it knows there is a saint called John; strangely enough, it also gets this one wrong:

the dr who lives on rodeo dr who is better than the dr I met in the st john’s st park

I could play with this all day, or all night, and indeed I have for the last couple of days, but now it’s your turn. Download the new Haiku Detector and paste your favourite novels, theses, holy texts or discussion threads into it.

If you don’t have a Mac, you’ll have to make do with a few more haiku from the New Scientist special issue on the brain which I mentioned in the last post:

Being a baby
is like paying attention
with most of our brain.

But that doesn’t mean
there isn’t a sex difference
in the brain,” he says.

They may even be
a different kind of cell that
just looks similar.

It is easy to
see how the mind and the brain
became equated.

We like to think of
ourselves as rational and
logical creatures.

It didn’t seem to
matter that the content of
these dreams was obtuse.

I’d like to thank the people of the xkcd Time discussion thread for writing so much in so many strange ways, and especially Sciscitor for exporting the entire thread as text. It was the test data set that kept on giving.

, , , , , , , , , , , , ,

1 Comment

Haiku Detector Update

On Monday I posted a quick-and-dirty Haiku Detector Mac application I’d written which finds haiku (in terms of syllable counts and line breaks, not aesthetics) in any given text. Since then I’ve made it less dirty and maybe more quick. It now shows progress when it’s busy looking for haiku in a long text, and gives you a count of the sentences it looked at and the haiku it found. You can also copy all the haiku (Copy All Haiku in the Edit menu) or save them to a file (Save in the File menu.) Here’s where you can download the new version, which should still work on Mac OS X 10.6 and later. And here are a few more haiku I’ve found with it.

There’s only one (not counting a by-line) in the feature articles of the April 27 edition of New Scientist:

Inside a cosy
new gut the eggs hatch and the
cycle continues.

From Flatland: a romance of many dimensions, by Edwin Abbott Abbott:

On the reply to
this question I am ready
to stake everything.

“I come,” said he, “to
proclaim that there is a land
of Three Dimensions.”

Man, woman, child, thing—
each as a Point to the eye
of a Linelander.

This was the Climax,
the Paradise, of my strange
eventful History.

Here are a few more from Flatland which I’m editing this post to add, since I liked them more on the second reading:

Let us begin by
casting back a glance at the
region whence you came.

Therefore, pray have done
with this trifling, and let us
return to business.

Even if I were
a baby, I could not be
so absurd as that.

From Last Chance to See, by Douglas Adams and Mark Carwardine, which I somehow ended up with a text file of many years ago and eventually got a book of:

I’ve been here for five
days and I’m still waiting for
something to go right.

We each went off to
our respective rooms and sat
in our separate heaps.

They’re nocturnal birds
and therefore very hard to
find during the day.

It looked like a great
horn-plated tin opener
welded to its face.

We keep searching for
more females, but we doubt if
there are any more.

The very laws of
physics are telling you how
far you are from home.

Foreigners are not
allowed to drive in China,
and you can see why.

`Just the one left,’ she
said, putting it down on the
ground in front of her.

Yet it was hunted
to extinction in little
more than fifty years.

And conservation
is very much in tune with
our own survival.

And here’s my own haiku about a particularly amusing passage in that book:

Here Douglas Adams
trudges through his anagram:
Sago mud salad.

Charles Darwin’s most popular work, The Formation of Vegetable Mould Through the Action of Worms with Observations on their Habits, only contained 12 mostly-lacklustre haiku, but I like to think this one is a metaphor:

Worms do not always
eject their castings on the
surface of the ground.

Something about lack of worm castings being only skin-deep.

But most of these don’t mention nature or seasons, as haiku should. So here are some from Sylva, or A Discourse of Forest Trees:

Dieu, and thence rode to
Blois and on to Tours, where he
stayed till the autumn.

How graphic, and how
refreshing, is the pithy
point thus neatly scored—

Meteorology, or Weather Explained, by J.G. M’Pherson contains some very poetic-sounding unintentional haiku:

“It’ll pe aither
ferry wat, or mohr rain”—a
poor consolation!

“Beware of rain” when
the sheep are restive, rubbing
themselves on tree stumps.

The brilliant flame, as
well as the smoky flame, is
a fog-producer.

Till ten o’clock the
sun was not seen, and there was
no blue in the sky.

But, strange to say, there
is a healing virtue in
breathing different air.

There is much pleasure
in verifying such an
interesting problem.

Unfortunately, there are no haiku in Dijkstra’s ‘Go To Statement Considered Harmful‘.

The app still uses a lot of memory if you process a novel or two, and may have trouble saving files in that case; It looks like it’s a bug in the speech synthesis library (or my use of it) or simply a caching strategy that doesn’t work well when the library is used in this rather unusual and intensive way. Anyway, if you ever try to save a file and the Save dialog doesn’t appear, try copying instead, and relaunch the program.

Next I think I’ll experiment with finding the best haiku based on the parts of speech at the ends of lines. But first, I’d better start working on the thing I’ve plan to do for the six of hearts.

If you’ve found any nice unintentional haiku, or if you can’t run Haiku Detector yourself but have ideas for freely-available texts it could be run on, let me know in the comments.

, , , , , , , , ,


Five of Hearts: Haiku Detector

A few weeks ago, a friend linked to Times Haiku, a website listing unintentional haiku found in The New York Times, saying ‘I’d actually pay for a script that could check for Haiku in my writings. That would make prose-production a lot more exciting! Who’s up to the script-writing-challenge?’

I knew I could do it, having written syllable-counting code for my robot choir (which I really need to create an explanation page about.) I told her I’d make it that weekend. That was last weekend, when I decided at the last moment to write an article about neutron stars and ISOLTRAP, and then chickened out of that and wrote a poem about it. So I put off the haiku program until yesterday. It was fairly quick to write, so here it is: Haiku Detector. It should work on Mac OS X 10.6 and above. Just paste or type text into the top part of the window, and any detected haiku will appear in the bottom part.

Haiku Detector looks for sentences with seventeen syllables, and then goes through the individual words and checks whether the sentence can be split after the fifth and twelfth syllables without breaking a word in half. Then it double-checks the last line still has five syllables, because sometimes the punctuation between words is pronounced. The Times Haiku-finding program has a database of syllable counts per word, but I didn’t need that since I can use the Mac OS X speech synthesis API to count the syllables. Haiku Detector makes no attempt to check for kigo (season words.)

The first place I looked for haiku was the Wikipedia page for Haiku in English. Due to the punctuation, it didn’t actually find any of the example haiku on the page, but it did find this:

Robert Spiess (Red Moon
Anthology, Red Moon Press,

How profound. Next, having declared myself contributing troubadour for New Scientist magazine, I fed this week’s feature articles through it, and found:

A pill that lowers
arousal doesn’t teach shy
people what to do

Meanwhile, there are signs
that the tide is turning in
favour of shyness.

So by 4000
years ago, the stage was set
for the next big step.

This heat makes the air
spin faster, so pulling the
storm towards the city.

Some will be cooler
and less humid — suitable
for outdoor sports, say.

The last ones seem almost seasonal.

I needed to stress-test the app with a large body of text, so I grabbed the first novel of which I had the full text handy: John Scalzi‘s Old Man’s War, which I had on my iPad on my lap to read while my code was compiling. This book has at least one intentional haiku in it, which Haiku Detector detected. Apart from that, some of my favourites are:

I hate that her last
words were “Where the hell did I
put the vanilla.”

As I said, this is
the place where she’s never been
anything but dead.

“I barely know him,
but I know enough to know
he’s an idiot.”

She’d find me again
and drag me to the altar
like she had before.

A gaper was not
long in coming; one swallow
and Susan was in.

They were nowhere to
be found, an absence subtle
and yet substantial.

And it stares at me
like it knows something truly
strange has just happened.

I haven’t got up to that fifth one in the novel yet, but it mentions a swallow, which I understand is (when accompanied by more swallows) a harbinger of Spring or Summer depending on which language you get your idioms from, so there’s the kigo.

Next I figured I should try some scientific papers — the kinds of things with words that the Times haiku finder would not have in its syllable database. You probably can’t check this unless your workplace also provides access to Physics Letters B, but I can assure you that the full text of the ISOLTRAP paper about neutron stars does not contain any detectable haiku. However, the CMS paper announcing the discovery of the boson consistent with the Higgs does:

In the endcaps, each
muon station consists of
six detection planes.

As is usual for CMS papers, the author and institute lists are about as long as the paper itself, and that’s where most of the haiku were too. Here are a few:

LHC Higgs Cross Section
Working Group, in: S.

of California, Davis,
Davis, USA

That’s ‘one hundred and two’ in case anyone who doesn’t say it that way was wondering.

And here are some from my own blog. I used the text from a pdf I made of it before the last JoCo Cruise Crazy, so the last few months aren’t represented:

Beds of ground cover
spread so far in front of him
they made him tired.

Apologies to
those who only understand
half of this poem.

I don’t remember
what colour he said it was,
but it was not green.

His eyes do not see
the gruesome manuscript scrawled
over the white wall.

• Lines 1 to 3 have
four syllables each, with stress
on the first and last.

(That’s not how you write a haiku!)

I don’t wear armour
and spikes to threaten you, but
to protect myself.

A single female
to perpetuate the genes
of a thousand men.

Kerblayvit is a
made-up placeholder name, and
a kerblatent cheat.

He wasn’t the first,
but he stepped on the moon soon
after Neil Armstrong.

He just imagined
that in front of him there was
a giant dunnock.

(there are plenty more where that one came from, at the bottom of the page)

She was frustrated
just trying to remember
what the thing was called.

Please don’t consider
this a failing; it is part
of your programming.

While writing this program, I discovered that that the speech API now has an easier way to count syllables, which wasn’t available when I wrote the robot choir. The methods I used to separate the text into sentences and the view I used to display the haiku are also new. Even packaging the app for distribution was different. I don’t get to write Mac software often enough these days.

Yet again, I didn’t even bother to deal out the cards because I already had something to inspire me. In my halfhearted attempt to find a matching card, I came across one about electronics in the service of ALICE, so I ran the latest instalment of Probably Never, by Alice, into it, and got this:

Or well, I have to
put up with getting called a
fake girl all the time.

The jackhole who called
me a “he/she” recognized
that he crossed the line.

If that sounds interesting, subscribe to Probably Never, and I could probably forward you the rest of that episode if you want.

And finally, two unintentional haiku from this very post:

Haiku Detector
makes no attempt to check for
kigo (season words.)

(there are plenty more
where that one came from, at the
bottom of the page)

Wait; make that three!

And finally, two
unintentional haiku
from this very post:

Have fun playing with Haiku Detector, and post any interesting haiku you find in the comments. Also, let me know of any bugs or other foibles it has; I wrote it pretty quickly, so it’s bound to have some.

I know what I’m doing for the six of hearts; I’ve planned it for a long time but still haven’t actually started it. It’s musical, so it will probably be terrible; brace yourselves. By the way, I keep forgetting to mention, but They Might Not Be Giants will be published in Offshoots 12. Yay!

, , , , , , , , , , ,


%d bloggers like this: