Misspelling Can Bite You

[Gene Wirchenko]

Dear JavaScripters:

Once again, it is time to thank me for pointing a possible source
of errors in JavaScript code or to chuckle over how I messed up again,
maybe both.

I misspelled the name of the control variable for a switch. I
had
switch (this.DateType)
when it should have been
switch (this.DataType)

JavaScript was quite happy to execute the code. After all, the
switch might have had a case to handle undefined.

Sincerely,

Gene Wirchenko

[Richard Cornford]

Really? Which browser, because the language's syntax won't stand for a -
case - clause inside a Block statement (and a - switch - isn't much use
without at least one - case - clause), so there should have been a
syntax error to prevent compilation, and so prevent any execution.

However, the situation you describe, of mistyping a keyword and still
having code execute is a possibility. One example would be mistyping -
if - to give:-

id(x)
{
y = 12;
}

- which would produce a runtime error, "id is not a function", rather
than a syntax error.

That (and possibly your code above) raises a point about the placement
of braces, because if you do:-

id(x){
y = 12;
}

- with the opening brace on the end of the same line as the mistyped -
if - keyword and its expression, you get a syntax error instead of a
runtime error. That is because the absence of a carriage return prevents
the language's automatic semi-colon insertion from acting to 'correct'
the syntax of the latter in the way it 'corrected' the syntax of the
former.

Personally I prefer the coding style represented in that final example,
but I have to admit that over the years I have been looking for a (any?)
justification for choosing that coding style over its alternatives and
have been unable to find anything substantial to indicate a choice of
any one common style over another with regard to the placement of
braces.

This syntax error in preference to a runtime error (syntax errors are
always reported while runtime errors only occurred if the code
containing them is actually executed, and that is difficult to guarantee
during testing) is pretty much the only justification for a 'brace at
the end of the first line' coding style that I have seen, and there may
be direct counter examples that exists but haven't yet been presented,
or occurred, to me.

Incidentally, you should find yourself an editor with a customisable
syntax highlighting system and have the keywords colored such that they
are distinct from all other sequences of characters in the code. That
way it becomes very visually apparent if you mistype a keyword, and you
get to spot/find the mistakes sooner/easier and so fix them more
quickly.

Richard.

[Gene Wirchenko]

On Fri, 3 Feb 2012 01:37:31 -0000, "Richard Cornford"
<Ric...@litotes.demon.co.uk> wrote:

>Gene Wirchenko wrote:
>> Dear JavaScripters:
>>
>> Once again, it is time to thank me for pointing a possible
>> source of errors in JavaScript code or to chuckle over how I
>> messed up again, maybe both.
>>
>> I misspelled the name of the control variable for a switch.
>> I had
>> switch (this.DateType)
>> when it should have been
>> switch (this.DataType)
>>
>> JavaScript was quite happy to execute the code. After all,
>> the switch might have had a case to handle undefined.
>
>
>Really? Which browser, because the language's syntax won't stand for a -

IE 9.

>case - clause inside a Block statement (and a - switch - isn't much use
>without at least one - case - clause), so there should have been a
>syntax error to prevent compilation, and so prevent any execution.

Huh? The sequence was syntactically correct. The only problem
with it was a misspelling of a property name. That corrected, the
code ran fine.

>However, the situation you describe, of mistyping a keyword and still
>having code execute is a possibility. One example would be mistyping -
>if - to give:-

I described the result of misspelling a property name, not a
keyword.

[snip]

Sincerely,

Gene Wirchenko

[Thomas 'PointedEars' Lahn]

Richard Cornford wrote:

> Gene Wirchenko wrote:
>> Once again, it is time to thank me for pointing a possible
>> source of errors in JavaScript code or to chuckle over how I
>> messed up again, maybe both.
>>
>> I misspelled the name of the control variable for a switch.
>> I had
>> switch (this.DateType)
>> when it should have been
>> switch (this.DataType)
>>
>> JavaScript was quite happy to execute the code.

JavaScript may not even have been used.

>> After all,
>> the switch might have had a case to handle undefined.
>
> Really? Which browser, because the language's syntax won't stand for a -

There is not *the* language with regard to browsers, unless you consider
"JavaScript" to be a stand-in for "ECMAScript (implementations)", which is
wrong.

> case - clause inside a Block statement (and a - switch - isn't much use
> without at least one - case - clause), so there should have been a
> syntax error to prevent compilation, and so prevent any execution.

I think he meant

switch (this.bar)
{
case undefined:
…
}

where the object referred to by `this' had a `foo' property instead. Only
that in this case `foo' and `bar' were much more similar.

Barring `this' referring to a host object or `bar' being a specially
designed accessor property¹, any conforming implementation of ECMAScript Ed.
1 to 5.1 would execute that code without throwing an exception, with or
without the `case undefined' clause, because you can attempt to access any
property of the object referred to by `this' or the object converted from
the value of `this' (ES 5.x strict mode) without it throwing an exception.
Accessing a non-existing property merely results in `undefined'. (I am
presently not aware of any feature in ES 5.1 that would change that
behavior. Strict mode is not it. SpiderMonkey's __noSuchMethod__ property²
only allows it for method calls. ES Harmony is going to allow it with
dynamic proxies, but not on regular objects.³)

¹ Object.defineProperty(this, "bar", {
get: function () {
throw new TypeError();
}
});

² this.__noSuchMethod__ = function (id, args) {
throw new Error(id + "(" + args.join(", ") + ")");
};

this.bar();

³ var o = Proxy.create({
get: function (rcvr, name) {
if (name == "bar")
{
throw new TypeError();
}
}
}, this);

o.bar;

See also
<http://www.youtube.com/watch?v=A1R8KGKkDjU&feature=plcp&context=C38a9548UDOEgsToPDskJEdhctNqhQUmYwp-
S69WbK>

> However, the situation you describe, of mistyping a keyword and still
> having code execute is a possibility. One example would be mistyping -
> if - to give:-
>
> id(x)
> {
> y = 12;
> }
>
> - which would produce a runtime error, "id is not a function", rather
> than a syntax error.

> […]

While this is valuable information indeed, I think you have misunderstood
the OP.


PointedEars
--
Sometimes, what you learn is wrong. If those wrong ideas are close to the
root of the knowledge tree you build on a particular subject, pruning the
bad branches can sometimes cause the whole tree to collapse.
-- Mike Duffy in cljs, <news:Xns9FB6521286...@94.75.214.39>

[Dr J R Stockton]

In comp.lang.javascript message <po1mi75fb0u5rqp9or413nv26799lpnndj@4ax.
com>, Thu, 2 Feb 2012 14:02:39, Gene Wirchenko <ge...@ocis.net> posted:

Putting the following script element last, or making it a body onload
function, will list all the script words of more than 2 letters, in
alphabetical order, with the number of times used. The result will be
browser-dependent; but not gravely so. Demo in my js-misc1.htm.


<script type="text/javascript">

var skripts = document.getElementsByTagName("SCRIPT")
var L = skripts.length, Str = "", Out = []
while (L--) Str += skripts[L].innerHTML
Str = Str.split(/\W+/).sort()
var was = "", K = 0
L = Str.length
while (L--) {
if (Str[L] == was) K++ ; else {
if (was.length > 2 && /^[a-z]/i.test(was)) Out.push(K + " " + was)
K = 1 ; was = Str[L] } }
document.write("<pre>" + Out.reverse().join("\n") + "<\/pre>")

</script>


The code could, no doubt, be optimised. The output might better be to a
textarea or a new page. It could be well to NOT scan words in strings
or comment.


Code could be added to pick out cases where two adjacent words are
unreasonably similar.

--
(c) John Stockton, nr London, UK. ?@merlyn.demon.co.uk Turnpike v6.05.
Website <http://www.merlyn.demon.co.uk/> - w. FAQish topics, links, acronyms
PAS EXE etc. : <http://www.merlyn.demon.co.uk/programs/> - see in 00index.htm
Dates - miscdate.htm estrdate.htm js-dates.htm pas-time.htm critdate.htm etc.

[Mel Smith]

Dr. J said:

> Code could be added to pick out cases where two adjacent words are
> unreasonably similar.

Dr J and Gene:

This morning I was caught (and mislead) by the following set of two
statements:

var rgn = document.getElementById("iselectrgn") ;
var rgnopts = rgn.getElementsByTagname("option") ;

Because I *thought* I had checked the statements thoroughly, I looked
everywhere *after* this set, and finally went back and checked them again
letter-by-letter, and then ...

-Mel Smith

[Dr J R Stockton]

In comp.lang.javascript message <9p5i9v...@mid.individual.net>, Sat,
4 Feb 2012 13:21:14, Mel Smith <med_cuto...@aol.com> posted:

Readily found by Firefox/Opera/Safari Error Console, Chrome JavaScript
Console, and by an IE8 pop-up, though.

--
(c) John Stockton, nr London, UK. ?@merlyn.demon.co.uk Turnpike 6.05 WinXP.
Web <http://www.merlyn.demon.co.uk/> - FAQ-type topics, acronyms, and links.
Command-prompt MiniTrue is useful for viewing/searching/altering files. Free,
DOS/Win/UNIX now 2.0.6; see <URL:http://www.merlyn.demon.co.uk/pc-links.htm>.

[Mel Smith]

Dr. J said:
>>
>> var rgn = document.getElementById("iselectrgn") ;
>> var rgnopts = rgn.getElementsByTagname("option") ;
>>
>> Because I *thought* I had checked the statements thoroughly, I looked
>>everywhere *after* this set, and finally went back and checked them again
>>letter-by-letter, and then ...
>

> Readily found by Firefox/Opera/Safari Error Console, Chrome JavaScript
> Console, and by an IE8 pop-up, though.

Yes, I use IE 7 on my dev machine, and 'once-in-a-while' (when I'm
nearing the production stage) I check out the appearance on Chrome and FF.

However, I didn't *this* time and spent a couple of confusing and
frustrating hours tracking down this 'devil'.

-Mel Smith

[Dr J R Stockton]

In comp.lang.javascript message <9pagtq...@mid.individual.net>, Mon,
6 Feb 2012 10:28:19, Mel Smith <med_cuto...@aol.com> posted:

One fault with those Error Consoles is that they lack a "Stay On Top"
checkbox, or one for "Jump up when written to".

Re my earlier code : given a reference to a script element, one can get
.innerHTML and (in all my browsers) read as a string the script as
transmitted. Is there any easy way of getting that string, MINUS all
comment and MINUS all inner strings (or, in each case, removing the
content of the comments and strings) while leaving all the other code?