Eliminate Forward Slashes

Rivet

I have my forms in pretty good shape now and pretty well understand what's going on at each step of the way.

But (there's always a but), I've reached another level of ignorance. I'm trying to filter the Message input that's entered. I've pretty well handled most everything (strip, validate, sanitize, etc.) and I can NOT find a way to damage a URL so it won't function, but will still be recognizable, at least to me. I'd like to eliminate "http:// completely, change "www" to "vvv" and replace forward slashes with a dash. 😕
No filters are available that I can figure out.
So I've tried things like string replace, preg replace, substr replace, and a couple things I've probably left out.

What would you consider the "best" but not complex way of doing this and what function, assuming a function would be best?

Your thoughts would be most appreciated.

Regards, & TIA,

Rivet`

bradgrafelman

What you described can be accomplished with a single call to [man]str_replace/man (or [man]str_ireplace/man). What did you try?

Rivet

Ah, thank you! That's just what I needed! I missed str_ireplace() completely - that'll help a lot and shorten the line considerably.
I think I get too close to the trees sometimes and lose track of the forest!
It's much clearer now and even works! lol.

Thanks,

Rivet`

Rivet

Well, I'm baack!! I said it was working, and is is: partly. Problem is, str_ireplace() is stopping midway through the text I'm feeding it. No errors, no notices or warnings; just an incomplete application of the function.

The easiest way I could think of for testing was to remove vowels from text in the Message Textarea. It succeeds with a nice, short message (named "comments" in my code). To start off with I used a lorem ipson paste for the text. About a third+ part of the way through, the replacements stopped. Figureing there might be some foreign text or something in it I switched to my own text, entered right here on my keyboard, for the Message; same thing; OK for short messages, stopped replacing about half way through. So, I've missed something!

For example, if I use the following in the textarea, it works:
"
This is some code to see how ireplace works in removing vowels from $comments.
Therefore, once str_ireplace() is applied, there should be no vowels in it.
"

But, if I extend the text as so, about the same number of characters as Lorum Ipsum I first used:
"
This is some code to see how ireplace works in removing vowels from $comments.
Therefore, once str_ireplace() is applied, there should be no vowels in it.
This is some code to see how ireplace works in removing vowels from $comments.
Therefore, once str_ireplace() is applied, there should be no vowels in it.
This is some code to see how ireplace works in removing vowels from $comments.
Therefore, once str_ireplace() is applied, there should be no vowels in it.

This is some code to see how ireplace works in removing vowels from $comments.
Therefore, once str_ireplace() is applied, there should be no vowels in it.
This is some code to see how ireplace works in removing vowels from $comments.
Therefore, once str_ireplace() is applied, there should be no vowels in it.
"

Then it fails to do all of the replacements partway through. Here are the replaced results:
"
COMMENTS
Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts. Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t. Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts. Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t. Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts. Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t. Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts. Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t. Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts. Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t. This is some code to see how ireplace works in removing vowels from $comments. Therefore, once str_ireplace() is applied, there should be no vowels in it. This is some code to see how ireplace works in removing vowels from $comments. Therefore, once str_ireplace() is applied, there should be no vowels in it. This is some code to see how ireplace works in removing vowels from $comments. Therefore, once str_ireplace() is applied, there should be no vowels in it. This is some code to see how ireplace works in removing vowels from $comments. Therefore, once str_ireplace() is applied, there should be no vowels in it. This is some code to see how ireplace works in removing vowels from $comments. Therefore, once str_ireplace() is applied, there should be no vowels in it.
"

I italicised the text where it stopped working, for clarity.

Here's the code I'm using:

$comments=$_POST['comments'];
echo "COMMENTS"."<br />";
str_replace("/",$_POST['comments'],"");
  $comments= stripslashes($_POST['comments']);
  trim($comments);
$vowels = array("a", "e", "i", "o", "u", "A", "E", "I", "O", "U");
$onlyconsonants = str_replace($vowels, "", "$comments");
echo $onlyconsonants;
  $comments = strip_tags($comments);
  $comments = htmlspecialchars($comments);
echo $comments."<br />";

I've read & re-read the manual & other places and they all seem to indicate this should work, including the "gotchas" to no avail.

What am I missing?

Regards,

Rivet`

bradgrafelman

It didn't stop working, you just discarded its results. You echo $onlyconsonants once, but then you go back to using $comments after that.

Here are a few other problems I noticed with your example code:

```
str_replace("/",$_POST['comments'],"");
```
This statement does absolutely nothing. (Perhaps you meant to store the return value into a variable?)
```
$comments= stripslashes($_POST['comments']);
```
You should never need to call [man]stripslashes/man on incoming data. If incoming data has slashes automatically prepended in front of certain character (see [man]addslashes/man), then the solution is to prevent PHP from trashing the data in the first place. See the manual page [man]security.magicquotes[/man] for more info.
```
  trim($comments);
```
Again, this statement does absolutely nothing. Same issue as in item #1.
```
$vowels = array("a", "e", "i", "o", "u", "A", "E", "I", "O", "U");
$onlyconsonants = str_replace($vowels, "", "$comments");
```
Note that the "_ireplace" variant would cut your array size in half, or you could also use [man]preg_replace/man with a regexp pattern like '/[aeiou]/i'.

$comments = strip_tags($comments);

Why strip_tags()? What if I want to use less-than and greater-than symbols to depict placeholders? For example, why would you want to turn this:

When using <insert programming language> for the first time, make a 'Hello World' program!

into this:

When using  for the first time, make a 'Hello World' program!

Rivet

It didn't stop working, you just discarded its results. You echo $onlyconsonants once, but then you go back to using $comments after that.

But ... WHY, when I use ' echo "$onlyconsonants" ', does that echo result in only a partial removal of vowels? Again, the relevant code is"

$vowels = array("a", "e", "i", "o", "u", "A", "E", "I", "O", "U");
$onlyconsonants = str_replace($vowels, "", "$comments");
echo $onlyconsonants;

I just removed all the extraneous $comments echoes, leaving only the vowel str_ireplace and echo $consonants and reran the code: the results are exactly the same; Only partial removal of vowels.

THAT is the echo that returns only partial removal of the vowels. $onlyconsonants should be $comments with the vowels removed, should it not?

Here are a few other problems I noticed with your example code:
1.
PHP Code:

str_replace("/",$_POST['comments'],"");

Quite right: Its a leftover from a previous testing attempt before I turned to you again. It should not (and doesn't) exist any longer.
This statement does absolutely nothing. (Perhaps you meant to store the return value into a variable?)

2.
PHP Code:

$comments= stripslashes($_POST['comments']);
You should never need to call stripslashes() on incoming data. If incoming data has slashes automatically prepended in front of certain character (see addslashes()), then the solution is to prevent PHP from trashing the data in the first place. See the manual page security.magicquotes for more info.

Hmm, that's an interesting thought. For a text area though it seems logical to me at least in this case. It's intended to be a "Plain Text" type of entry, with a max of ~240 words (forgt how many characters that's based on).
What pre=pended slash might I expect to see? I AM interested in any details so the question is far from moot.

3.
PHP Code:

trim($comments);

Again, this statement does absolutely nothing. Same issue as in item #1.

Does nothing because of other

4.
PHP Code:

$vowels = array("a", "e", "i", "o", "u", "A", "E", "I", "O", "U");
$onlyconsonants = str_replace($vowels, "", "$comments");

Note that the "_ireplace" variant would cut your array size in half, or you could also use preg_replace() with a regexp pattern like '/[aeiou]/i'.

Agreed; I knew better, just left the code alone for testing; shouldn't have done that. Course I should've cleaned up other things, too.

5.
PHP Code:

$comments = strip_tags($comments);

Why strip_tags()? What if I want to use less-than and greater-than symbols to depict placeholders? For example, why would you want to turn this:

Code:
When using <insert programming language> for the first time, make a 'Hello World' program! into this:

Good Point; I agree.

Code:

When using for the first time, make a 'Hello World' program! ?

Actually, I did, and it worked fine; just like the example I showed initially. When it's short and sweet, it works. When it gets past a certain length, which I noticed happens to a a double line-feed, the replacement stops working.

i'm sorry if I'm not making this clear; there is SOMETHING that makes the replacements actually stop after a certain "whatever" (like a C/R?) happens. It has/had nothing to do with the extraneious $comments echoes. As it is now, the only code not commented out is the $consonants stuff; it does't complete the vowel replacemtn if the data in the textarea is too long. I'm breaking a law somewhere but darned if I know where.

Regards,

bradgrafelman

Rivet;11029637 wrote:
But ... WHY, when I use ' echo "$onlyconsonants" ', does that echo result in only a partial removal of vowels?

It doesn't - your example proved that. You went from this:

This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.
This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.
This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.

This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.
This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.

to this:

Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts. Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t. Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts. Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t. Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts. Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t. Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts. Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t. Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts. Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.

as per the "replaced results" you posted in reply #4.

Re-inserting the line breaks where they belong, that jumbled mess looks like:

Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts.
Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.
Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts.
Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.
Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts.
Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.

Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts.
Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.
Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts.
Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.

which is what you started with sans all the vowels.

Rivet

I don 't recall ever seeing, OR pasting anything even similar to a completely vowel-removed message if it was over 4 lines or so in length. Thus, I have NO idea where the successful removal quote came from; it wasn't me, that's for sure. THIS is the only result I've ever had OR posted as the failed result;

"
[FONT=Verdana]Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts. Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t. Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts. Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t. Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts. Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t. Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts. Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t. Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts. Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t. This is some code to see how ireplace works in removing vowels from $comments. Therefore, once str_ireplace() is applied, there should be no vowels in it. This is some code to see how ireplace works in removing vowels from $comments. Therefore, once str_ireplace() is applied, there should be no vowels in it. This is some code to see how ireplace works in removing vowels from $comments. Therefore, once str_ireplace() is applied, there should be no vowels in it. This is some code to see how ireplace works in removing vowels from $comments. Therefore, once str_ireplace() is applied, there should be no vowels in it. This is some code to see how ireplace works in removing vowels from $comments. Therefore, once str_ireplace() is applied, there should be no vowels in it. [/FONT]
"

I respectfully submit that for some reason we are at odds with each other and I don't appreciate your intimating I did something I did not. Perhaps you missed that the code you refer to is what WORKED, not what happened when I copy/pasted it to a size a little over twice the length of what you indicated. I never posted a resulting vowel removal that worked over the entire content of the messge when it was longer than some limitation I'm not aware of.

Sorry if I sould critical, but I do believe in speaking my mind; it's just how I am. And I know it can irritate some people.

Regards,

Rivet`

bradgrafelman

Take a look at the first 535 characters you posted there between the double quotes. That's exactly what I posted above.

EDIT: Assuming you still don't believe me, try executing the following code snippet (which is based entirely on your post in reply #4 above):

// execute the following via the CLI SAPI (on the console)
// no need to complicate things using a web browser since we're not generating an HTML document

$comments='This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.
This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.
This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.

This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.
This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.';

echo "COMMENTS\n";
$vowels = array("a", "e", "i", "o", "u", "A", "E", "I", "O", "U");
$onlyconsonants = str_replace($vowels, "", "$comments");
echo $onlyconsonants;

I get the following as output:

COMMENTS
Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts.
Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.
Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts.
Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.
Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts.
Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.

Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts.
Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.
Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts.
Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.

EDIT2: I'll also point out again why we're only considering the first 535 characters of the output you posted:

bradgrafelman;11029605 wrote:
It didn't stop working, you just discarded its results. You echo $onlyconsonants once, but then you go back to using $comments after that.

That last sentence is of the most importance; your code snippet above from reply #4 was outputting the data twice, but only the first one used the copy of the data where you removed all of the vowels. The second output was based on manipulations of the original data (not the copy you just outputted that had all of the vowels removed).

Rivet

bradgrafelman;11029667 wrote:
Take a look at the first 535 characters you posted there between the double quotes. That's exactly what I posted above.

EDIT: Assuming you still don't believe me, try executing the following code snippet (which is based entirely on your post in reply #4 above):
// execute the following via the CLI SAPI (on the console)
// no need to complicate things using a web browser since we're not generating an HTML document

$comments='This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.
This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.
This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.

This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.
This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.';

[B]Yes, that works. But, it's the SHORT version of $comments(now $onlyconsonants, NOT the longer one. I have a hunch it's what you are seeing, not me. Something sure isn't jiving here.[/B]

echo "COMMENTS\n";
$vowels = array("a", "e", "i", "o", "u", "A", "E", "I", "O", "U");
$onlyconsonants = str_replace($vowels, "", "$comments");
echo $onlyconsonants;
I get the following as output:

Notice the Echo is for $onlyconsonants! As I mentioned earlier, I have removed the Echoes for $Comments.

code]onlyconsonants
Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts.
Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.
Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts.
Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.
Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts.
Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.

Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts.
Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.
Ths s sm cd t s hw rplc wrks n rmvng vwls frm $cmmnts.
Thrfr, nc str_rplc() s ppld, thr shld b n vwls n t.[/code]

That should have failed, and does, on my machine, at the first /n.

EDIT2: I'll also point out again why we're only considering the first 535 characters of the output you posted:
That last sentence is of the most importance; your code snippet above from reply #4 was outputting the data twice, but only the first one used the copy of the data where you removed all of the vowels. The second output was based on manipulations of the original data (not the copy you just outputted that had all of the vowels removed).

That's incorrect: On my last check, 692 characters worked: lengthening the textarea input to 1385 characters failed.

Andk the cli doesn't interest me at the moment: a browser is what's going to see and interpret what the server sends back sI that's where I need to keep my attention at this moment.

I have no idea why we're having such a disconnect; I think the only thing to do is to re-title and start a new post, woulddn't you agree? Witha more concise , complete and accurate description of the situation. I never meant for this to drift so far from the slashes to the state it's in now anyway. You've taught me a great deal throughout this discussion and I certainly appreciate that.

Regards,

Rivet`

bradgrafelman

Rivet;11029671 wrote:
That's incorrect: On my last check, 692 characters worked: lengthening the textarea input to 1385 characters failed.

Can you give me some sample input as well as the exact code you're using?

Rivet;11029671 wrote:
Andk the cli doesn't interest me at the moment: a browser is what's going to see and interpret what the server sends back sI that's where I need to keep my attention at this moment.

Okay, but when you're outputting text that contains line breaks and then copying-and-pasting what the browser renders, it's really hard to read since you lose all of those line breaks.

Rivet;11029671 wrote:
I have no idea why we're having such a disconnect

Do you at least agree that the output you posted in reply #4 is exactly as expected? At that point, your sample code had two echo's: one to output a copy of the original data without vowels, and one to output a copy of the original data with tags stripped as well as replacing applicable characters with their HTML-safe entities (both of which resulted in no change based on the sample data you gave).

Rivet;11029671 wrote:
I thin k the only thing to do is to re-title and start a new, more concise and accurate of the situation.

Feel free, but make sure you provide a set of input data, a code snippet that completely and fully illustrates what you're doing to that data, a copy of the output you get, and either a copy of what you expected to get or at least an explanation that describes it. Perhaps then you'll realize that you truly are getting the correct result above (which is that PHP isn't a human and doesn't randomly decide it's bored after str_ireplace()'ing a certain number of characters and just decides to leave the rest untouched).

bradgrafelman

Apologies for the double post, but I just wanted to add one more illustration and leave it at that.

Using the following sample code:

$sample = 'This is some code to see how ireplace works in removing vowels from $comments. 
Therefore, once str_ireplace() is applied, there should be no vowels in it.
';
$num_times = 50;

$comments = str_repeat($sample, $num_times);
$vowels = array("a", "e", "i", "o", "u", "A", "E", "I", "O", "U");
$onlyconsonants = str_replace($vowels, "", "$comments");

file_put_contents("original.txt", $comments);
file_put_contents("modified.txt", $onlyconsonants);

are you suggesting that $num_times can be adjusted such that the text placed in "modified.txt" is no longer correct? If so, can you give me an example of what value would do that? Using the given example of 50 times, the resulting text in "original.txt" is 7,900 characters in length (including all CRLF line breaks), and the text in "modified.txt" is still correct.