I'd like Alexa to be able to accept a variable-length list of English letters to my custom skill. It will allow users to search based on a string.
There's two steps to this:
Getting good representation for individual letters that Alexa can understand
Enumerating sample utterances with variable number of letters
For the first, one way would be to define a custom slot that has as its enumerated values of the English alphabet:
SLOT_LETTER
ay
bee
see
dee
ee
eff
gee
... etc
but that feels hacky. Does Amazon support any way to do this or is there a cleverer way?
I'd really rather not use NATO phonetic ("alpha bravo charlie" for "A-B-C") because it's a terrible user experience and very few people actually know them.
For the second issue (sample utterances), for AMAZON.LITERAL I want to define something like:
SpellIntent find me things starting with {first second|SLOT_LETTER}
SpellIntent find me things starting with {first second third|SLOT_LETTER}
SpellIntent find me things starting with {first second third fourth|SLOT_LETTER}
But I don't think Amazon will let you define a variable length LITERAL using a custom slot (since they are different "types")?
It isn't very well documented, but you can use "a.", "b.", "c." etc to represent the letter, as opposed to the sound. Create a custom slot and use these as the values. That should do you for the slot.
For the intent, create an intent with, say, five slots, all with the same slot type. Create five utterances against the intent, with one, two, three, four and five slots filled. When the user spells something, the intent will be invoked. Any slots the user did not specify will be null.
Having two slots in one intent not separated by a word often does not perform well. But try it and see. With a restricted vocabulary like this, it could do OK.
Lastly, if it has trouble distinguishing, say, "b." and "v.", you might try adding NATO call codes to your list. Alpha, Bravo, Charlie, etc. Then, in your processing, just take the first character of whatever value comes in for the slot.
You might enable the "Star Lanes" skill and experiment with the "Set Call Sign to X Y Z" intent. I do the above in that skill and it works fairly well.
I would avoid this kind of interface because it can be difficult for users to spell things, and difficult for Alexa to differentiate letters (b or v, perhaps). But if you want to try this, consider using the literal type and asking for letters in groups of three or four.
AV: "Spell the first four letters now"
User: "see ay are tee"
AV: "You spelled C, H, R, T. Would you like to add more letters, make a correction or search now?"
User: "Cancel. I'll just use my damn phone."
Related
Got an interesting one, and can't come up with any solid ideas, so thought maybe someone else may have done something similar.
I want to be able to identify strings of letters in a longer sentence that are not words and remove them. Essentially things like kuashdixbkjshakd
Everything annoyingly is in lowercase which makes it more difficult, but since I only care about English, I'm essentially looking for the opposite of consonant clusters, groups of them that don't make phonetically pronounceable sounds.
Has anyone heard of/done something like this before?
EDIT: this is what ChatGpt tells me
It is difficult to provide a comprehensive list of combinations of consonants that have never appeared in a word in the English language. The English language is a dynamic and evolving language, and new words are being created all the time. Additionally, there are many regional and dialectal variations of the language, which can result in different sets of words being used in different parts of the world.
It is also worth noting that the frequency of use of a particular combination of consonants in the English language is difficult to quantify, as the existing literature on the subject is limited. The best way to determine the frequency of use of a particular combination of consonants would be to analyze a large corpus of written or spoken English.
In general, most combinations of consonants are used in some words in the English language, but some combinations of consonants may be relatively rare. Some examples of relatively rare combinations of consonants in English include "xh", "xw", "ckq", and "cqu". However, it is still possible that some words with these combinations of consonants exist.
You could try to pass every single word inside the sentence to a function that checks wether the word is listed inside a dictionary. There is a good number of dictionary text files on GitHub. To speed up the process: use a hash map :)
You could also use an auto-corretion API or a library.
Algorithm to combine both methods:
Run sentence through auto correction
Run every word through dictionary
Delete words that aren't listed in the dictionary
This could remove typos and words that are non-existent.
You could train a simple model on sequences of characters which are permitted in the language(s) you want to support, and then flag any which contain sequences which are not in the training data.
The LangId language detector in SpamAssassin implements the Cavnar & Trenkle language-identification algorithm which basically uses a sliding window over the text and examines the adjacent 1 to 5 characters at each position. So from the training data "abracadabra" you would get
a 5
ab 2
abr 2
abra 2
abrac 1
b 2
br 2
bra 2
brac 1
braca 1
:
With enough data, you could build a model which identifies unusual patterns (my suggestion would be to try a window size of 3 or smaller for a start, and train it on several human languages from, say, Wikipedia) but it's hard to predict how precise exactly this will be.
SpamAssassin is written in Perl and it should not be hard to extract the language identification module.
As an alternative, there is a library called libtextcat which you can run standalone from C code if you like. The language identification in LibreOffice uses a fork which they adapted to use Unicode specifically, I believe (though it's been a while since I last looked at that).
Following Cavnar & Trenkle, all of these truncate the collected data to a few hundred patterns; you would probably want to extend this to cover up to all the 3-grams you find in your training data at least.
Perhaps see also Gertjan van Noord's link collection: https://www.let.rug.nl/vannoord/TextCat/
Depending on your test data, you could still get false positives e.g. on peculiar Internet domain names and long abbreviations. Tweak the limits for what you want to flag - I would think that GmbH should be okay even if you didn't train on German, but something like 7 or more letters long should probably be flagged and manually inspected.
This will match words with more than 5 consonants (you probably want "y" to not be considered a consonant, but it's up to you):
\b[a-z]*[b-z&&[^aeiouy]]{6}[a-z]*\b
See live demo.
5 was chosen because I believe witchcraft has the longest chain of consonants of any English word. You could dial back "6" in the regex to say 5 or even 4 if you don't mind matching some outliers.
Not 100% if this is possible but I would like to convert any outbound call that does not match my DID range to a set phone number.
With our carrier in Australia if the ANI is not from their supplied range the call is blocked as part of new regulations.
What I am looking for is something like this.
if not +61 2 XXXX XXXX - +61 2 XXXX XXXX then send as +612XXXX XXXX
I apologise I have no true understanding of regex and do not know even where to begin.
I am starting to work on my knowledge of it though. please be kind. If anyone can point me to an "idiots guide" link I would be appreciative as I am just getting into this.
Of course it's possible. It's just a matter of how much work you want to do. I'm not quite sure what you want to mask and what you want to pass on unmutilated. A couple of particular examples would help. How many different formats, countries, and so on do you need to support?
With these problems, I tend to follow this approach:
Normalize the data. Make them all look the same. So, remove all non-digits, for example. +61 2 XXXX XXXX turns into 612XXXXXXXX. In this step, you'd also fill in implicit information, like a local number that does not include the country code. Number::Phone may be interesting, but, also note is was the largest distro on CPAN for awhile.
Now it should be easier to recognize the number and it's components (because if it isn't, you didn't do Step 1 right). Instead of a regex, you might use a parser. That is, get the country code, and then from that, decide what has to happen next. That's the sort of thing I have to do with ISBNs in Business::ISBN, which have a group code then a publisher code (both of which are variable length.
Once you can recognize the number, it's easy to select a range. If it's in the range, you know what to replace.
I recently started working with ontologies and I am using Protege to build an ontology which I'd also like to use for automatically classifying strings. The following illustrates a very basic class hierarchy:
String
|_ AlphabeticString
|_ CountryName
|_ CityName
|_ AlphaNumericString
|_ PrefixedNumericString
|_ NumericString
Eventually strings like Spain should be classified as CountryName or UE4564 would be a PrefixedNumericString.
However I am not sure how to model this knowledge. Would I have to first define if a character is alphabetic, numeric, etc. and then construct a word from the existing characters or is there a way to use Regexes? So far I only managed to classify strings based on an exact phrase like String and hasString value "UE4565".
Or would it be better to safe a regex for each class in the ontology and then classify the string in Java using those regexes?
An approach that might be appropriate here, especially if the ontology is large/complicated or might change in the future, and assuming that some errors are acceptable, is machine learning.
An outline of a process utilizing this approach might be:
Define a feature set you can extract from each string, relating to your ontology (some examples below).
Collect a "train set" of strings and their true matching categories.
Extract features from each string, and train some machine-learning algorithm on this data.
Use the trained model to classify new strings.
Retrain or update your model as needed (e.g. when new categories are added).
To illustrate more concretely, here are some suggestions based on your ontology example.
Some boolean features that might be applicable: does the string matches a regexp (e.g the ones Qtax suggests); does the string exist in a prebuilt known city-names list; does it exist in a known country-names list; existence of uppercase letters; string length (not boolean), etc.
So if, for instance, you have a total of 8 features: match to the 4 regular expressions mentioned above; and the additional 4 suggested here, then "Spain" would be represented as (1,1,0,0,1,0,1,5) (matching the first 2 regular expressions but not the last two, is a city name but not a country name, has an uppercase letter and length is 5).
This set of feature will represent any given string.
to train and test a machine learning algorithm, you can use WEKA. I would start from rule or tree based algorithms, e.g. PART, RIDOR, JRIP or J48.
Then the trained models can be used via Weka either from within Java or as an external command line.
Obviously, the features I suggest have almost 1:1 match with your Ontology, but assuming your taxonomy is larger and more complex, this approach would probably be one of the best in terms of cost-effectiveness.
I don't know anything about Protege, but you can use regex to match most of those cases. The only problem would be differentiating between country and city name, I don't see how you could do that without a complete list of either one.
Here are some expressions that you could use:
AlphabeticString:
^[A-Za-z]+\z (ASCII) or ^\p{Alpha}+\z (Unicode)
AlphaNumericString:
^[A-Za-z0-9]+\z (ASCII) or ^\p{Alnum}+\z (Unicode)
PrefixedNumericString:
^[A-Za-z]+[0-9]+\z (ASCII) or ^\p{Alpha}+\p{N}+\z (Unicode)
NumericString:
^[0-9]+\z (ASCII) or ^\p{N}+\z (Unicode)
A particular string is an instance, so you'll need some code to make the basic assertions about the particular instance. That code itself might contain the use of regular expressions. Once you've got those assertions, you'll be able to use your ontology to reason about them.
The hard part is that you've got to decide what level you're going to model at. For example, are you going to talk about individual characters? You can, but it's not necessarily sensible. You've also got the challenge that arises from the fact that negative information is awkward (as the basic model of such models is intuitionistic, IIRC) which means (for example) that you'll know that a string contains a numeric character but not that it is purely numeric. Yes, you'd know that you don't have an assertion that the instance contains an alphabetic character, but you wouldn't know whether that's because the string doesn't have one or just because nobody's said so yet. This stuff is hard!
It's far easier to write an ontology if you know exactly what problems you intend to solve with it, as that allows you to at least have a go at working out what facts and relations you need to establish in the first place. After all, there's a whole world of possible things that could be said which are true but irrelevant (“if the sun has got his hat on, he'll be coming out to play”).
Responding directly to your question, you start by checking whether a given token is numeric, alphanumeric or alphabetic (you can use regex here) and then you classify it as such. In general, the approach you're looking for is called generalization hierarchy of tokens or hierarchical feature selection (Google it). The basic idea is that you could treat each token as a separate element, but that's not the best approach since you can't cover them all [*]. Instead, you use common features among tokens (for example, 2000 and 1981 are distinct tokens but they share a common feature of being 4 digit numbers and possibly years). Then you have a class for four digit numbers, another for alphanumeric, and so on. This process of generalization helps you to simplify your classification approach.
Frequently, if you start with a string of tokens, you need to preprocess them (for example, remove punctuation or special symbols, remove words that are not relevant, stemming, etc). But maybe you can use some symbols (say, punctuation between cities and countries - e.g. Melbourne, Australia), so you assign that set of useful punctuation symbols to other symbol (#) and use that as a context (so the next time you find an unknown word next to a comma next to a known country, you can use that knowledge to assume that the unknown word is a city.
Anyway, that's the general idea behind classification using an ontology (based on a taxonomy of terms). You may also want to read about part-of-speech tagging.
By the way, if you only want to have 3 categories (numeric, alphanumeric, alphabetic), a viable option would be to use edit distance (what is more likely, that UA4E30 belongs to the alphanumeric or numeric category, considering that it doesn't correspond to the traditional format of prefixed numeric strings?). So, you assume a cost for each operation (insertion, deletion, subtitution) that transforms your unknown token into a known one.
Finally, although you said you're using Protege (which I haven't used) to build your ontology, you may want to look at WordNet.
[*] There are probabilistic approaches that help you to determine a probability for an unknown token, so the probability of such event is not zero. Usually, this is done in the context of Hidden Markov Models. Actually, this could be useful to improve the suggestion given by etov.
This might be a hard one (if not impossible), but can anyone think of a regular expression that will find a person's name, in say, a resume? I know this won't be 100% accurate, but I can't come up with something.
Let's assume the name only shows up once in the document.
No, you can't use regular expressions for this. The only chance you have is if the document is always in the same format and you can find the name based on the context surrounding it. But this probably isn't the case for you.
If you are asking your applicants to submit their résumé online you could provide a separate field for them to enter their name and any other information you need instead of trying to automatically parse résumés.
Forget it - seriously.
Or expect to get a lot of applications from a Mr C Vitae
In my experience, having written something very similar (but a very long time ago), about 95% of resumes have the person's name as the very first line. You could probably have a pretty loose regex checking for alpha, hyphens, periods, and assume that's the name.
Obviously there's no way to do this 100% accurately, as you said, but this would be close.
Unless you wanted to build an expression that contained every possible name, or-ed together, the expression you are referring to is not "Regular," with a capital R. A good guess might be to go looking for the largest-font words in the document. If they follow a pattern that looks like firstname-lastname, name-initial-name, etc., you could call it a good guess...
That's a really hairy problem to tackle. The regex has to match two words that could be someone's name. The problem with that is that some people, of Hispanic origin, for example, might have a name that's more than 2 words. Also, how would you define two words to match for a name? Would you use a database of common first and last name fields? That might work unless someone has an uncommon name.
I'm reminded of a story of a COBOL teacher in college told me about an individual of Asian origin who's name would break every rule the programmers defined for a bank's internal system. His first name was "O." just the letter O.
The only remotely dependable way to nail down the regex would be if you had something to set off your search with; maybe if a line of text in the resume began with "Name: " then you'd know where to start looking.
tl;dr: People's names and individual resumes are too heavily varied for a regular expression to pick apart.
You could do something like Amazon does for book overviews: SIPs. This would require some after-the-fact double checking by humans but you might find the person's name(s) in there.
Although this seems like a trivial question, I am quite sure it is not :)
I need to validate names and surnames of people from all over the world. Imagine a huge list of miilions of names and surnames where I need to remove as well as possible any cruft I identify. How can I do that with a regular expression? If it were only English ones I think that this would cut it:
^[a-z -']+$
However, I need to support also these cases:
other punctuation symbols as they might be used in different countries (no idea which, but maybe you do!)
different Unicode letter sets (accented letter, greek, japanese, chinese, and so on)
no numbers or symbols or unnecessary punctuation or runes, etc..
titles, middle initials, suffixes are not part of this data
names are already separated by surnames.
we are prepared to force ultra rare names to be simplified (there's a person named '#' in existence, but it doesn't make sense to allow that character everywhere. Use pragmatism and good sense.)
note that many countries have laws about names so there are standards to follow
Is there a standard way of validating these fields I can implement to make sure that our website users have a great experience and can actually use their name when registering in the list?
I would be looking for something similar to the many "email address" regexes that you can find on google.
I sympathize with the need to constrain input in this situation, but I don't believe it is possible - Unicode is vast, expanding, and so is the subset used in names throughout the world.
Unlike email, there's no universally agreed-upon standard for the names people may use, or even which representations they may register as official with their respective governments. I suspect that any regex will eventually fail to pass a name considered valid by someone, somewhere in the world.
Of course, you do need to sanitize or escape input, to avoid the Little Bobby Tables problem. And there may be other constraints on which input you allow as well, such as the underlying systems used to store, render or manipulate names. As such, I recommend that you determine first the restrictions necessitated by the system your validation belongs to, and create a validation expression based on those alone. This may still cause inconvenience in some scenarios, but they should be rare.
I'll try to give a proper answer myself:
The only punctuations that should be allowed in a name are full stop, apostrophe and hyphen. I haven't seen any other case in the list of corner cases.
Regarding numbers, there's only one case with an 8. I think I can safely disallow that.
Regarding letters, any letter is valid.
I also want to include space.
This would sum up to this regex:
^[\p{L} \.'\-]+$
This presents one problem, i.e. the apostrophe can be used as an attack vector. It should be encoded.
So the validation code should be something like this (untested):
var name = nameParam.Trim();
if (!Regex.IsMatch(name, "^[\p{L} \.\-]+$"))
throw new ArgumentException("nameParam");
name = name.Replace("'", "'"); //' does not work in IE
Can anyone think of a reason why a name should not pass this test or a XSS or SQL Injection that could pass?
complete tested solution
using System;
using System.Text.RegularExpressions;
namespace test
{
class MainClass
{
public static void Main(string[] args)
{
var names = new string[]{"Hello World",
"John",
"João",
"タロウ",
"やまだ",
"山田",
"先生",
"мыхаыл",
"Θεοκλεια",
"आकाङ्क्षा",
"علاء الدين",
"אַבְרָהָם",
"മലയാളം",
"상",
"D'Addario",
"John-Doe",
"P.A.M.",
"' --",
"<xss>",
"\""
};
foreach (var nameParam in names)
{
Console.Write(nameParam+" ");
var name = nameParam.Trim();
if (!Regex.IsMatch(name, #"^[\p{L}\p{M}' \.\-]+$"))
{
Console.WriteLine("fail");
continue;
}
name = name.Replace("'", "'");
Console.WriteLine(name);
}
}
}
}
I would just allow everything (except an empty string) and assume the user knows what his name is.
There are 2 common cases:
You care that the name is accurate and are validating against a real paper passport or other identity document, or against a credit card.
You don't care that much and the user will be able to register as "Fred Smith" (or "Jane Doe") anyway.
In case (1), you can allow all characters because you're checking against a paper document.
In case (2), you may as well allow all characters because "123 456" is really no worse a pseudonym than "Abc Def".
I would think you would be better off excluding the characters you don't want with a regex. Trying to get every umlaut, accented e, hyphen, etc. will be pretty insane. Just exclude digits (but then what about a guy named "George Forman the 4th") and symbols you know you don't want like ##$%^ or what have you. But even then, using a regex will only guarantee that the input matches the regex, it will not tell you that it is a valid name.
EDIT after clarifying that this is trying to prevent XSS: A regex on a name field is obviously not going to stop XSS on its own. However, this article has a section on filtering that is a starting point if you want to go that route:
s/[\<\>\"\'\%\;\(\)\&\+]//g;
"Secure Programming for Linux and Unix HOWTO" by David A. Wheeler, v3.010 Edition (2003)
v3.72, 2015-09-19 is a more recent version.
BTW, do you plan to only permit the Latin alphabet, or do you also plan to try to validate Chinese, Arabic, Hindi, etc.?
As others have said, don't even try to do this. Step back and ask yourself what you are actually trying to accomplish. Then try to accomplish it without making any assumptions about what people's names are, or what they mean.
I don’t think that’s a good idea. Even if you find an appropriate regular expression (maybe using Unicode character properties), this wouldn’t prevent users from entering pseudo-names like John Doe, Max Mustermann (there even is a person with that name), Abcde Fghijk or Ababa Bebebe.
You could use the following regex code to validate 2 names separeted by a space with the following regex code:
^[A-Za-zÀ-ú]+ [A-Za-zÀ-ú]+$
or just use:
[[:lower:]] = [a-zà-ú]
[[:upper:]] =[A-ZÀ-Ú]
[[:alpha:]] = [A-Za-zÀ-ú]
[[:alnum:]] = [A-Za-zÀ-ú0-9]
It's a very difficult problem to validate something like a name due to all the corner cases possible.
Corner Cases
Anything anything here
Sanitize the inputs and let them enter whatever they want for a name, because deciding what is a valid name and what is not is probably way outside the scope of whatever you're doing; given the range of potential strange - and legal names is nearly infinite.
If they want to call themselves Tricyclopltz^2-Glockenschpiel, that's their problem, not yours.
A very contentious subject that I seem to have stumbled along here. However sometimes it's nice to head dear little-bobby tables off at the pass and send little Robert to the headmasters office along with his semi-colons and SQL comment lines --.
This REGEX in VB.NET includes regular alphabetic characters and various circumflexed european characters. However poor old James Mc'Tristan-Smythe the 3rd will have to input his pedigree in as the Jim the Third.
<asp:RegularExpressionValidator ID="RegExValid1" Runat="server"
ErrorMessage="ERROR: Please enter a valid surname<br/>" SetFocusOnError="true" Display="Dynamic"
ControlToValidate="txtSurname" ValidationGroup="MandatoryContent"
ValidationExpression="^[A-Za-z'\-\p{L}\p{Zs}\p{Lu}\p{Ll}\']+$">
This one worked perfectly for me in JavaScript:
^[a-zA-Z]+[\s|-]?[a-zA-Z]+[\s|-]?[a-zA-Z]+$
Here is the method:
function isValidName(name) {
var found = name.search(/^[a-zA-Z]+[\s|-]?[a-zA-Z]+[\s|-]?[a-zA-Z]+$/);
return found > -1;
}
Steps:
first remove all accents
apply the regular expression
To strip the accents:
private static string RemoveAccents(string s)
{
s = s.Normalize(NormalizationForm.FormD);
StringBuilder sb = new StringBuilder();
for (int i = 0; i < s.Length; i++)
{
if (CharUnicodeInfo.GetUnicodeCategory(s[i]) != UnicodeCategory.NonSpacingMark) sb.Append(s[i]);
}
return sb.ToString();
}
This somewhat helps:
^[a-zA-Z]'?([a-zA-Z]|\.| |-)+$
This one should work
^([A-Z]{1}+[a-z\-\.\']*+[\s]?)*
Add some special characters if you need them.