# Tagged: formal languages

# Myhill-Nerode Theorem in Practice

When it comes to practical side of computer science, we often work with **regular** and **context-free** languages. Regular languages are most common for expressing syntax through the widely used **regular expressions**. Somewhat stronger context-free grammars dominate in the field of compilers and various language processors.

As we know from the Chomsky hierarchy of formal languages, the class of regular languages is a subset (or a sort-of a special case) of context-free language class. This means that every regular language can be described not only by finite automatons, regular expressions or regular grammars but also by context-free grammars, pushdown automatons and also (for the hard-core folks) a turing machine.

In theory this is fine, but it’s not very good for practice. I mean, if you were developing some app that requires syntax checking, would you rather develop some sort of lineary bounded automaton or use a library for regular expressions?

Sometimes we want to prove, that we don’t need to do magic to check the correct syntax of an email address. There are some ways of doing so. For example **pumping theorem**, which is fairly straight forward. But pumping lemma can be only used to prove, that a language is **not **regular. The lemma doesn’t provide sufficient condition for a language to be regular. That’s where the folks Myhill and Nerode come in!

The Myhill-Nerode theorem on the contrary provides necessary and sufficient condition for the language to be regular! Yay! But as you can imagine it’s not that simple :-P.

## Formal definition

**Definition 1.** Let Σ be an alphabet and ~ a equivalence relation on Σ^{*}. Then relation ~ is right invariant if for all u, v, w ∈ Σ^{*}:

- u ~ v <=> uw ~ vw

**Definition 2.** Let L be a language (not neccessarily regular) over Σ. We define relation ~_{L} called **prefix equivalence** on Σ^{*} as follows:

- u ~
_{L}v <=> ∀w ∈ Σ^{*}: uw ∈ L <=> vw ∈ L

**Theorem 1.** (*Myhill-Nerode*) Let L be a language over Σ. Then these three statements are equivalent:

- L is accepted by some deterministic finite automaton.
- L is the union of some of the equivalence classes of a
**right invariant**equivalence relation of finite index. - Relation ~
_{L}is of finite index.

## Informal description

The heart of this theorem is the fact that finite automaton has only a finite number of states. With this, any language that can be described by a finite automaton must consist of only finite number of string patterns. This is expressed by the right invariant relation with finite index.

The prefix equivalence is a form of the right invariant equivalence, but stronger. It’s tied to the respective language and says that if some strings are ~_{L}-equivalent they both are included or excluded from the language after we extend them. The Myhill-Nerode theorem says, that a regular language always has a finite number of equivalence classes, i.e. there is only a finite number of word patterns that can be repeated through the string.

Example proof

Now a little example of how to show, that a language is not regular by using this theorem. Let’s take for instance the most classic context-free language of all times L = { a^{n}b^{n} | n >= 0 }.

We’ll be interested in the third part of Myhill-Nerode theorem which states, that **relation ~ _{L} is of finite index**. We need to find a way of showing that it’s not and therefore the language is not regular (since Myhill-Nerode theorem is an equivalence).

In this case, the most elegant way is proof by contradiction. We will show, that L = { a^{n}b^{n} | n >= 0 } is not reglar.

**Proof 1.** Suppose that L is a regular language. Let be two natural numbers so . Then consider words and a sequence of words over a^{1}, a^{2}, …, a^{i}.

Now let’s assign some values to strings *u*, *v* and *w*:

.

According to the definition of prefix equivalence (**definition 2**),

we can see, that none of the words *a ^{i}* from the sequence are

**~**-equivalent. The sequence is infinite, so the

_{L}**index of the relation will be infinite**. But that’s a contradiction with the Myhill-Nerode theorem. Therefore, the language is

**not**regular.

## Sources

- http://www.mec.ac.in/resources/notes/notes/automata/Myhill%20theorm.htm
- http://cs.wikipedia.org/wiki/Myhillova-Nerodova_v%C4%9Bta
- Češka M., Vojnar T., Smrčka A.. Teoretická informatika: studijní opora. FIT VUT v Brně. 2010.

# Language in Computer Science

The last post was about strings, so now something about **languages**. From the theoretical point of view a *language* is a set of words or more precisely sentences (strings). Possibly and usually an infinite set. Only sentences that are present in the set are part of the language, nothing else. Neat feature of languages is, that they interconnect two separate formal theories/models into one thing. Let me explain.

Language is a **set**. As in mathematical set. So anything that applies to sets applies to languages as well (like union, intersection, complement, Cartesian product). Members of any language are only **strings**. And there are some operations defined over strings (that is concatenation, reversal, sub-strings etc.). We can take advantage of this fact and abstract those, so they work with language as well. There is generally **a lot** you can do with a formal language either from the set or string point of view.

**Definition 1.** A **formal language** *L* over an alphabet Σ is a subset of Σ^{*}. Formally, L ⊆ Σ^{*}. The magical Σ^{*} denotes a set of all existing words and sentences over Σ. Sometimes it’s called an alphabet iteration.

Like I said earlier, all operations defined for sets work with languages as well. For example **language union** L_{1} ∪ L_{2} = {s; s ∈ L_{1} ∨ s ∈ L_{2}} etc. More interesting are the string operations that can be abstracted so they work on languages. There is** language concatenation**:

**Definition 2.** Let L_{1} be a language over Σ_{1} and L_{2 }be a language over Σ_{2}. **Concatenation** of these two L = L_{1}⋅L2 = {xy; x ∈ L_{1}, y ∈ L_{2}} is a language over alphabet Σ = Σ_{1} ∪ Σ_{2}.

Language concatenation as well as the string equivalent is not commutative operation. There is a couple of special cases:

- L ⋅ {ε} = L
- L ⋅ ∅ = ∅

The first, {ε} is a language with a single word — an empty string, in the second case we’re talking about an empty set. Concatenation can be looked at a sort of multiplication between the two languages (remember Cartesian product?). As a result of this thought, we define **language power**.

**Definition 3.** Let *L* be a language over Σ, then *L ^{n}* denotes n-power of language and

- L
^{0}= {ε} - L
^{n+1}=L^{n}⋅L

With language power defined, we can advance to the last operation called **language iteration**. It’s basically an abstraction of language power.

**Definition 4**. L^{+} denotes positive iteration of language L defined accordingly

- L
^{+}= L ∪ L^{2}∪ … ∪ L^{∞}

**Definition 5**. L^{*} denotes iteration of language L defined accordingly

- L
^{*}= {ε} ∪ L ∪ L^{2}∪ … ∪ L^{∞}

## Summary

In this post were introduced some of the most basic and most common operations that are defined on languages. The most important thing to remember is, that language is a **set** of **sentences** (or words) and the operations that we define originate from these two areas.

We have some set operations like **union**, **intersection**, **complement **as for generic sets. And we have abstractions for string operations like **concatenation** (notice, that it’s somewhat similar to cartesian product), **language power** and **iteration**.

Anyway, these couple of operations is nowhere near to the complete list of all existing functions for languages! After all, if you dive into some algebraic structures a little you can easily define your own :-).

# Basic Computer Science

There is a couple of very basic definitions and axioms in computer science. I consider them to be very important, because everything that comes later is based on them. And if you don’t fully comprehend the basic stuff, it will be very hard to understand anything further. That’s why I decided to write a whole post on these trivial definitions.

## Alphabet

Yeah, I’m not kidding. There’s a formal definition for alphabet. And what’s worse: it actually makes sense. I just wonder what would a kinder-gardener say on this :-D. Anyway, here it goes:

**Definition 1.1.** An **alphabet** is a finite non-empty set of *symbols.*

Alphabets are usually denoted by Greek big sigma * Σ*. A set

*Σ*can be referred to as alphabet when it’s not empty, but also not infinitely big.

*Σ = {a, b, c}*is an alphabet*Σ = {0, 1}*is an alphabet*Σ = {}*is**NOT**an alphabet*Σ = all integers*; is**NOT**an alphabet

Content of an alphabet are *symbols*. A symbol is some indivisible or atomic unit that can have some meaning (not necessarily). For example, if* Σ = {righteous, dude} *were alphabet, **righteous** and **dude** would be considered to be atomic indivisible elements (even though they contain multiple characters). Some folks also choose to omit the *finite* word from the definition. But that’s for some very advanced stuff and I’m guessing that finite alphabets will be more than sufficient for us.

## String

Another fundamental thing is a string. **String** or a **word** is a sequence of some symbols (the order is important). Strings usually contain symbols from only one alphabet. In that case we say string over is a finite sequence of symbols from . The formal definition is this:

**Definition 1.2**. Let be an alphabet.

- empty string is string over
- if is a string over and is a symbol from , then is a string over
*y*is string over only when we can derive it by applying rules 1 and 2.

There are some more basic definitions like string length, concatenation, reversation that are useful and important, but they might be a little intimidating at first, so let’s skip ahead to languages.

## Language

What is a **formal language**? We need to define first. is a set of all existing strings over alphabet . For example, let , then all positive binary numbers .

Then a formal language **L** over alphabet is a subset of . Formally said, . Basically, language is a *set of strings*. In context of programming languages, a language would be a set of all possible programs in that programming language. You see, that in most practical cases the set will be **infinite** (as it is in the programming language case). So, it’s not very practical to describe languages by enumerating all possible sentences of the language. One way of describing a language is with things called **formal grammars**. But we’ll get on to that later.

## Sources

# Introduction to Computer Science

In the upcoming semester I’ll be taking class called Theoretical Computer Science. Which is said to be *the hardest* thing you can attend here at BUT. Only half the people pass the bar every year. It’s brutal. And since I’d really like to be a part of the lucky half, I thought I could dig into the theory a little earlier and see how bad it is.

So** Computer Science**, right? What the hell is all that about? *Theoretical computer science* or *theoretical information technology* (as referred by some people) is a formal foundation for the things we like to call **computers**. What computers do? They essentially compute stuff. Current computer hardware is built and programmed to work with numbers. Who cares about a bunch of numbers? But the numbers represent some sort of **information**. In broader terms, computers work with information. A computer takes some information and transforms it into another information. What Wikipedia says:

Computer sciencedeals with the theoretical foundations of information, computation, and with practical techniques for their implementation and application. Computer scientists invent algorithmic processes that create, describe, and transforminformationand formulate suitable abstractions to model complex systems.

Right, so we have information. How do people work with these anyway? People share their thoughts though **languages**. We talk and sometimes listen, we also read and sometimes write. Other people like to draw, play charades or sing. In all of these cases the information is encoded into some sort of *language* that others can understand. Sure, we’re people, it comes naturally to us, but what about the machines? A coffeemaker will most certainly not learn to talk to you or even understand your needs. Machines are dumb. That’s where we (the nerdy guys from engineering department) come in with** formal languages** — a substantial part of computer science. By formalizing common languages we make the machines understand our instructions.

Ever heard of a **programming language**? Programming language is basically a sequence of instructions we use when we tell the computers what to do. It’s a language that we use when we talk to the computers. It sounds a little weird, but that’s it. Computer science sets up some ground rules so the computers can algorithmically analyze the instructions and process them. It explores possibilities of computers — what can be processed by a computer? Is there anything that computers can’t solve, why? Bunch of interesting stuff.

Interesting, but sometimes very hard to understand. I wanted to put some basics here too, but I don’t want to scare you off too early. I got stuck for a while with the very basics at first (I figured it out eventually). The math will come in a stand-alone post shortly after. Stay tuned ;-).