gMp_iWyIJ_E

so go there and welcome to another
tutorial my name is Henry Bakshi and
this time we're gonna be going over well
searching through data sets and so this
time let's just say we have a little
data set with a few people's names or
movies names or any data set now we want
to search for something inside of that
data sets now I already know I have
associated visas part two on this which
there will be a link to in the
description but this time it's a bit
different in Swift databases part two
and also in this method I have my own
little package where I just essentially
keyword search meaning I take the thing
that you're searching for everything
inside of the data set and everything in
the data set that contains what you're
searching for will become result now I'm
also doing it in this method but there's
a catch I'm using a Levenstein string
distance algorithm from a set of code
org in order to rank those final results
and give it to you in order of relevance
and so what's essentially going to
happen let's say you have a data set
like something like test
Frank and tester just to show you how it
works now let's say you want to search
for everything in this data set that has
the word test in how would you do it
first of all let's just say this is my
function my search function okay now
what you need to give it in is a data
set and something to search or a search
search word I'll just write that a bit
better so you actually understand that
so it needs a search word what you want
to search inside of that data set and
then this function will output a final
array of results
so let's say you were to input this data
set into the search function this data
set sorry and let's say you want to
search for the word test so this is what
we're searching for
how would we do it inside of the search
function now well first of all we're
going to do is we're going to go into
each keyword first of all we're going to
rank will actually determine does test
contain the word test yes it does so
this is a successful candidate next does
Tanmay contain the word test no it does
not
does Frank contain the word test no it
does not
does tester contain the word test yes it
does now what can we do in order to see
which one of these two elements is more
relevant and so just as a little example
I can also put the word testing and this
would also be as well considered as a
candidate because it has the word test
in it now the thing is how are we going
to rank each individual word well first
of all we need to understand is we have
to use the Levenstein string distance
algorithm and again one more time this
is from rosetta code org and there will
be a link to it in the description and
so what happens is the levenshtein
distance algorithm will tell us how many
characters at minimum we need to change
in a string to convert it into another
string for example if I wanted to do
testing and test
how many care Kurt sorry how many
characters wanting to change in testing
in order to get tests well we can easily
see that if we were to remove in we
would have test this means three
characters removed that means it has a
Levenstein
score of three and this means that the
more score it has the less likely it is
the less score has the more likely it is
and if it's a score of zero it's spot on
accurate and so all we really need to do
is create a dictionary first of all to
actually an array in the dictionary an
array that says okay
there's test tester and testing so let's
just write these down test tester and
testing and then in a dictionary we all
right there Levenstein scores for test
it's zero for tester it's two and for
testing is three we know it's two here
because er removed that's two characters
are moved and we get test three
characters are moved and we get test now
all we really need to do is rank the
array actually sort the array because
dictionaries can't be sorted they're
just completely random sort the array
depending on their actual dictionary
values and that is least at the top most
at the bottom and you're done display
that in the UI tableview and you're done
searching and so theoretically our app
currently if we give it this data set
and research for test it will give us
these three results in this exact order
so now that I've actually explained the
algorithm to you why not actually go to
the Mac part and show you the code
a little demonstration and then there
will also be source code if you want to
check it out in the description yourself
so without further ado let's get
straight to the Mac part so welcome back
to the Mac part now I'm going to
actually give you a quick demo of the
app then we'll go to the code of the app
and see exactly how it works
again we are using a custom string
matching just a little if statement and
then we are using the 11g I now agree
them from rosetta code org which will
rank our results by relevance so let's
just search for as you can see quickly
though I have a UI tableview with tons
of movies names I have a little text
view over here where I can search for
movies so let's just say I wanted to
search for the movie with the word
sponge in it sorry sponge and as you can
see it gives us the two results sponge
on water and the spongebob squarepants
movie and the thing is due to the fact
that sponge out of water has sponge but
very few other characters it's saying
this is more relevant because it's
mainly sponge however since the
spongebob squarepants movie has sponge
but a lot of extra characters saying
maybe it's not that much sponge as it is
others and so it's ranking that next
after sponge out of water however if I
were to search for something like the
word thee as you can see lots of movies
have the word D in it and it gives us
all of these movies names next after
thee is another feature this is case
sensitive so if I put small T it gives
us these movies because they have a
small team with their look and so that's
what gives us these movies and that's
why first of all this is case sensitive
it will rank by relevance as I said and
it will use a custom search matching
meaning all results must contain this
word in order to be able to be ranked as
a be made as a candidate and that was a
simple demo and one more thing if you
were to remove everything from the text
view and click on enter it realizes that
there's nothing
the text view and just resets the entire
thing okay so now let's actually get to
the code of this app as you can see I've
warped over here to the code let me just
make this a bit smaller and move my face
around I don't want my face to be that
big and let's continue now so as you can
see for the sole this is the UI over
here
and our main storyboard file right as it
loads up sometimes Xcode takes a while
okay so you can see this is a textview
first of all pretty self-explanatory
move this up a bit this textview will
actually has a delegate to our view
controller no referencing outlet because
we really don't need it because when
it's a delegate we can just say okay
textview did return or it should return
and it will give us the textview i mean
field to get the text out of so there's
no iboutlet required here however from
the table view that's an entirely
different story we have a datasource
hooked up to our view controller
delegate object to our view controller
and a referencing outlet name table view
hooked up to our view controller as well
now comes the coding part of things so
which is the harder part and before I
explain this class or anything else that
I'm doing here I just want to bring your
attention attention sorry to one
function called search it now first of
all I don't need this code so delete
this next continuing this is the search
function you give it a list of strings
and something to compare in that list of
strings and it returns a list of strings
that it thinks in relevance are
potential things that the user is
searching for and the way this works is
first of all we're defining this levy
I'm going to say this one more time is
from rosetta code org a link will be in
the description and possibly on screen
right now continuing as you can see we
are first creating our final
stringer these are the final candidates
which will be then ranked by our
Levenstein this ins algorithm next what
we're doing is we are looping through
this entire list which is the list that
you gave us which is the list of movies
in this case and what we're doing is
we're checking if the current element
were on in the list contains the string
that you're comparing to then append
this into our finals meaning take as a
candidate next create a finals int
dictionary and so this is going to do is
this is essentially going to say okay
this candidate has a Levenstein distant
score of this much and again as I
explained in the whiteboard whiteboard
part the lower the levenshtein distance
score the better this result is meaning
the less characters are going to change
in a string in order to get a string
that's why it's called the distance
algorithm and then what we're doing is
we're looping through our finals and
we're saying okay in finals int set this
candidate score to the levenshtein
distance of compare verse that that
specific candidate so we're getting the
Levenstein distance of you what you want
to compare and the candidate and putting
it into finals int for that candidate
next what we're doing is we're sorting
finals in place finals not sort in place
self-explanatory by finals int dollar
mark 0 less than finals into dollar mark
1 so this is doing is if this element
the first element over here is smaller
than the second element over here swap
them meaning basically all the smallest
elements will come to the top of your a
and this way we are sorting our results
by relevance and the
that simple and then what we're doing is
we are just returning our finals and
that's how circuit function works very
simple if you think about it and it's
just taking a list and something you
want to compare and returning an array
of strings above what it thinks are good
results that's what I was looking for
results anyway continuing now let's just
present that function isn't there it's
already predefined
even though I created it it's just there
now now let's actually go into the main
part of the code the viewcontroller
class now we know that this is the class
you controller which inherits from
uiviewcontroller UI tableview datasource
UI tableview delegate and even UI text
field delegate this time so that we can
actually get input from the text field
and see when the user clicks on the
return function next as you can see we
have our IDL at for our table view we
have our variable named data which has
this list of movies we have our
temporary data which will hold our
search results for the UI tableview and
then our just our normal view did load
function in our text field should return
function first of all let me just tell
you we are just disappearing the
keyboard if it's on
if the keyboards up bring it back down
force it back down on our view meaning
self mute and editing true meaning force
then what we're doing is we're checking
if our text fields text is not equal to
nothing meaning there's something inside
of our text field and so this is going
to do now is if this is true we will set
our temporary data to the result of
searching using our normal data and
we're comparing the text fields text
then we're reloading our table view and
our table view will automatically adjust
saying okay there's something inside of
temporary data this means that instead
of loading normal data I'm going to load
temporary data and that's how it works
however else if
there is actually if there's nothing in
the text field we're just going to blank
out the temp data and reload our table
you and it's going to say oh there's
nothing in the temp data just use the
data that's it and then of course we
just returned true as we always do with
text field should return then in our
table view number of rows in section
function
we're just saying return and then an
inline conditional and we're checking if
the if temp - count is greater than zero
meaning there is something inside of
temp data then return temp data's count
or else return data's count meaning if
there is something in the search element
use the search result if there is not
then just use the normal data then in
our cell for our index path function all
we're really doing is we're taking our
cell as a uitableviewcell
setting it's text labels text to again
an inline conditional we're checking if
10 theta has something inside of it if
it does give 10 data's index path row
element or give data's index past row
element if this inline conditional is
false then we return the cell and then
we finish our table view once it is all
done and so next what we're doing
actually it's pretty much it so let's
just revise here because that was a bit
quick first of all we have a textview
the delegate is set to the view
controller when you enter something
inside of it we're using the circuit
functional which i've explained on the
whiteboard and over here we're using our
movies ArrayList we are comparing and
we're giving it to compare the stuff
that you entered in the text field then
we're reloading our table view meaning
put the search results then once you put
nothing into the text field it'll clear
out the search results then reload our
table view so that it refreshes the
normal data back in and that is the
simplest explanation possible
explanation possible for this search bar
and again I got this levenshtein
distance algorithm from rosetta code org
there will be a link to it in the
description below and that was actually
pretty much it
again this is an extension to Swift
databases part two and because there I
did explain a quick search bar but this
is just extending off of that and making
the search bar much better in the other
search bar I just compared here I'm
comparing and ranking and that's pretty
much it for this video again
if you like this video or it helped you
in any way please make sure to like this
video and again if you're new to my
channel or you like my cotton you want
to see more of it please make sure to
subscribe to the channel
it does help out and apart from that you
can even follow me on at tad Jimani on
twitter my twitter handle will be in the
description below if you can't spell
that you can leave comments down below
for any questions ideas tips app ideas
science math pretty much anything you
can even email me at admin at gmail.com
with any of these stuff i just said or
even a video question of you saying
something and I might just feature it in
my next video and that's going to be it
goodbye