Events Add an event Speakers Talks Collections
 
Duration 45:13
16+
Play
Video

PGConf India 2020 - Full Text Search in PG12 - Oleg Bartunov - Postgres Professional

Oleg Bartunov
CEO at Postgres Professional
  • Video
  • Table of contents
  • Video
PGConf India, 2020
February 27, 2020, Bengaluru, India
PGConf India, 2020
Request Q&A
PGConf India, 2020
From the conference
PGConf India, 2020
Request Q&A
Video
PGConf India 2020 - Full Text Search in PG12 - Oleg Bartunov - Postgres Professional
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Add to favorites
359
I like 0
I dislike 0
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
  • Description
  • Transcript
  • Discussion

About speaker

Oleg Bartunov
CEO at Postgres Professional

Design and development of IR systems, major developer of PostgreSQL, programming, quality assurance, it consulting and auditing, management of big IT-projects Specialties: database programming, information retrieval, postgresql, perl, mason, multi-tier architecture, search engines, open source, algorithms and abstract data type, soa, digital libraries, BIG DATA

View the profile

About the talk

Full-Text Search ( FTS ) is a search for the documents, which satisfy query and, optionally, return them in some order. The most usual case is to find documents containing all query terms and return them in order of their similarity to the query. Notions of query and similarity are very flexible and depends on application. Document can live on a hard drive or to be completely virtual as a result of complex query.

I will explain the importance of built-in FTS, the basics of FTS (configurations, query language and indexes) and present several new feautures of PG12, which improve FTS.

Share

I'm very happy to be here the first time in India and I really enjoyed staying here. People that's community in, India is growing and my intent to visit here in India, is to help engine community. And why I'll say, one of my lovely project I am and we I work was working on this project for 20 years, maybe. So, it's constantly improving improving. And today, I will talk about that idea in the what's new in the fidget 12. And what's expected in the PG-13,

So the entire half full Dixie lecture about 5 hours are so don't expect some big details but I'm here for several days. If you have any questions so I can talk to you, no problem. I am pulling this developer contributor sins, 1995. So and and in my main interest is expensive. And also I working for the community in which country is making lectures, organized confidence in someone or like three freaking making pictures and old. Good rock music. This is what I like, of course, I like my kids, my family and so long here is a picture

of the projects which I participated on, you know, the Jean room. And so, this is what I have made my contribution to hold you tonight at 6. Is there a local support? As you know, the police was only seven beats bass and I made, suppose he has available for other countries, body languages. And I still believe that this was my first small patch was the most important contribution to positive. Okay, let's talk. So what do you think what is the full text search? Is the process of finding the documents which which security,

but usually people see that it contains liquidy but actually created content can be very complex. They can use some pretty language so we will come clean. But the match find documents with the word for sorted by really was, but in progress, of course, you can reach me by e-mail today, like by change the time. Stand, for example, you search documents and just show the most prettiest commence, sodium in ranking, really want to document. So did you comment in database is different than we usually? Don't know what is a web page

you get from the Google from other search engines in database? You text can be wary of you. Got your little man. Is there a result of very big very complex? Select. Can you join with other tables? So it's never materialize. what you can use it for when you can treat it as a document and you can and you can create a document dog is enjoying, don't forget about calling to ask about because in full text search if you concur, Okay, doesn't matter. You can be adjusted. Open-source

complicated Twitty. You have free search like open source database. You have you require, is it one? Word not included in the commence one? What should be included. So it's like, it's a language. We need to pull in because there are so many external tools, like, Alex Torres. People say why we need this. Question is, of course, you know that comfortable. What's the problem that external search engines if they continue solder commence? As I sit looking as could be

completely virtual. Do Winn-Dixie sell bikes to notice. You need to feed list, and sometimes you can hold the external search engines. They don't have access to the previews. So, for example, you can buy at a beauty which stored in database which is not in the index by this but this engine I also external Mission accepted search engine. You have to maintain. If you have a electric storage or other, you have to have separate sewer system administration work since no one in production,

sometimes you craft a requirement that your software should be certified. So if you could have pulled his already certified for the use, for example, for government in Russia, with this requirement software control for his cousin, a good security and we improved it a bit but it's very difficult to another Woody Folsom, get to external search engines. Can't provide instant access into instant search. If you insult document, you can find it with external search engines with no time. You answer, then you need to wait until these documents

will be fitted to the exported to the search engine. For some for some people don't care. Don't care about the Sirius system which works with the money. You need to be very careful about consistency in public, will know that if you can find it. What's in the you know that when you search sound like Google, you see the documents you click. List of Whittier open cabins with the database is. So that's why when you used to go, we started to develop built-in text, search in Portuguese. And if you use external search engine

oil fuel system, it looks like this. So it will work. But looks to block to block external tools to your database. It said you have to, to do something, you know, something ugly tricks. If you care, if you can search engine, it is Nature's, natural. It's native to database. You can use any ideas for a search engine, you're free and the requirements for the Full integration with database engine. That means that transaction support concurrency recovery online in the sink, another important

requirements is configured. Use of can configure how to pass document, how to treat warts, how to normalize work, it's just text, don't confuse this text because with text, we can with this search engine, we can in the actual experience of experience in the sink chemical diagram for electrical formulas in. Just a matter of Pastor, how you the document and how will you end it? Searching, Jinkins and Copa support. Is there a good good? I like and do something. You can use it and people actually a few use them and let's examples how to search the X-ray

in the database. It work sexually, but there is no vision. You don't freeze up a plural and singular forms of the same. What will be different? And there is no stop work with this ain't no, it's worse and no ranking is old documents or equal to the committee. You can tell you don't understand how to rank Yes, so it is slow because not mostly support. So it should feel sequential scan in, 93, inflatable, wake me up at the end of support for Operator, Operator, if you install PGT are getting extension, actually

can use the least afraid of and self Kindle support, but looks like you have to write it but no, it's nice check yesterday there. It's easier. You can start investing. Nepali. So, let's see, explain what is the basics of the building blocks of text? Actually very simple. Today's a tight one is Taylor Swift dating site so you can do you get stripped of its country. Counties is a sorted array of Elections, which country is Mount constitution. In binary representation by another pipe is. It's a textual did the

bike for chlamydia with somebody and soul. If this is a traitor to and sign in, so I might show up later. So this works for example. All right, why can't why can't his committee and fixtures? What is it? Actually a fuel filter and just quit. It can be constructed or right by. You can use external to Fluor Field total lights on and actually, on Total Protection of text in total, and then it's up to the database. So we use database only fostering storing insertion, what position do commands and clearly was

wait outside. Yes. So these precincts Victor and what is a building blocks, so you can build different search engines. Is there a Safeway in Dixon images formula but provides the infrastructure for fixing. It provides the functions to make conversion from text document to the kids later. Pearl white Endocrinology for acceleration text search results in obtaining search slippers. So, so this is basically and oldest controlled, by which configuration will

commence Department, Camille's dictionaries? Which used to normalize work from this. And so, Parker brakes, the document onto and type of token and each token broken, go to to buy dictionaries. You can throw away. It's when you show this to princess Lexi and you can block dictionaries and then use it will interface and we have to face. So inside a school, you can create all through the group. Hold this picture changes. You can completely Weather in Justin. Just several flights leave. Yes, so how

to tell sister is a configuration name and this big zit look again. And then we process is the token and talk and type and reapply for each token with dictionaries. Once the dictionary to the what? It's just not into. This committee will just text, it contains some pretty good at language and stars and is a Play-Doh video. So stop on the right side. It won't do some idiot. Comes with super nova to Supernova and shortening S10 and starts, can you add it to the single form? And then you Butte

Lake, really looks like Supernova or as Sin and star and then it comes to the same So what's the difference to process the credit language? And the City of Bartow. It depends on your dictionary. Of course you can see, you know, if you can teach the text abbreviation, everything we kept going to three types of Tokyo's in for each. For each book in which religion, apply a different dictionaries. Example of consecration for Russian language. So, we see that for asking what we apply English, Taemin, official function, TSX eyes you apply

english team to the stars and it just make it from Pluto. Russians to me. Tokens process by simple dictionary which has just the lower case and support. So where is it for a used? Just one usually, but later I will show more complex example. This is so bogus documentation, for example, Is our special dictionary for paugus. For example, for PJ's Grill, Fiji will be so, will you several officials? Who do onions in the English dictionary? So it's come your thumb

today route. Malibu Cafe, English Standard English dictionary. It's so it's so good luck. You can use for this Country Creations for for inducing puzzle, specific emails. The program, just a program one. What could be when you when you lose? Yes, and look up special. TS Nick size function, which you can use to, to check the, Your Dictionary API allows you to develop your own. If you want to invest integers integers is intended. You just leave it on the several

ways to write white. White or you can write f f f. Another. So with this dictionary you can just say so when you you don't need to write several combinations, white or ffo in Seoul So it's up to you how to write addiction is keita's. Example example of a few in the historical documents. You know that we use Arabic this but in historical documents people use Roman and this just make it. So, when you search for example nineteenth-century, you don't need to write complete ambulation, Roman and Arabic

Deluxe examples. We using the Throne of me will use a PC Area library with. So you can write with this week. Several templates for dictionary Configure yourself but you don't need to program example, simple and it suits, the 4th. So you just put away cuz I'm going beyond the file with the stopwatch simple but it is it is it Bejeweled public school is a PDS School in 4th grade. Some people say Bulgaria and right, post, it also PJ's Grill. $2 to replace phrase. Bye bye. Bye. Another phrase

for the English, but the Project's Noble, which provides their facility to create singer for different languages. So now we use oldest languages supported by and supports. so this example how it works, I spill spill Huntsville, my spell, you know will this dictionary switch used for swelling in LibreOffice? Open Office on someone else available for Liberal use? Look up in the country. And sometimes it's very useful when you removed their special accent characters

just remove and then filter the token to the next picture late. We will use food stamps on Tuesday because each dictionary, it's a 20 GB memory Trail dictionary. It's just one copy and much faster, but then we'll be Instant access to the dictionary. And here are the numbers English. Indian English, just some less than one second and bigger. because when you selections, We did. Pics of installed as a shirt. He does example how we create the share English dictionary. I still file.

We have almost the same completely sold less than 100 seconds compared to the second. So used to share memory if you want, if you have carload real Skyward, Look up functions to support conversion of query and documents, for example, to Taylor Swift, or function on input Jason or Jason day. So, we function, automatically understand there is a good label default exercise complete. So, but I recommend. This is why the configuration because Jason is a image of a function if you don't and if you don't you supposed to find the configuration, the police will be will slow down your credit because

it's doesn't understand that it's commutable. Coherence examples. Olympics. And I also understand your language and other function for what people bother to write inside. So plain to Tesco flynntix without simple. Please do. Because if you're looking for the, it preserve the difference between the words in the order. So this is also supported and it's like a Google go to language. So if you like to write Google play today, Of course, we have several look up for indexes, like in dicks. It would have driven index and

become a room. Looks so did the difference is in the performance of search of the ranking of updating. So just because it's a good score for a bit and it's okay. Yeah. But it's not very scalable Moravian Tail Fairy Tail. So actually, it's a little awkward course. Nottoli for text to wish but for winning money, order to gation. And what is like a gym plus plus because it contains inside the exam additional information, you can specify which information you want to store in the next, and you can use this

information to Run results. So it's very fast if you for rank results really fast. For example, if you want to find look against the commands, so Who is George and ranking will be held in the in Dixon. You don't need to go to the Sheep. Pasture what's the price is at its biggest size. Of course, because to store some more information. We use it. Mostly for mostly. It is only for clothes. So, in summary, for flexible, very flexible, and they provide some collection of

weeks, you can combine to make a sweaty different. You can provide custom Sportster custom, dictionaries and what's really important. You can use it won't power up at 12. So you can combine full text search, which, for example, with special, with virtual with temporal data can be Example, working example, just use this oral. You are able to switch with increased alkire its use the custom poster. Which example, how would you feel parts of different from their

Captain Parker? Because we will use its relation to search, not just English to Russian. So, here is a search So I can reference so you can you can just to see how to make in one query in the Asian Federation people off of like you write something inside. Just so we have been looking for a support for some pretty language. No one important about Jason D, how to search Jason day because it's very important Community V8 engine. What is the Divi engine is The Benchmark for the database and

Mike Will Oracle Ms. SQL data, 1 months old. Any stop growing when we implemented Jason day because so many other people from law school, community and important for both us, and our support, for text to John, Jason be. As I said, and she was example. Look just recently but also Jason of course, Jason and Jason preserve spaces. So this way the result is different. So we searched inside value of specific key. So we don't care for extended Which official function

knows about. Jason B2K selector. What else do numeric and key. You can specify it Who is example of how you can induce everything. Bridget wheeler in support of Jason Park, which is the wedding potent part of the school's resource, for which you can easily described projection, of Json data, you want to use, Here's the example, example, MPG 11, you have to write this query. If it's real, it's really simple and easy to understand. Using Jason for you. Can you can You can use in Texas. For example, IMDb database.

To accompany its name. And this is a reaction erase and some structure you can construct you can in the different parts of this. IMDb Jason that's really sweet of you to 3 and you just in the jackpot. It's the fourth of July rated columns. Before we just, we love, we always said that, if you want to update FTS TSX and create, It's so much easier. So this example, which he actually is still at work, one will commit people that if we do you want, you kept 255ds querrey win a million times. And you

everywhere before it was really difficult it was slow sometimes very slow because What temperature's searching for him without any problem with performance? It's because you can now use immutable functions in from statement and then resort to this William examples of a blend of different planet, how it looks like. Another come eat. It's a void full, scan of Jewel In The Swim possible. He is example of how it looks like. If you read, scientific books on the last page, as you see the ended, so the race of fidget where this world capitals,

the problem is this, if you Look for sample, sample for the query. Like, Look for you can't use index, you have to scan the coolant leaks. And before before the committee likes, this example, what's the weather, what it's cool because What? No? After this, it has become more smart and don't use this as an example is. He took away two free and coffee seconds to milliseconds and Imperial sit in the same is March March faster because you don't need to read the cool in the waters, in the

read 180. Because you don't, you don't read that in the sky. So this is the flight is available. You can make picture which flights available and everything I use in my presentation. Yeah so I will another references, you know that it is what it is. It's so big country India because no supporting public sexually. This problem but I truly just want they work for him to support. If you somebody wants to drive to make the contribution come to me, I will show you how to make things

because candy is standard algorithm exist in Snowbowl Home, Depot, Home Depot. Canada is she was example yesterday. I was talking to some guys here and show them and their mobile phones in Polish. you see that I am the second flight, their difference is this true and false because of the plural and single Elkins system standard algorithm will remain. This is like a Nepali Improvement for the party because we sure we need to treat special. So we got another when we can make it 4:14

because there is no less noble than another state. It's so could be a good contribution for beginners developer. So this is a lot like, so old now, and thanks for the addition and I want to why people like friends, when you're both asked me to promote their political figures in Kathmandu University, so you can still. So you can see the whole, it's not far from injury yesterday. Yes, we have been submitted. Bissell. So if you could questions that I will be able to hear. You can knock me no problem. I have several books. You can take this.

It's a good for really familiar. If you want to start working with bogus, you can read this company to the Polish Community. It's written in English is a book. Specially for people who wants to stop and not afraid because sometimes it's very good for a beginners. You read and write and you get the positive results please. And just one week and you you can say that I know how to write some crazy ass in Portuguese. How to use the indexing how to use Jason VA in so long. So I feel like

I don't know, how are you? Just, I don't know. Look up this book available for free. You can just do Lord. Yes. Yeah, Usually, how you said I prefer them woman first? and then, thank you. Thanks for letting thank you for the talk.

Cackle comments for the website

Buy this talk

Access to the talk “PGConf India 2020 - Full Text Search in PG12 - Oleg Bartunov - Postgres Professional”
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free

Standart

Get access to all videos “PGConf India, 2020”
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Ticket

Interested in topic “IT & Technology”?

You might be interested in videos from this event

September 28, 2018
Moscow
16
179
app store, apps, development, google play, mobile, soft

Similar talks

Ivan Panchenko
Deputy CEO at Postgres Professional
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Apoorva Aggarwal
Database Engineer at Clarisights
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Venkat Susarla
Cloud Solution Architect, Cloud Professional Services at Societe Generale Global Solution Centre
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free

Buy this video

Video

Access to the talk “PGConf India 2020 - Full Text Search in PG12 - Oleg Bartunov - Postgres Professional”
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free

Conference Cast

With ConferenceCast.tv, you get access to our library of the world's best conference talks.

Conference Cast
816 conferences
32658 speakers
12329 hours of content