00;00;00;00 - 00;00;06;06
Speaker 2
Hey, Jim and I am recording this for podcast. Okay?
00;00;06;08 - 00;00;09;19
Speaker 3
Sure thing. What's the topic of your podcast?
00;00;09;22 - 00;00;52;09
Speaker 2
Okay, so I have this project that I'm learning. I'm learning how to use and, and, and, and a web browser together to create a, a web app, I guess you could say where you can go and you can type your city name in and what that web app does is it goes and finds the, the link to the public library request to book or submit, request to book for purchase or, that kind of thing, submit a book for purchase or, or any variation of that.
00;00;52;09 - 00;01;13;00
Speaker 2
Right. They say it in a couple different ways. But I want to build something that ingests the city name and then finds that library page and returns it.
00;01;13;03 - 00;01;21;24
Speaker 3
That's a cool idea. So to clarify, you're building a web app where someone enters a city name and the app finds the link to that city's public library page for requesting book purchases. Right?
00;01;21;26 - 00;01;26;21
Speaker 2
Yep.
00;01;26;23 - 00;01;32;28
Speaker 3
Love it. So what kind of tech are you thinking of using for this?
00;01;33;00 - 00;01;43;01
Speaker 2
Like a web front and that interacts with an innate and back end? Is the right terminology.
00;01;43;03 - 00;01;59;11
Speaker 3
You got it. So for the front end, you could use HTML, CSS, and JavaScript. Then for the back end with no end, you'll probably want to use its Http request node to grab data from library websites. How does that sound?
00;01;59;13 - 00;02;24;02
Speaker 2
Yeah, so that's what I actually wanted to discuss today. I've got a bit of it built. I've got, the webhook right. The like you, you you have the submit button or like a textbox with the submit button, you hit the submit, it goes to the webhook, and sees it, ingest it and can, and can work from there.
00;02;24;05 - 00;02;32;17
Speaker 3
It's awesome that you've already got the webhook an and part working. So what's the next step you're thinking of tackling?
00;02;32;20 - 00;02;59;07
Speaker 2
Okay, so one thing that I don't understand, is when I talk to ChatGPT on the web app, like the chat GPT open AI web app and I'm like, hey, can you find me a link to, submit book for purchase? Or what is the phrase that they usually say for the link?
00;02;59;09 - 00;03;10;24
Speaker 3
Libraries usually have a few different ways to say it. You could try suggest a purchase, recommend a title, or even just request a book. Do any of those sound like what you're looking for?
00;03;10;26 - 00;03;15;06
Speaker 2
Yes.
00;03;15;09 - 00;03;24;18
Speaker 3
So once you figure out the exact phrasing, you can use that in your innate and workflow to search for the link on the library's website.
00;03;24;20 - 00;03;41;06
Speaker 2
Okay. So that was actually the part of the workflow that I wanted to brainstorm with you today. Of it. But, to start out, I wanted to go over something really quick.
00;03;41;09 - 00;03;48;01
Speaker 3
So let's brainstorm. Do you want to figure out how to find the link on the website, or did you have something else in mind?
00;03;49;12 - 00;04;14;19
Speaker 2
Okay, so when I open the app on my phone for GPT five and I ask it something like, can you please send me a link to the purchase or suggest to purchase or suggest the book page of the San Mateo County Library, or of the Boston, Massachusetts Library, or the Boise, Idaho Library, or whatever. It does an amazing job.
00;04;14;19 - 00;04;39;18
Speaker 2
It sends me back the link, and it goes to the right place and it all works. But when I do the same exact thing in and a right, I have a, a text box that sends the city name into a prompt. And in that prompt in the AI, OpenAI GPT five node, it has everything that is supposed to be there.
00;04;39;18 - 00;05;17;09
Speaker 2
Like, please send me this city and find the purchase, a book page and blah blah, blah, blah, blah. And it's all set up, right? When that node sends me back a link, it goes to the right library. But it's almost always broken. So there's a disconnect somewhere in there. Like, why or why is the an GPT five node sending me broken links and the GPT five OpenAI app sending me great working links.
00;05;21;23 - 00;05;24;02
Speaker 2
Gemini, can you hear me?
00;05;25;07 - 00;05;30;28
Speaker 2
Gemini please respond if you can.
00;05;31;00 - 00;05;47;09
Speaker 3
But. So you want to brainstorm how to find the link on the website. One way to do that is by using web scraping, which is where you extract data from websites. You could use an eight ns Http request node to grab the website's HTML, and then use a tool like cheerio to parse the HTML and find the link you need.
00;05;47;16 - 00;05;51;29
Speaker 3
Have you worked with web scraping before?
00;05;52;01 - 00;05;58;14
Speaker 2
No. I need you to go way simpler.
00;05;58;16 - 00;06;13;20
Speaker 3
Okay. No problem. Let's try something simpler. You could use an eight and the web page node to grab the website's content, and then use the find text node to search for the phrase request a book that might give you the link you need. How does that sound?
00;06;13;22 - 00;06;31;21
Speaker 2
Okay, we're we're we're pretty granular here. I want to talk about ways that we can kind of do this. I don't want to I don't want to discuss specific nodes like necessarily. I want to think about it more high level.
00;06;31;24 - 00;06;43;23
Speaker 3
But thinking high level. So besides web scraping, you could also check if the library has an API. APIs are a more structured way for websites to share data. If they have one, it might be easier to get the means. Or you could try.
00;06;43;23 - 00;07;15;11
Speaker 2
Searching and I can use it for say what? I don't want to do that. I want to just be able to, I don't want to look for an API. I want this to be I want this to be a web app where you can put in a city name and just find this page super easy, reliably, without a bunch of fluff technology.
00;07;15;13 - 00;07;32;19
Speaker 3
Okay, I understand you want a straightforward way for your web app to find library pages without relying on APIs or complex methods? In that case, you could try using a search engine like Google or DuckDuckGo. You can construct a search query with the city name and keywords like request book purchase to see if the library page shows up in the results.
00;07;32;21 - 00;07;37;27
Speaker 3
Would you like to explore that approach?
00;07;37;29 - 00;08;06;08
Speaker 2
Okay, so what I would like to do actually, is I would like to discuss a couple of ideas that I have for you. Right. So so I can go and I can have a node that searches for purchase request library in this city. Right. And then it tries to find that page off of the search requests. And in that way Google does a lot of the heavy lifting on their side.
00;08;06;08 - 00;08;39;01
Speaker 2
Right. Chances are on the first page will be that purchase request thing or what I could do is I could just search for that library's website, which is even easier, right? And, but then I go into the site map of that library and scrape the different pages until I find it. Are you seeing how these are? These are doing the same thing, but they're slightly different processes.
00;08;39;04 - 00;08;44;26
Speaker 2
And I want your feedback on this stuff.
00;08;44;28 - 00;09;04;20
Speaker 3
Okay, I see what you're thinking. Using a search engine like Google is definitely a good way to leverage their search capabilities and find the library page you need. Searching within the library's website or sitemap could also work, but it might be more dependent on how the website is structured. Both methods have their pros and cons. Do you want to explore how to implement either of these methods?
00;09;04;22 - 00;09;30;26
Speaker 2
Okay. Let's start. Let's start by digging into the different ways that we can do this. Right. Method one is probably a decent option. We pick an ideal, search query for Google, and then what we use is we use AI to pick the right link out of the ten.
00;09;30;28 - 00;09;41;26
Speaker 3
Okay. Let's explore that first method. So you'd craft a specific search query for Google and then use AI to pick the right link from the results. Right. What kind of query are you thinking?
00;09;41;28 - 00;09;49;18
Speaker 2
I don't know if I wanted to find this page, you would know better than I would. Can you suggest some?
00;09;49;20 - 00;10;01;07
Speaker 3
Or you could try something like request book, purchase city name, public library or city name library. Recommend a title. Do you want to try a few variations to see which works best?
00;10;01;09 - 00;10;11;16
Speaker 2
Which is okay? I want you to save them. I want you to list them out individually, slowly.
00;10;11;18 - 00;10;25;28
Speaker 3
Okay. Here are a few options. Request. Book. Purchase. City name. Public library. City name. Library. Recommended. Title. City name. Public library. Suggest a purchase or city name. Library. Request a book. Do any of these sounds like a good starting point?
00;10;26;02 - 00;10;31;13
Speaker 2
Does the order matter in this kind of stuff?
00;10;31;15 - 00;10;40;04
Speaker 3
Search queries. Word order can sometimes matter, but search engines are pretty good at figuring out what you mean, so it's worth trying a few different orders to see what gives the best results.
00;10;41;03 - 00;11;06;14
Speaker 2
Okay, so this, I grew up in a town called Pescadero on the West coast. And we didn't have a Pescadero library, so this wouldn't work in this situation. We had, we had the having Bay library with the heron Bay library was part of the San Mateo County Library system.
00;11;06;17 - 00;11;17;23
Speaker 3
Okay, I understand in that case, you could use San Mateo County Library. Request to book purchase as your search query. Would you like to explore how to implement that search in your app?
00;11;17;25 - 00;11;30;04
Speaker 2
Know what I want I'm talking about here is if we have a single search query, even if it's linked to the city name, that that's not always the right call.
00;11;30;07 - 00;11;48;05
Speaker 3
Right. A single search query might not always work. In that case, you could try using a few different queries and see which one gives the best results. You could also try searching for the library's website directly, and I'm looking for a request to book page.
00;11;48;07 - 00;12;05;26
Speaker 2
So, I don't know if you ever answered, but why, why would I have this? I in a in node look for links. The links that it gives back to me are bad.
00;12;05;29 - 00;12;13;19
Speaker 3
You could have the I node look for links that contain the words request, purchase or recommend to try and find the right page. Does that make sense?
00;12;13;20 - 00;12;34;17
Speaker 2
You know what? I'm or you're not answering the question what? When I use the web app, it works on OpenAI. But when I'm when I'm in any and those we those links are like old or something. It's not getting good links. Why is that?
00;12;34;19 - 00;12;48;29
Speaker 3
Links are outdated. It might be because the library's website has changed. You could try updating your search query with more recent keywords, or checking the library's website directly for the most current links. Does that help?
00;12;49;02 - 00;13;11;18
Speaker 2
Does any and not, have up to date data like when when I have any in installed on a vpn's? Does it not have access to the internet like the OpenAI web app does?
00;13;11;21 - 00;13;27;28
Speaker 3
It and itself doesn't have its own data. It relies on the information it gets from the websites you're scraping. So if the links on those websites are outdated now and will pick that up, the fact that you're running it on a VPN shouldn't affect its access to the internet. Though.
00;13;28;01 - 00;13;49;20
Speaker 2
Okay, so we still have not figured out why, the n a n I keep coming back bad links, but open. I in its own app gives me good links. Is the app like checking the internet for something?
00;13;49;23 - 00;14;05;26
Speaker 3
To use a n, it's possible that the links that retrieves are outdated because the website structure has changed since those links were indexed by search engines, OpenAI's web app might have access to more up to date information, or it could be using a different method to find the links.
00;14;05;28 - 00;14;20;16
Speaker 2
It must be using a different method or something like that. But I'm using GPT T5 on a node and I'm using GPT five on an app.
00;14;20;18 - 00;14;36;02
Speaker 3
So it's possible that the AI model in your node is configured differently than the one used in the web app. You might want to check the settings of your N8 and node to see if there are any options to adjust the model or its parameters.
00;14;36;05 - 00;14;55;09
Speaker 2
Okay. I don't know really if we're getting any further in this part of the conversation. So let's go back to, using, brainstorming methods for getting good links in any and.
00;14;55;11 - 00;15;09;19
Speaker 3
Okay. Let's explore some other methods. You could try using any web page node to scrape the library's website directly. Or if the library has a sitemap, you could use that to find the page you need.
00;15;09;21 - 00;15;13;28
Speaker 2
Okay, describe this method.
00;15;14;00 - 00;15;24;28
Speaker 3
Or you could use the web page node to grab the HTML of the library's website, and then use CSS selectors to find the link you need. Have you worked with CSS selectors before?
00;15;30;04 - 00;15;31;15
Speaker 2
No, I have not.
00;15;31;17 - 00;15;52;05
Speaker 3
Okay. No problem. CSS selectors are just patterns that let you pick a okay, no problem. CSS selectors are just patterns that let you pick out specific parts of a website's code. It might be a little tricky if you haven't used them before, but there are lots of tutorials online that can help.
00;15;52;08 - 00;16;18;11
Speaker 2
Okay, so we've got the method of finding the right, library's web page and scraping the page. We have the method of just searching on Google and having I pick the right link by like its best judgment. And then let's come up with a third, totally different method.
00;16;18;14 - 00;16;32;08
Speaker 3
Okay. For a third method, you could try using a library, database or directory. Many library systems have online databases that list all their branches and services. You might be able to find a database like that and extract the link you need from there.
00;16;32;10 - 00;16;42;14
Speaker 2
Now I want it to be something that is, like you can do it with any library.
00;16;42;17 - 00;16;53;12
Speaker 3
Okay, then how about this? You could try searching for the library's contact information and then calling or emailing them to ask for the link directly. It's a bit old school, but it would definitely work for any library.
00;16;53;15 - 00;17;30;20
Speaker 2
Now, I would like to stay within kind of the type workflow of, finding the link, right? So I don't want to or I don't want to deal with email or calling or whatever. This other stuff is all I want. My main goal is to go the shortest distance between someone putting in their city name, and then return the system, returning back the request to book purchase link.
00;17;30;22 - 00;17;46;05
Speaker 3
Okay, I understand you want a streamlined method to get that link. In that case, you could try using a specialized search engine that focuses on library resources. There are search engines specifically designed to find information within library websites and databases. Would you like me to find some example?
00;17;46;07 - 00;17;50;03
Speaker 2
Yes, please.
00;17;50;05 - 00;18;03;29
Speaker 3
Here are a few options. One is called OCLC WorldCat. It's a huge global catalog of library collections. Another option is Library of Congress online catalog, which is great for U.S. libraries. Would you like me to find more options?
00;18;04;02 - 00;18;08;05
Speaker 2
Yes.
00;18;08;07 - 00;18;23;02
Speaker 3
But also try using Z Tok, which is a search engine that focuses on journal literature and conference proceedings. Or there's Capac, which is a union catalog for UK and Irish libraries. Would you like me to find even more options?
00;18;23;04 - 00;18;31;16
Speaker 2
No. I'd like you to tell me more about this OCLC WorldCat.
00;18;31;19 - 00;18;49;03
Speaker 3
CLC WorldCat is basically a massive online catalog that lets you search the collections of libraries all over the world. It's a great way to find books, movies, music, and more. You can even see which libraries near you have, the item you're looking for.
00;18;49;06 - 00;19;01;24
Speaker 2
Okay. Will this be able. Will I be able to access, purchase or recommend for purchase links via this?
00;19;01;26 - 00;19;22;14
Speaker 3
WorldCat is a great resource for finding library materials. It doesn't have a built in feature to directly request or recommend purchases. It's more of a catalog to see what's available. You might be able to find a link to the library's website within WorldCat, and then use your other methods to find the purchase request page.
00;19;22;17 - 00;19;31;29
Speaker 2
So what is this? This world? What is it? WorldCat.
00;19;32;02 - 00;19;39;21
Speaker 3
WorldCat is the name. It's a huge online catalog that lets you search the collections of libraries worldwide.
00;19;39;24 - 00;19;46;17
Speaker 2
Okay, tell me about the Library of Congress. One.
00;19;46;19 - 00;19;56;17
Speaker 3
The Library of Congress is the largest library in the world. It's located in Washington, D.C. and was established way back in 1800. It has a massive collection of.
00;19;56;20 - 00;20;07;24
Speaker 2
Recordings from Gemini. Use. The second one you said in that list was like Library of Congress Online. Something, something.
00;20;07;27 - 00;20;17;14
Speaker 3
Right. It's the Library of Congress online catalog. It's a great tool for searching the Library of Congress vast collection, but it doesn't have a direct way to request purchases.
00;20;21;01 - 00;20;38;10
Speaker 2
Gemini, you were just talking about, search engine specific to libraries and stuff like that, and you mentioned that was like the Library of Congress online and in you were talking about how like, maybe it was a list of well.
00;20;38;10 - 00;20;38;21
Speaker 3
The Library.
00;20;38;21 - 00;20;39;08
Speaker 2
Libraries.
00;20;39;14 - 00;20;53;20
Speaker 3
Catalog is a great resource, but it's really focused on the Library of Congress is own collection. It might not be the best for finding resources in your local public library. We can explore other options, though. What kind of library are you trying to find?
00;20;53;23 - 00;21;10;27
Speaker 2
Is there anywhere that someone has compiled a list of, like public libraries across the US and like, is linked out to them?
00;21;11;00 - 00;21;26;01
Speaker 3
Yeah. The Institute of Museum and Library Services has a directory of public libraries in the US. It's called the Public Library Survey. And you can find it on their website.
00;21;26;04 - 00;21;35;19
Speaker 2
Okay. That's pretty cool. Tell me more about this institute.
00;21;35;22 - 00;21;53;08
Speaker 3
Institute of Museum and Library Services is a US government agency that helps support libraries and museums. They offer grants, conduct research, and work on developing policies to help these institutions better serve their communities. Do you want to learn more about the grants they offer or their research?
00;21;53;10 - 00;22;01;23
Speaker 2
I want to know if they have a database of all the public libraries in the United States.
00;22;01;26 - 00;22;12;11
Speaker 3
The Institute of Museum and Library Services has a database called the Public Library Survey. It has information on over 9000 public libraries in the US. Would you like to know more about this database?
00;22;12;13 - 00;22;16;27
Speaker 2
Yes, please.
00;22;17;00 - 00;22;33;20
Speaker 3
A public library survey is a great resource. It has information like library locations, contact details, and even statistics on things like visits and circulation. It's a really comprehensive look at public libraries across the country.
00;22;33;22 - 00;22;50;06
Speaker 2
And what, how can view this public library survey? How recent is it? Is it a database? Does it have links? All these questions?
00;22;50;09 - 00;23;08;25
Speaker 3
Public library survey is an annual survey, so the data is pretty up to date. You can access it on the Imls website, and it's available as a downloadable data set. It doesn't have direct links to library websites, but it does have contact information, so you could potentially use that to find the links you need.
00;23;08;27 - 00;23;15;00
Speaker 2
Okay. And is this just like an Excel file?
00;23;15;02 - 00;23;25;03
Speaker 3
Public library survey data is available in a few different formats, including CSV, which you can open with Excel. It's a pretty big data set, so it might take a little while to download.
00;23;25;05 - 00;23;32;22
Speaker 2
And you said this is in the I am less website.
00;23;32;25 - 00;23;51;04
Speaker 3
That's correct. The public library survey is on the Institute of Museum and Library Services website. You can find it by searching for public library survey on their site, or by looking in their Data and Statistics section.
00;23;51;07 - 00;24;04;28
Speaker 2
Okay. Tell me about this CSV if I were to download it. What, how many how many columns do you think it has with headers?
00;24;05;01 - 00;24;23;07
Speaker 3
CSV file from the Public Library Survey has a lot of columns. It includes things like the library's name, address, phone number, website, and even information about their collections and services. There are over 100 columns in total. So it's a pretty comprehensive data set. Do you want to know about specific columns?
00;24;23;09 - 00;24;26;26
Speaker 2
Yes, please.
00;24;26;29 - 00;24;44;29
Speaker 3
Okay. Some of the key columns include the library name, the physical address, the mailing address, the phone number, and the website URL. There are also columns with statistics about the library's finances, staff, and collections. Let me know if you want to know more about any of the other columns.
00;24;45;02 - 00;24;49;24
Speaker 2
Do you know all of the columns?
00;24;49;27 - 00;25;07;24
Speaker 3
Well, there are a lot of columns, but some of the most important ones are the columns with the library name, address, phone number, and website URL. There are also columns with information about the library budget, staff, and collections. Would you like me to find the columns with information about the library services?
00;25;07;26 - 00;25;11;16
Speaker 2
Sure.
00;25;11;19 - 00;25;27;24
Speaker 3
Okay, there are columns that show the number of books, ebooks, and audio materials available. There are also columns about the library's programs for children and adults, and whether they offer online resources or meeting rooms. Is there anything else you'd like to know about the columns?
00;25;27;27 - 00;25;47;00
Speaker 2
Yeah, let's say that on this thing that I was making, I wanted to let people know how many, how many books were ordered every single year.
00;25;47;02 - 00;26;02;16
Speaker 3
You'll want to look at the columns labeled Total Circulation that shows how many items were borrowed from the library each year. There are also separate columns for children's materials and electronic materials. If you want to get more specific.
00;26;02;18 - 00;26;31;00
Speaker 2
What, what columns are there that pertain to the number of books and the speed of turnover? And, it would give people like a little bit of a view inside of what their request could mean. Maybe number of requests received.
00;26;31;02 - 00;26;57;12
Speaker 3
Public library survey CSV has columns like Total Circulation, which indicates the number of items borrowed, giving you a sense of turnover for the number of books. You look for columns related to print materials, ebooks, or audio materials within the collection section. Unfortunately, there isn't a direct column for requests received in the public documentation, but the circulation data can give you an idea of demand.
00;26;57;14 - 00;27;08;15
Speaker 2
And if it's updated annually, can you, upload all this data and be able to see trends in it?
00;27;08;17 - 00;27;23;09
Speaker 3
Since the public library survey is annual. You can definitely download the data for multiple years and analyze trends. You could look at things like how circulation changes over time or how library funding has shifted. What kind of trends are you interested in exploring?
00;27;23;11 - 00;27;45;27
Speaker 2
Cool. So I could make something that was like, the that picked like a certain library over like, ten year period and could show, I don't know, like how their circulation has changed.
00;27;45;29 - 00;28;03;19
Speaker 3
Exactly. You could analyze how a specific library circulation has changed over the past decade. It might be interesting to see if there are any patterns or if certain events, like the pandemic had a noticeable impact.
00;28;03;22 - 00;28;39;23
Speaker 2
And if I took the like, if I took the print materials and like, let's say the print materials was like, you know, 10,000 or many books or in the library. And then the next year it was, 9500, and then the following year it was like 9200. You'd be able to see how many books were like, at least added, minus subtracted or whatever.
00;28;39;25 - 00;28;40;13
Speaker 2
The change.
00;28;40;13 - 00;28;59;26
Speaker 3
And by comparing the print materials column across different years. You're looking at the total circulation column. You can see the total number of items borrowed each year. To figure out the change. You'd subtract the circulation number from one year from the number in the previous year. That'll give you the difference.
00;28;59;29 - 00;29;30;09
Speaker 2
Okay, so, what columns from this MLS Public Library Survey, databases do you think are helpful in what we're trying to do here? We're right. We're inputting a city name and we're getting back a good, reliable link. Right. Obviously the website URL is going to be a big one. Anything else?
00;29;30;11 - 00;29;43;06
Speaker 3
Besides the website URL, the physical address and mailing address columns can be super helpful. You could use those to double check that you've got the right library in case there are multiple branches in the same city.
00;29;43;08 - 00;29;47;12
Speaker 2
Okay, that's a good idea.
00;29;47;15 - 00;29;57;08
Speaker 3
Great. So to recap, we've got the website URL, physical address, and mailing address from the Imls Public Library survey. Do you want to talk about how to use these in your code?
00;29;57;11 - 00;30;02;09
Speaker 2
Yes, please.
00;30;02;11 - 00;30;21;22
Speaker 3
Okay. So first you'll need to download the Public Library Survey CSV file from the Imls website. Then you can use a library like Pandas and Python to read the CSV and access the columns we discussed. You'll probably want to filter the data by city name to find the specific library you're looking for.
00;30;21;24 - 00;30;32;05
Speaker 2
Okay, I probably will just get ChatGPT here to look at the CSV and find the city name and in the data, right? So you can.
00;30;32;05 - 00;30;44;26
Speaker 3
Use Python with pandas to read the CSV file. Filter by city name and then grab the website URL. You'll probably want to double check that the address matches, just in case there are multiple libraries in the same city.
00;30;44;28 - 00;30;55;18
Speaker 2
Okay, why would I use Python with pandas instead of like, just an AI node?
00;30;55;20 - 00;31;12;01
Speaker 3
Definitely use an AI node for that. But Python with pandas gives you a bit more control over the data. You can easily filter, sort, and manipulate the data to find exactly what you need. It's a powerful way to work with large data sets like the Public Library Survey.
00;31;12;03 - 00;31;21;00
Speaker 2
Yeah, but I don't really need that. I'd rather have something that's. Is it faster?
00;31;21;02 - 00;31;33;14
Speaker 3
Hey, if speed is a priority, you could try using polars instead of pandas. It's written in rust and tends to be much faster, especially for larger datasets. The syntax is pretty similar to pandas, so the switch shouldn't be too difficult.
00;31;33;17 - 00;31;56;04
Speaker 2
And, I don't want to do all this. I want to keep it really simple. I'd like to just be able to ask, chat GPT or Claude or whatever. Can you go into the CSP and get the website URL of this specific library in this specific city?
00;31;56;07 - 00;32;13;03
Speaker 3
Hey, if you want something simple, you could try using a library like requests in Python to send a query to the Imls website search API. You just need to format your request with the library name and city, and then parse the Json response to get the website URL. Want you to help you write out an example request.
00;32;13;06 - 00;32;19;17
Speaker 2
Just what we're doing this in in a in.
00;32;19;19 - 00;32;33;12
Speaker 3
Here's an example of how you could use the requests library to get the website URL for a specific library. You'll need to replace library underscore name and city underscore name with the actual names you're searching for. You can also use the library's physical address to filter your.
00;32;33;12 - 00;32;33;21
Speaker 2
Search.
00;32;33;26 - 00;32;36;18
Speaker 3
Which might help if there are multiple libraries that they need.
00;32;36;21 - 00;32;50;11
Speaker 2
I please stop talking. We're getting way off track here. I want you want you to stop and take a breath. And I'm going to reset us.
00;32;50;13 - 00;32;56;09
Speaker 3
But we can take a break and come back to this later.
00;32;56;11 - 00;33;38;00
Speaker 2
Okay, so we're in. Net and on a virtual private server. And what I'm doing now is I would like to have a text box that, and maybe an AI node, like a an I know it's Claude or an AI node with GPT five. And what it does is it in it takes a CSV. And when I put a city name in it, it finds the library that I want and gives me back the URL from that.
00;33;38;02 - 00;33;45;09
Speaker 2
From that census or whatever that CSV.
00;33;45;12 - 00;33;57;03
Speaker 3
Okay. You want to build a chat bot that uses an AI node like Claude or GPT four to take a city name as input and return the website URL of a library in that city from the CSV file. Is that right?
00;33;57;06 - 00;34;04;26
Speaker 2
Yes.
00;34;04;29 - 00;34;24;21
Speaker 3
Okay, I understand in any den you would typically use a read CSV node to load the public library's survey data. Then you'd use a filter node to find the row that matches the city name provided by the user. Finally, you'd extract the website URL from that filtered row, while an AI node could be used to interpret the user's request and even help construct the filter.
00;34;24;26 - 00;34;30;12
Speaker 3
The actual data processing would be done by these other end to end nodes.
00;34;32;04 - 00;35;05;25
Speaker 2
Okay, Gemini. But, I still have the problem of, for instance. I type in Las Vegas, right? I mean, the type in Las Vegas for the V las space. Vegas. But my finger slips and I type in Las Vegas. Las Vegas. If I do the read CSV filter and extract within, then the read CSV is going to kind of error up, because there's not going to be a city called Las Vegas.
00;35;05;25 - 00;35;25;12
Speaker 2
Right? And and I will know that it's it doesn't mean Las Vegas. It means for Vegas. Like obviously there's no Las Vegas, Nevada. But if I do the CSV filter extract method, that's not going to work unless I squeeze an AI in there somewhere.
00;35;25;14 - 00;35;44;10
Speaker 3
You're right. If the user misspelled Las Vegas, the read CSV and filter nodes won't find a match an AI node like the one you mentioned. It could be really helpful here. It could correct the spelling or even suggest the correct city name to the user.
00;35;44;13 - 00;35;49;06
Speaker 2
Where would I place it in the process?
00;35;49;08 - 00;36;04;16
Speaker 3
Put the AI node right after the user typed in the city name. It would clean up the input and make sure it matches the city names in your CSV file. After that, you'd use the read CSV and filter nodes to find the library and extract the URL.
00;36;04;19 - 00;36;27;29
Speaker 2
Okay. But let's say that we have a situation where I put in pescadero.com, right? This is too small. The closest public library is the next town, which is Half Moon Bay. If I type in Pescadero, it's going to look for a Pescadero Library and then send back nothing. How do I overcome that?
00;36;28;01 - 00;36;46;12
Speaker 3
Yeah. If you type in Pasadena and the closest library is actually in Deer Park. You might run into some trouble. In that case, the AI node could be a lifesaver. It could use some fuzzy matching to find the closest match in your CSV file. Even if the city names aren't an exact match.
00;36;46;15 - 00;36;58;07
Speaker 2
Okay. So how would that? I know that Half Moon Bay is close enough to Pescadero to me to do something with the G.P.S..
00;36;58;09 - 00;37;10;21
Speaker 3
AI node could use fuzzy matching algorithms to compare Pasadena with the city names in your CSV. It would look for the closest match, even if the spelling isn't exact. Would you like me to find some examples of fuzzy matching algorithms?
00;37;10;23 - 00;37;29;10
Speaker 2
No. So the situation here is it's a geographical situation, right? I have an app, or I have a city that doesn't have a library, and you have to go to the next city over. How does the. I know to do that?
00;37;29;12 - 00;37;40;26
Speaker 3
AI node could handle that too. You could use a function that calculates the distance between Pasadena and other cities in your data set, and then return the closest library. Would you like to explore some distance calculation method?
00;37;40;29 - 00;37;44;28
Speaker 2
Yes, please.
00;37;45;00 - 00;38;07;09
Speaker 3
Distance calculations. You could use the never seen formula, which calculates the great circle distance between two points on a sphere. Given their longitude and latitude. Or if you want something simpler, you could use the Euclidean distance formula, which calculates the straight line distance between two points in a plane.
00;38;07;11 - 00;38;16;14
Speaker 2
I think that this, I think that this method is proving to be kind of difficult.
00;38;16;16 - 00;38;35;16
Speaker 3
Right. The having seen formula is perfect for calculating distances between cities. You'll just need the latitude and longitude coordinates for both Pasadena and the other cities in your data set. Once you have those, you can use the formula to find the closest library.
00;38;35;19 - 00;38;44;10
Speaker 2
Okay. I think, I don't know, this one. I like the idea of having a database.
00;38;44;10 - 00;38;58;13
Speaker 3
So we've got the AI node figuring out the closest point, and we're using the harbor scene formula to handle the next town over situations. We're getting close to having a complete workflow. Are you ready to talk about how to connect everything in in a.
00;38;58;16 - 00;39;04;13
Speaker 2
Yeah. Go ahead and do that for this method.
00;39;04;16 - 00;39;25;27
Speaker 3
Okay. Here's how you could structure it. First you'd have your AI node. It would take the city name the user enters. And if it's misspelled or not an exact match, it would use fuzzy matching to find the closest city in your CSV. Then you'd use the everything formula to calculate the distance to nearby libraries and return the closest ones URL.
00;39;25;29 - 00;39;38;02
Speaker 2
Okay. Can you describe the have seen formula and how that fits in here in in a more basic way?
00;39;38;04 - 00;39;59;22
Speaker 3
For thing. The have seen formula is basically a way to calculate the distance between two points on a sphere like the Earth. In our case, we'd use it to figure out the distance between the city the user enters and all the libraries in our data set. The formula takes into account the Earth's curvature, so it's more accurate than just drawing a straight line on a map.
00;39;59;25 - 00;40;06;13
Speaker 2
Yeah. What is is I going to be able to do all this stuff?
00;40;06;15 - 00;40;23;18
Speaker 3
The AI node can handle all of this. It can take the city name the user enters, use fuzzy matching to correct any misspellings, and then use the formula to find the closest library. It can also extract the website URL from the CSV data.
00;40;23;20 - 00;40;39;03
Speaker 2
Okay, so tell me specifically what I would need to do to do this tab or seeing formula thing, do I? I take the city right. And then what happens?
00;40;39;05 - 00;41;05;12
Speaker 3
Is the formula. First you'd get the latitude and longitude of the city the user entered. Then for each library in your CSV, you'd also get its latitude and longitude. You'd plug these coordinates into the received formula to calculate the distance between the user city and every library. Finally, you'd find the library with the shortest distance, and that would be your closest match.
00;41;05;15 - 00;41;08;08
Speaker 2
This.
00;41;08;10 - 00;41;28;18
Speaker 2
Okay, that makes sense. In the CSV. So what I would need to start with doing is giving the latitude and longitude points of every row in the CSV.
00;41;28;21 - 00;41;47;27
Speaker 3
Public library survey. CSV files should have columns for latitude and longitude for each library. You'll want to look for columns with names like latitude, longitude, lat, or lon. If those aren't directly available, you could use the address information in the CSV with a geocoding service to get the coordinates.
00;41;48;00 - 00;41;56;17
Speaker 2
There's the Public library survey database, have latitude and longitude columns.
00;41;56;19 - 00;42;14;04
Speaker 3
Oh, the public library survey data set doesn't include latitude and longitude columns. However, it does have address information which you can use with a geocoding service to get the coordinates you need.
00;42;14;07 - 00;42;38;12
Speaker 2
Okay, this is getting a little bit crazy, but let's let's let's let's, let's, go through it. What is what is it? Geocoding service. Does it cost money? Can I access it through the an API?
00;42;38;15 - 00;42;55;26
Speaker 3
So geocoding is basically turning addresses into coordinates that computers can understand. Think of it like that. You give the service an address like 123 Main Street, Anytown, and it gives you back the latitude and longitude for that spot. There are a few different services you could use, and many have free options depending on how many addresses you need to convert.
00;42;55;28 - 00;43;02;11
Speaker 3
Would you like me to find some example?
00;43;02;13 - 00;43;13;20
Speaker 2
Would I do this through an API or something like that, where I put the address in and I get latitude and longitude back?
00;43;13;22 - 00;43;23;06
Speaker 3
Many geocoding services offer APIs that allow you to send an address and receive the latitude and longitude coordinates in return. Would you like me to find some examples of geocoding APIs?
00;43;23;08 - 00;43;29;23
Speaker 2
Yes, please.
00;43;29;25 - 00;43;39;05
Speaker 3
Absolutely. Here are a few popular geocoding APIs you could use. Google Maps Platform geocoding API is a solid choice, and it's free for up to.
00;43;43;07 - 00;43;53;15
Speaker 2
Gemini. Can you give me, your. We cut off, but you're giving me a geocoding service API list.
00;43;53;17 - 00;44;12;10
Speaker 3
You'll need to sign up for an API key with your chosen geocoding service. Then you'll use a node in Nate in, like the Http request node to send the address from your CSV file to the API. The API will return the latitude and longitude, which you can then save back into your end to end workflow.
00;44;12;12 - 00;44;23;28
Speaker 2
No, you're going to give me a list of actual geocoding service APIs. There was like Google Maps and some others.
00;44;24;00 - 00;44;46;11
Speaker 3
Mentioned a few earlier, but here they are again. The Google Maps Platform Geocoding API is a popular choice, and it's free for a certain number of requests. Location IQ geocoding API is another good option, especially if you're working with OpenStreetMap data and Mapbox geocoding API is known for being developer friendly. Do any of these sound like a good fit for your project?
00;44;46;13 - 00;44;55;13
Speaker 2
Okay, probably just a simple, Google Maps one.
00;44;55;15 - 00;45;11;21
Speaker 3
Okay. If you're going with Google Maps, you'll use their geocoding API. You'll send an address to the API and it will return the latitude and longitude coordinates. You'll need to get an API key from Google to use this service.
00;45;11;23 - 00;45;48;04
Speaker 2
Okay, so, although this is a little bit more complicated, it feels a little bit more kind of foolproof, right? We have a complete list of the public libraries with the actual URLs. And maybe what we can do is we can, write somebody can put in a random ass town name, and it will give us the closest library to them, based on latitude and longitude.
00;45;48;04 - 00;45;57;03
Speaker 2
With that process, take a lot of compute, searching the CSV for the closest library.
00;45;57;05 - 00;46;16;10
Speaker 3
Greece. Using latitude and longitude is definitely more precise. It's a bit more work up front, but it pays off with accuracy, especially for those next town over situations. We'll get those coordinates using the Google Maps Geocoding API, and then use the header scene formula to find the closest library. It's a solid plan.
00;46;16;12 - 00;46;23;08
Speaker 2
Right, but using the harrassing formula on a CSV file with 9000.
00;46;23;08 - 00;46;24;03
Speaker 1
Rose.
00;46;24;04 - 00;46;31;00
Speaker 2
I feel like that's going to be work intensive.
00;46;31;02 - 00;46;45;23
Speaker 3
Right. Running the ever seen formula on a 9000 row CSV would be pretty resource intensive. To make things more efficient. We could use a spatial indexing technique. This would help narrow down the search area and reduce the number of calculations needed.
00;46;45;25 - 00;46;50;12
Speaker 2
Okay, can you explain that?
00;46;50;15 - 00;47;07;29
Speaker 3
Right. So spatial indexing is like creating a special map of your data. It organizes the latitude and longitude points in a way that makes it super fast to find the closest ones. Instead of checking every single library in your CSV, it lets you zoom in on the area you need. There are a few different ways to do this, but one common method is called a KD tree.
00;47;08;05 - 00;47;14;06
Speaker 3
It's like a binary search tree, but for points in space.
00;47;14;08 - 00;47;17;09
Speaker 2
Okay, I think we're kind of getting.
00;47;17;09 - 00;47;21;01
Speaker 1
A little bit crazy here.
00;47;21;04 - 00;47;24;29
Speaker 2
Let's let's go back to our method trace.
00;47;24;29 - 00;47;25;08
Speaker 1
Right.
00;47;25;08 - 00;47;29;14
Speaker 2
So this is method three. It's kind of really intricate.
00;47;29;14 - 00;47;31;02
Speaker 1
It's very accurate. But it's a.
00;47;31;04 - 00;47;35;21
Speaker 2
It's got a lot going on. I think our method one okay.
00;47;35;22 - 00;47;50;21
Speaker 3
Let's simplify things. We'll stick with the have received formula for now as it's the most accurate for calculating distances between latitude and longitude points. We can explore spatial indexing later if needed, but for now let's focus on getting the core functionality working.
00;47;50;23 - 00;47;59;07
Speaker 2
Actually, I want to brainstorm, a new, less complicated method. Right. Let's kind of lean on.
00;47;59;07 - 00;48;00;11
Speaker 1
Google search a little.
00;48;00;11 - 00;48;18;20
Speaker 2
Bit. Maybe instead of like, downloading this giant database and, like, checking the database against formulas, blah, blah, blah. Why don't we, like Google's already done all this stuff. Like, why don't we intelligently.
00;48;18;27 - 00;48;20;07
Speaker 1
Search.
00;48;20;10 - 00;48;27;21
Speaker 2
In Google and then vet a handful of links?
00;48;27;24 - 00;48;45;21
Speaker 3
That's a smart idea. We could use Google Search to find the closest library. You'd enter the user search term like library near me, and then parse the search results to extract the library names in URLs. This would eliminate the need for the CSV file and the seen formula.
00;48;45;23 - 00;48;47;12
Speaker 2
Okay. I like the idea.
00;48;47;12 - 00;49;02;02
Speaker 1
Of like a library near Pescadero, California, right? It's probably going to pop up at the half movie library.
00;49;02;04 - 00;49;20;02
Speaker 3
It looks like the Half Moon Bay library, which is part of the San Mateo County Libraries, serves the Pescadero area. You can find more information at link to SMC. Claud. There's also a bookmobile stop at 202 Stage Road, Pescadero, California. 94060.
00;49;20;05 - 00;49;21;08
Speaker 2
Okay, I know exactly.
00;49;21;08 - 00;49;23;16
Speaker 1
Where that is. That's really cool.
00;49;23;17 - 00;49;28;16
Speaker 2
I think this is going to be the way to do this. How do we do that?
00;49;28;19 - 00;49;31;24
Speaker 1
Links that we're finding. Right.
00;49;31;28 - 00;49;36;03
Speaker 2
So let's say that we find the library near blank. And so we have.
00;49;36;03 - 00;49;36;28
Speaker 1
The right library.
00;49;36;28 - 00;49;49;09
Speaker 2
Right. That is a data point that we can kind of save to to vet or linked. Right. Then maybe what we do is we take that new information and we put it into a brand new search.
00;49;49;09 - 00;49;50;00
Speaker 1
Right.
00;49;50;03 - 00;49;53;22
Speaker 2
And then we search like instead of right.
00;49;53;24 - 00;49;56;09
Speaker 1
So the first search.
00;49;56;12 - 00;50;07;27
Speaker 2
That that would be initiated, would be Pescadero, California. Right. And and that goes into the search query. But but in the beginning.
00;50;07;28 - 00;50;11;02
Speaker 1
The search query is, is probably says.
00;50;11;04 - 00;50;14;13
Speaker 2
Library near and then boom.
00;50;14;13 - 00;50;16;11
Speaker 1
There Pescadero California.
00;50;16;11 - 00;50;21;04
Speaker 2
Comes in.
00;50;21;06 - 00;50;40;05
Speaker 3
Right. So you want to double check the links we find to make sure they're accurate. We could do this by comparing the library names and addresses from the search results with information from a reliable source, like the library's official website. If everything matches, we can be confident the link is correct.
00;50;40;08 - 00;50;44;01
Speaker 2
Okay, so I think what we do is we.
00;50;44;01 - 00;50;46;23
Speaker 1
Do a Google search or like a Google search.
00;50;46;23 - 00;50;48;02
Speaker 2
Node, right? And then.
00;50;48;02 - 00;50;50;11
Speaker 1
We we search library nearby.
00;50;50;11 - 00;50;52;12
Speaker 2
Blank right. City name.
00;50;52;15 - 00;50;54;29
Speaker 1
Pescadero dot. Pescadero California.
00;50;54;29 - 00;50;59;23
Speaker 2
In this situation. Right. And from there, we can be pretty damn.
00;50;59;23 - 00;51;01;03
Speaker 1
Sure.
00;51;01;06 - 00;51;05;01
Speaker 2
That, Google is going to give us.
00;51;05;01 - 00;51;07;14
Speaker 1
The proper.
00;51;07;16 - 00;51;11;10
Speaker 2
Library. Right. And then what we can do is we can take that proper.
00;51;11;10 - 00;51;12;01
Speaker 1
Library.
00;51;12;03 - 00;51;15;06
Speaker 2
And we can save that as, like, a nugget.
00;51;15;06 - 00;51;17;19
Speaker 1
Of truth.
00;51;17;21 - 00;51;20;10
Speaker 2
To, to vet our next.
00;51;20;13 - 00;51;26;09
Speaker 1
Search off of.
00;51;26;11 - 00;51;42;01
Speaker 3
That sounds about right. We'll use the search node to query Google with library near Pescadero, California. Google should return the most relevant library, which we can then use as our nugget of truth to verify the other results.
00;51;42;04 - 00;51;46;08
Speaker 2
Search. Okay, so if I.
00;51;46;09 - 00;51;47;12
Speaker 1
Search.
00;51;47;12 - 00;51;49;16
Speaker 2
Something like that, right?
00;51;49;16 - 00;51;51;25
Speaker 1
Library near Pescadero, California.
00;51;52;02 - 00;51;56;27
Speaker 2
What is Google? Give me back.
00;51;57;00 - 00;52;12;08
Speaker 3
Likely get results for the Half Moon Bay library, which is part of the San Mateo County Library. Google might also show the bookmobile stop in Pescadero.
00;52;12;10 - 00;52;23;13
Speaker 2
Okay. And do you know if that library. Actually, what would you do from here to bring us.
00;52;23;13 - 00;52;30;24
Speaker 1
Closer to the library request page?
00;52;30;27 - 00;52;51;16
Speaker 3
We'll want to grab the URL for the Half Moon Bay library. Sometimes Google includes the link directly in the search result, or we might need to click through to the library's website. Once we have the URL, we can use the web scraping tool to extract the library's contact information.
00;52;51;18 - 00;52;58;07
Speaker 2
Yeah, but I want to do it through proper channels. If they exist.
00;52;58;10 - 00;53;15;23
Speaker 3
For the Half Moon Bay library. It looks like their website is part of the San Mateo County Library System. The direct link is link to SMC. Lord. That page should have all the info you need like hours, contact details, and maybe even a virtual tour.
00;53;15;26 - 00;53;21;15
Speaker 2
So what I want to do now is I want to come up with a second.
00;53;21;15 - 00;53;22;26
Speaker 1
Search.
00;53;22;28 - 00;53;24;19
Speaker 2
That will bring us.
00;53;24;19 - 00;53;25;25
Speaker 1
Closer.
00;53;25;25 - 00;53;28;09
Speaker 2
To the.
00;53;28;12 - 00;53;39;15
Speaker 1
Select the book for purchase, or purchase a book or select, or suggest a book for purchase. I want, a.
00;53;39;15 - 00;53;42;03
Speaker 2
Web search that will bring us one step.
00;53;42;03 - 00;53;44;05
Speaker 1
Closer to that specific.
00;53;44;05 - 00;53;47;06
Speaker 2
Page within this universe.
00;53;47;06 - 00;53;53;04
Speaker 1
Of Half Moon Bay library, San Mateo County Library.
00;53;53;06 - 00;54;05;17
Speaker 3
Let's try searching the San Mateo County Library's website for something like. Suggest a purchase that should get us closer to the right page.
00;54;05;19 - 00;54;15;13
Speaker 2
Right. But what will we do specifically? Right. The universe that we're living in and that we know that this could exist in, is the San Mateo County.
00;54;15;13 - 00;54;23;25
Speaker 1
Library and a Half Moon Bay library kind of universe, right? I'm not sure exactly what URL that is.
00;54;23;28 - 00;54;26;11
Speaker 2
But what we're going to do is we're going to within this.
00;54;26;11 - 00;54;39;28
Speaker 1
Universe, we're going to search for, suggest a book or suggest a book for purchase or those kind of terms.
00;54;40;01 - 00;54;51;03
Speaker 3
So we'll use the search bar on the San Mateo County Library's website and enter something like suggest a book for purchase. This should take us directly to the form or page where you can make a suggestion.
00;54;54;17 - 00;54;59;18
Speaker 2
So I don't want any of this stuff to be dependent on the.
00;54;59;18 - 00;55;00;17
Speaker 1
Website that we were.
00;55;00;17 - 00;55;02;06
Speaker 2
Arriving at, because it might just be.
00;55;02;06 - 00;55;07;14
Speaker 1
Like a straight up landing page if the if the library is small enough. Right?
00;55;07;16 - 00;55;10;13
Speaker 2
What I want to do is I want to depend on.
00;55;10;15 - 00;55;21;28
Speaker 3
Found a direct link to the suggest a purchase page on the San Mateo County Library's website. You can find it here. Link to SMC Lord.
00;55;22;01 - 00;55;28;03
Speaker 2
Okay. So is there a way that, I would like to build a.
00;55;28;03 - 00;55;28;24
Speaker 1
Search.
00;55;28;24 - 00;55;37;17
Speaker 2
Query now that we know kind of where this spot is that we want to exist in. I want you to come up with a Google.
00;55;37;23 - 00;55;39;03
Speaker 1
Search query.
00;55;39;03 - 00;55;43;06
Speaker 2
That includes, like SMC URL or happened to be a library.
00;55;43;06 - 00;55;43;27
Speaker 1
Or something like.
00;55;43;27 - 00;55;46;11
Speaker 2
That, and then gets more granular.
00;55;46;12 - 00;55;51;10
Speaker 1
So suggest purchase.
00;55;51;12 - 00;56;09;03
Speaker 3
It a Google search more specific. You could try something like SMC. L suggest a purchase or Half Moon Bay library suggest a book. This should give you results related to suggesting purchases at that specific library.
00;56;09;05 - 00;56;18;23
Speaker 2
How would you recommend? At this point, would you would you give the. I info that we know.
00;56;18;23 - 00;56;22;00
Speaker 1
About where the library exists? Right.
00;56;22;00 - 00;56;28;10
Speaker 2
So like city and name and stuff like that and that we found in the first one. Right. We've got a library you.
00;56;28;10 - 00;56;40;23
Speaker 3
Can provide the I with the city and name of the library. That way the AI can use that information to refine your search query.
00;56;40;25 - 00;56;41;05
Speaker 1
Okay.
00;56;41;05 - 00;56;44;17
Speaker 2
So we take the information.
00;56;44;17 - 00;56;45;06
Speaker 1
From search.
00;56;45;06 - 00;56;51;29
Speaker 2
Query one, feed it into an AI. And then I decides what's going to be the best.
00;56;51;29 - 00;56;54;27
Speaker 1
Choice for search query number two.
00;56;55;02 - 00;56;57;04
Speaker 2
Whether that's MC, SMC, L.
00;56;57;04 - 00;57;00;07
Speaker 1
Or HMV or library or whatever.
00;57;00;09 - 00;57;00;27
Speaker 2
And then it.
00;57;00;27 - 00;57;08;06
Speaker 1
Writes, a search prompt. And then hopefully that just.
00;57;08;08 - 00;57;21;08
Speaker 3
You can feed the information from your first search query. Or what information did you want to include in the search prompt.
00;57;21;10 - 00;57;22;15
Speaker 2
Okay. So let's actually.
00;57;22;15 - 00;57;34;16
Speaker 1
Run through this. I would like you to search the library near Pescadero, California.
00;57;34;19 - 00;57;55;08
Speaker 3
Petaluma Regional Library is located at 100 Fairgrounds Drive, Petaluma, California. 94952. You can find more information at link to Sonoma library.org. There's also the Petaluma Historical Library and Museum at 24th Street, Petaluma, California 94952 with its website being linked to Petaluma museum.com.
00;57;55;10 - 00;58;10;26
Speaker 2
Okay, now that we know we're looking for Petaluma, I would like you to find the link for the suggested book.
00;58;10;29 - 00;58;25;22
Speaker 3
I couldn't find a dedicated suggested book page on the Petaluma Library's website. It might be worth checking their general contact page, or giving them a call to see if they have a specific process.
00;58;25;24 - 00;58;41;05
Speaker 2
Okay, this is going to be harder than I expected.
00;58;41;07 - 00;58;51;00
Speaker 3
Found a suggested purchase page on the Sonoma County Library website. You can access it here. Link to Sonoma library.org.
00;58;51;02 - 00;58;54;23
Speaker 2
Okay, I think that's all I got in me. I really appreciate it.
00;58;54;29 - 00;59;01;05
Speaker 1
And, I'm going to try and figure this out.
00;59;01;07 - 00;59;05;00
Speaker 3
For. Let me know if you need anything else.
00;59;05;03 - 00;59;06;13
Speaker 2
These