Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    mkcrandall
    @mkcrandall
    Okay. Understood. Thank you for the effort. If you want, I can un-comment the <script> tags in the HTML. The styling gets a little off then, but the search function works then if it's useful for comparison. However, that will not be an idea solution as I'm hoping not to have to manually add that script tag to multiple hundreds of web pages, which is why we have a single javascript to inject shared content across all of them.
    Jaroslav Vaňkát
    @JVankat
    Hey @mkcrandall the search seems to be working, and the search box seems to have all classes now, are there still any issues for you?
    James
    @jmshey90
    @JVankat is there an easy way of changing the results links to relative?
    Jaroslav Vaňkát
    @JVankat
    @jmshey90 not really, you could try to use a postSearch (to replace the links by modifying the DOM), or a preRender (to replace the links directly in the search response) callback, but that would be the only options I can think of atm.
    James
    @jmshey90
    Thanks, I'll give them a go
    James
    @jmshey90
    PostSearch works for my purposes, thanks!
    Jaroslav Vaňkát
    @JVankat
    Perfect!
    mkcrandall
    @mkcrandall
    @JVankat Good morning (or good afternoon there). The search is only working because I uncommented the script that calls sitesearch360-v13.min.js that is part of the original HTML of the page. (Not injected by navb.js.) When the script is part of the original page, the search works. But I need to be able to inject it from an outside file (the navb.js) is the problem. When I remove the script from the actual page code and try to inject it solely with a javascript, that's when it won't work.
    Jaroslav Vaňkát
    @JVankat
    @mkcrandall and did you try the last code I've sent you? Otherwise I'm not at my computer anymore, so I'll be able to take a look on Monday.
    mkcrandall
    @mkcrandall
    If you're referring to the code here: https://pastebin.com/TAcZsa1K then, 'yes' and sadly, it did not seem to change the results. (The results being that sitesearch360-v13.min.js is still not being executed on the page.)
    Jaroslav Vaňkát
    @JVankat
    @mkcrandall looks like I didn't save the changes on Pastebin, please try the following one - https://pastebin.com/LSnfnQ8K
    mkcrandall
    @mkcrandall
    @JVankat Great! We've made headway! Thank you. This does execute the search function, however all it displays are the page image (or placholder icons) and the title. There is no sample text with the search term(s) highlighted as previously. I've been looking through your ss360Config configuration options and I'm not finding what I assume is missing that's causing this result. Any ideas?
    Jaroslav Vaňkát
    @JVankat
    @mkcrandall the search snippet is not being extracted since it's diabled in the control panel under search settings.
    mkcrandall
    @mkcrandall
    Great! I love an easy answer and an easy fix! :) That was probably changed when I was poking around trying to troubleshoot it last week.
    Now I just need to re-adjust the CSS. Do I need to call the CSS separately (with it's own .css file), or can I put a long string of it in the ss360Config settings using the "additionalCss" parameter? Which would you suggestion?
    Which would you suggest, that is. still waking up and my typoes are showing that.
    Previously I'd been calling the CSS from it's own .css file.
    Jaroslav Vaňkát
    @JVankat
    @mkcrandall having this in an own .css file is the better approach in my opinion, as long as this is easy enough for you to manage. The additionalCss parameter is mainly there to target some specific use-cases where the customer cannot easily access the site's stylesheets.
    mkcrandall
    @mkcrandall
    Yeah, I think that visually, it be much easier to manage and edit/troubleshoot that way too as opposed to having it all crunched into one long string. Thank you.
    Jaroslav Vaňkát
    @JVankat
    @mkcrandall definitely, also you don't need to worry about string escaping :) Glad to help, just let me know if you run into any other issues.
    AEKercher
    @AnneKercher
    Hello. We have many pages behind a custom login. I entered our login url, the login "form" xpath, as well as the username and password inputs and values. However, when I try to index a password protected page, it just indexes our login page (protected pages redirect to login page). Could you please advise on how to get this working?
    Jaroslav Vaňkát
    @JVankat
    Hi @AnneKercher - we don't currently support custom login screens that just trigger custom JS handlers (so basically a POST request on the form submit is required), might that be the issue here?
    AEKercher
    @AnneKercher
    @JVankat I dont think so. Our login Submit button sends a POST request to our login api and returns a token. our login inputs are not in their own form (they are in the main page form). Would that have something to do with it?
    Jaroslav Vaňkát
    @JVankat
    @AnneKercher can you send me a link to the login page? I will take a look.
    AEKercher
    @AnneKercher
    Jaroslav Vaňkát
    @JVankat
    @AnneKercher thank you, I'll take a look and will get back to you in ~30 minutes.
    Jaroslav Vaňkát
    @JVankat
    @AnneKercher looks like we don't currently support this type of authentication, I think the authentication should be rather easy to enhance on our side, but let me take a deeper look and I'll get back to you tomorrow.
    AEKercher
    @AnneKercher
    @JVankat Thank you. I could possibly modify our login page code behind to recognize when the requests are coming from your crawler, and authenticate then, but would need more info on how your crawler is indexing. I am trying to find a way to utilize the SS360 crawling capability as much as possible, so any suggestions on how to get your crawler to see our protected pages would be appreciated.
    Jaroslav Vaňkát
    @JVankat
    @AnneKercher we do use the Mozilla/5.0 (compatible; SiteSearch360/1.0; +https://sitesearch360.com/) user agent, and the IPs listed over here, if that's helpful. But if you can wait until tomorrow, then I'll definitely look into it since a very similar request came to our mail support earlier today.
    AEKercher
    @AnneKercher
    @JVankat Yes it can wait until tomorrow, thank you.
    Jaroslav Vaňkát
    @JVankat
    Hi @AnneKercher - I think I've got a fix, do you have a sample page that requires a login to be accessible?
    Jaroslav Vaňkát
    @JVankat
    @AnneKercher thank you, I will test this within the next hour and will keep you posted.
    Jaroslav Vaňkát
    @JVankat
    @AnneKercher looks like my fix wasn't even necessary, I did just have to add another parameter to the authentication settings, can you please confirm it works as expected now?
    AEKercher
    @AnneKercher
    @JVankat yes it seems to be working now! thank you so much! did you need to change something in our site setup or was the change on your end?
    Jaroslav Vaňkát
    @JVankat
    @AnneKercher perfect! I did just have to add an additional property to the login parameters in your settings - had to explicitly set the LOGIN event.
    mkcrandall
    @mkcrandall
    QUESTION: Does re-indexing your site impact user's ability to access the existing index? Example: I tell SiteSearch to 'Reindex All Configured Sources'. While this is happening (which according to the logs takes roughly an hour) does it impact my end-user's ability to use the search engine at all? Do they have access to new indexed pages/documents as the re-indexing is happening, or not until the re-index is complete?
    Jaroslav Vaňkát
    @JVankat
    Hi @mkcrandall - no, re-indexing doesn't have any impact on the search, so new pages should be retrieved as soon as they are indexed, and old pages (removed ones) won't be shown just after removed from the index (unless the query is cached, we do clear the query cache after the full re-index finishes).
    James
    @jmshey90
    @JVankat the extraction/index facilities don't appear to be working
    Just keeps loading
    Jaroslav Vaňkát
    @JVankat
    @jmshey90 what exactly isn't working for you?
    James
    @jmshey90
    The Test Extraction or Index Single Page services, I enter a URL, then the button, then it spins until it timesout
    Jaroslav Vaňkát
    @JVankat
    @jmshey90 do you have a sample URL I could use to test this?
    Jaroslav Vaňkát
    @JVankat
    Thanks, I'll get back to you in a few minutes.
    Jaroslav Vaňkát
    @JVankat

    @jmshey90 again, that's the same issue I've explained last time, something on your side seems to be blocking our servers, switching to a different one seems to do it. So back to my question from July 23rd:

    @jmshey90 so I did take a look, and actually not sure what went wrong, the only reason for this seems to be your server blocking (or rather ignoring) our server (usually due to too many requests). But I'm not 100% sure about this, are you aware of some component that might be causing this on your side?

    James
    @jmshey90
    Do you know which server it was blocking?
    Or the IP address
    I did contact our IT department and they can't see any blocking from the IPs I saw on your docs, but better to check when an actual problem occurred
    Jaroslav Vaňkát
    @JVankat
    This was our EU API, so it should be 88.99.162.232.