‹ Home
Exported

TOLOGIX - Infrastructure LawGuide (ILG)

Harsh Parikh, Tech Lead at DevITMelissa Cowell, General Manager at IndustrialDevaang BhattMorgan Maguire, CEO at TologixHiren PatelMarysia RaptisPaul MoonIrit WeinfeldPrerak Shah, Executive Director & Jt CEO at DevITPatrick Harris, Web Designer at IndustrialPiyush Kanpariya, DevITRob Wiesenberg, ContegraRadomir Mladenovic, ContegraKetan Sondarva, Technical Project Manager at DevITStephanie Ducasse, Web Developer at Industrial +10 others
Posted by Stephen Ceresia ·

PDF -> HTML conversion issues

Hi all,

As discussed during our last P3 call,  Anil Vaghela Anil noted issues were encountered converting PDF to HTML using an Adobe API as it is not supported with .NET Core.

A couple of alternatives were proposed:

- Using a windows service hosted on the Carbon60 server. While this would solve the issue, it would result in performance issues (e.g. document conversion taking up to several minutes). Other concerns include maintainability of this service.

- Using MVC instead of .NET Core to maintain use of the original APIs.

I've opened this thread to discuss these options to make sure we have explored all possible alternatives so Morgan has enough information to make a final decision on how to move forward.

Some questions (mainly for Anil Vaghela Anil at this point):

- What were the issues that were encountered with the Adobe API using .NET Core?
- Can you provide more details on this windows service? I assume it's a third-party solution?
- Issues with the search API were also noted during the call: are these also compatibility related?
- Are there alternative APIs that can be used in place of PDF -> HTML conversion and the Search?
- Can we get a better idea of how long it's expected document conversion would take using the windows service solution (e.g. for a 1 MB, 10 MB, or 50 MB PDF?)

Morgan Maguire, CEO Morgan Anil Vaghela Anil if I missed anything, feel free to add. 

Thanks,
Stephen

Comments & Events

Morgan Maguire, CEO Morgan Maguire, CEO
Hello all,

Just to elaborate on Stephen Ceresia, Industrial Stephen 's comments above, the crux of the issue is that .NET Core is hindering our ability to use certain third-party APIs that are critical to the P3 application (Adobe PDF/HTML converter and dtSearch Full Text Search - dtSearch is also important in ISLG).  My understanding is that the following options are available:

  1. Stop using .NET Core and go back to MVC, which is more compatible with third-party APIs;
  2. Continue using .NET Core and run Adobe and dtSearch on separate applications; or
  3. Continue using .NET Core and find alternative third-party APIs (or in-house built solutions) that are compatible with .NET Core.

Could Anil Vaghela Anil , Juan Silva, Industrial Juan and Derek Ethier, Industrial Derek please comment on each of these options, and provide recommendations. Also, if option 3 is preferred, please suggest the best course of action for replacing Adobe and dtSearch.

Thanks,

Morgan
Notified 7 people
Anil Vaghela Anil Vaghela
Hello All,

Just wanted to mention few things for above issues:

Issue 1 - Issue with PDF to HTML Conversion on Carbon60 Server's IIS:
For some reason AdobePro API gives error on Carbon60's IIS server when we are trying to convert PDF file to HTML file using P3 App. As a temporary solution we have created a separate window service and hosted on the Carbon60 server. This solution works well. 

Issue 2 - dtSearch API isn't supported by .Net Core:
dtSearch's third party DLL is not working with .Net Core application. This works well with MVC's latest version.

Hello Stephan, my feedback for your below questions:

- What were the issues that were encountered with the Adobe API using .NET Core?

For some reason AdobePro API gives error on Carbon60's IIS server when we are trying to convert PDF file to HTML file using P3 App. As a temporary solution we have created a separate window service and hosted on the Carbon60 server. This solution works well. Please note that this works well in our local environment when we try to covert PDF to HTML using P3 app (.Net core).

- Can you provide more details on this windows service? I assume it's a third-party solution?

We have developed a windows service which is not a third-party solution. We can develop a window service in Visual Studio .Net similar as a web application. We have full control on this window service and can modify it whenever needed. The only issue is window service is not a part of P3 app. P3 app and window service are two separate applications.

- Issues with the search API were also noted during the call: are these also compatibility related?

As per our research so far it looks like the issue is compatibility related. The dtSearch API works well with MVC's latest version but not working with .Net core.

- Are there alternative APIs that can be used in place of PDF -> HTML conversion and the Search?

We require to explore alternative APIs which are similar to dtSearch.

- Can we get a better idea of how long it's expected document conversion would take using the windows service solution (e.g. for a 1 MB, 10 MB, or 50 MB PDF?)

We checked the sample files provided by Morgan. Most of the files' size are below 1 MB. We converted few sample files using window service, this conversion would take 40-50 seconds. If file size is bigger then this may take more time.

Hi Morgan,

Please see my comments for your below points:

- Stop using .NET Core and go back to MVC, which is more compatible with third-party APIs;
          We can convert .Net Core application to MVC's latest version at any time. This can be done within a week. This could be the last solution if we will not have any other solution. 

- Continue using .NET Core and run Adobe and dtSearch on separate applications; or
     We can do this with very few limitations.

- Continue using .NET Core and find alternative third-party APIs (or in-house built solutions) that are compatible with .NET Core.
       For time being we can go with separate applications for Adobe and dtSearch. Once we complete all stories after that we can explore for alternative APIs.
Notified 7 people
Stephen Ceresia, Industrial Stephen Ceresia
Hi Anil Vaghela Anil Morgan Maguire, CEO Morgan ,

I discussed this with Derek Ethier, Industrial Derek today, and before making any technical suggestions we would like to get a better technical understanding behind some of the issues encountered, and also be clear on the reason for using dtSearch and the AdobePro API so we know exactly what options we have available.

Could we schedule a call for one day next week to discuss in more detail?

Let me know when works best for you and I'll schedule it in.

Thanks,
Stephen
Notified 7 people
Morgan Maguire, CEO Morgan Maguire, CEO
Sounds good to me Stephen Ceresia, Industrial Stephen . Could we discuss this during our call scheduled for next Thursday?

Thanks,

Morgan
Notified 7 people
Stephen Ceresia, Industrial Stephen Ceresia
That works Morgan Maguire, CEO Morgan . I'll update the invite.

Thanks,
Stephen
Notified 7 people