-
Notifications
You must be signed in to change notification settings - Fork 0
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
PML-TQ Search issues #5
Comments
Right - of course I tested a lot, but forgot that the PML interface reads from a Git that is not reachable from the live server; so I just copied the relevant data from the Git to the NFS, that solves many of the errors |
The 50 limit is easy to make editable, but since it has to do quite a few steps, it will be slow to display too many results at a time - ideally, I would be able to tell the API to give me the next batch of results - can it do that? |
And pinpointing the result is indeed tricky: it does now do it properly the other way around - if you click on any word or link to get to the PML view, it will show you which word you clicked on. But the pmltql server gives me back a node from any of the layers - since we are showing TEITOK XML, it needs to know what that corresponds to; and the TEI only contains the corresponding w-layer node for the token, not anything higher up. And it is tricky to quickly detemine the w-layer node corresponding to the result of the pmltq-server. It should work for PDT, where you can predict the correspondence (t-XXX-p2s4w1 will correspond to w-XXX-p2s4w1) but none of the other parts use systematic naming - so it is tricky to determine the filename as well - I had to resort to building a correspondence table between any s-type node in the PML to the corresponding s-id in TEITOK. |
And the clicking back is always tricky (strange that the back button does not work) - would opening the links in a new tab be better? |
Yes, this would do great.
Jiří
From: "Maarten Janssen" <[email protected]>
To: "ufal/teitok-tools" <[email protected]>
Cc: "Jiri Mirovsky" <[email protected]>, "Author" <[email protected]>
Sent: Wednesday, 27 January, 2021 15:22:48
Subject: Re: [ufal/teitok-tools] PML-TQ Search issues (#5)
The 50 limit is easy to make editable, but since it has to do quite a few steps, it will be slow to display too many results at a time - ideally, I would be able to tell the API to give me the next batch of results - can it do that?
—
You are receiving this because you authored the thread.
Reply to this email directly, [ #5 (comment) | view it on GitHub ] , or [ https://github.com/notifications/unsubscribe-auth/AEGAYYLIDLZJQD3E5OEV55LS4AOTRANCNFSM4WVJUBAA | unsubscribe ] .
|
I guess clicking on "context" might open the context in a new tab - then the list of results would stay intact.
JM
From: "Maarten Janssen" <[email protected]>
To: "ufal/teitok-tools" <[email protected]>
Cc: "Jiri Mirovsky" <[email protected]>, "Author" <[email protected]>
Sent: Wednesday, 27 January, 2021 15:28:45
Subject: Re: [ufal/teitok-tools] PML-TQ Search issues (#5)
And the clicking back is always tricky (strange that the back button does not work) - would opening the links in a new tab be better?
—
You are receiving this because you authored the thread.
Reply to this email directly, [ #5 (comment) | view it on GitHub ] , or [ https://github.com/notifications/unsubscribe-auth/AEGAYYLKVHE4G4N6VF3H24DS4APJ3ANCNFSM4WVJUBAA | unsubscribe ] .
|
I was hoping PMLTQ server could do that, but the API does not seem to support that; because now it has a next button by slicing from the API reponse, but that itself is limited and there you cannot do next (the limit is 10k though and not 50, but it is an uncomfortable mixture) |
But any idea how to quickly and reliably get the w-node for each result of a query in the API? |
Maybe the current author of the PML-TQ might know ("Matyáš Kopp" <[email protected]>). This goes for the other topic (50 results at a time) too.
From: "Maarten Janssen" <[email protected]>
To: "ufal/teitok-tools" <[email protected]>
Cc: "Jiri Mirovsky" <[email protected]>, "Author" <[email protected]>
Sent: Wednesday, 27 January, 2021 23:40:00
Subject: Re: [ufal/teitok-tools] PML-TQ Search issues (#5)
But any idea how to quickly and reliably get the w-node for each result of a query in the API?
—
You are receiving this because you authored the thread.
Reply to this email directly, [ #5 (comment) | view it on GitHub ] , or [ https://github.com/notifications/unsubscribe-auth/AEGAYYOJXJ3ZN64K2AQL7FDS4CI4BANCNFSM4WVJUBAA | unsubscribe ] .
|
When I submit a query in the "PML-TQ Search" section and then click on "PML" for a specific result, I get:
Fatal error: Uncaught Error: Call to a member function xpath() on null in /var/www/html/teitok/ufal/Sources/pmlview.php:453 Stack trace: #0 /home/janssen/Git/TEITOK/common/Sources/main.php(204): include() #1 /mnt/nfs/teitok/teitok-main/pdtc/index.php(14): include('/home/janssen/G...') #2 {main} thrown in /var/www/html/teitok/ufal/Sources/pmlview.php on line 453
When I run a PML-TQ query (Execute Query button), the interface looks like nothing is happening and that maybe I should press the button again. Here, it would want some more visible response clearly indicating "wait for the result".
You also can't get more than 50 results, it would be nice to be able to get more.
For individual results, it is not indicated which part of the sentence matches the query. Also, when I display the result tree, it is not possible to see which nodes match the query. But I understand that it would be difficult to do, the tool can be used without it. But if it was possible, it would be great, because displaying the results as a list of sentences is very nice and it would be an added value compared to using the original PML-TQ interface.
When I click on "context" for a specific result, I see the context but I cannot to get back to the list of results (or do not know how, the back button does not work).
Below the context, there are some links that also result in some errors (links "Dependencies", "PML-TQ Tree View").
The text was updated successfully, but these errors were encountered: