SPONSORS:






User Tag List

Results 1 to 10 of 10
  1. #1
    Senior Member
    Join Date
    Jul 2002
    Location
    Paris (France)
    Posts
    182
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    A twister of a problem: link recognition

    I have a problem that I have been unable to solve for quite some time. The problem is when there is a listing of HTMLlinks, SilkTest recognizes all of them as one big HTMLText block.

    The key that makes Silk view the objects incorrectly is that it doesn't quite see a Browser Object. JavaScript is being used to open a new browser, with all browser attributes removed (I have a demo site setup for you to see).

    The application under test actually is made up completely of these types of browser windows. I have gotten around the problem by declaring a DialogBox with a child BrowserChild window. This works throughout the application, except in windows that have a list of links. In which case all links are listed as one HTMLText file.

    So, the Question: How can I get SilkTest to reconizie these?

    I have setup a sample page at http://jeepdog.archerbrothers.com/silktest/ that shows this problem. Select the "Click to Open" link to get the problem page. Now, since this was a demo page, the links will not work, but it is the same link code as the original problem.

    For reference, here is the declaration of the problem window:


    Notice that the listing of HTMLLinks show up as only one HTMLText object (ChangeJobForThisEmployee).

    So, any advice on how to get the HTMLLink objects to be recognized?



    ------------------
    David Genrich
    Icarian
    555 North Mathilda Ave
    Sunnyvale, CA 94086
    davidg@icarian.com

  2. #2
    Advanced Member
    Join Date
    Jan 2006
    Posts
    995
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: A twister of a problem: link recognition

    *Imported from comp.software.testing*

    Sometimes I face that problem also. Recently I catched the problem of
    SilkTest, when it recognized HTML links as HTML headers. It's very possible,
    especially if links are connected with Java Scripts to open the link
    location in new browser window. From my experience I could suggest you three
    ways to solute that problem.

    1. Try to record your test frame and perform your web test under different
    type of browser. For Instance if you're trying to record frame declaration
    under IE 5.0., try to do that and run your tests under NN 4.0. or IE 4.0. I
    observed that SilkTest still has some problems with supporting IE 5.0.
    2. Reinstalling Silk. Sometimes it works out also. Try to reinstall Silk and
    do your recording again. Also if you use Silk 5.0.2., I would reccomend you
    to install back Silk 5.0.1.. I faced that Silk 5.0.1. works more stable
    than Silk 5.0.2. Perhaps we need to wait for next upgrade from Segue.
    3. If two previous was have not worked out, then you can declare custom
    objects in your big HTML text block. It's quite easy solution, especially if
    you HTML links are static, and contents of the group HTML links (recognized
    as HTML block) is not changed depending on the situation.
    Hopefully it will help.

    Best regards,
    Andrew Issaenko

    aissaenko@mail.com

  3. #3
    Senior Member
    Join Date
    Jul 2002
    Location
    Paris (France)
    Posts
    182
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: A twister of a problem: link recognition

    1. Problem is the same under IE or Netscape.

    2. Have installed Silk on other systems, no help. This problem has been around since 4.x days. Now I am on 5.01 (afraid to go to 5.02).

    3. Depending on the context of where I am and what user I am logged in as, the links my vary as far as which onces are present (only relative links are shown).



    ------------------
    David Genrich
    Icarian
    555 North Mathilda Ave
    Sunnyvale, CA 94086
    davidg@icarian.com

  4. #4
    Senior Member
    Join Date
    Jul 1999
    Location
    Burlingame CA 94010
    Posts
    502
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: A twister of a problem: link recognition

    One thing I've done may seem strange, but it works. First, see if the HtmlText object is a multi-line text field. Get its contents using lsContents = wObject.GetMultiText (). Get the position of the link in the list of string. Get the RECT of the HtmlText object, figure out where the item you want is in the list, then figure out the location to click on based on the number of pixels between each row. It sounds funky, but it works consistently. You can even write a routine to search all the HtmlText objects on the page to find the one that contains the text or link you are actually looking for.

    ------------------
    John W Green
    jwgreen@automationexpertise.com
    http://www.automationexpertise.com

  5. #5
    Senior Member
    Join Date
    Jul 2002
    Location
    Paris (France)
    Posts
    182
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: A twister of a problem: link recognition

    You were right on the money John

    Here is the function I wrote for handling the links:

    ------------------
    David Genrich
    Icarian
    555 North Mathilda Ave
    Sunnyvale, CA 94086
    davidg@icarian.com

  6. #6
    rg
    rg is offline
    Member
    Join Date
    Feb 2000
    Posts
    98
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: A twister of a problem: link recognition

    Try the following setting:

    BrowserPage.SetUserOption ("JoinStackedText", FALSE, USEROPT_DEFAULT)

    You need to run this line of code every time you start the browser and/or the agent.

    Hope that helps.
    RG

  7. #7
    Junior Member
    Join Date
    Dec 1999
    Location
    Pleasanton, CA
    Posts
    16
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: A twister of a problem: link recognition

    I have a funky problem that is similier to this one (well, it maybe different):

    I use silk 5.01 and IE 5.01. I want to click a link, but alway get a item not found error. I found a funly thing that every time I record the same window I get different result. sometime the link I want is in htmltable1, other time it is in htmltable3. also the tag is different. This makes it impossible to write a code to click the link. However, I could do so by record the testcase, that make the code with many long red tag information in it and hard to read and maintain. Still worse is that the object identifiers in the recorded code is totally different (table, identifier, and tag) from that in the recorded window declearation. Anybody out there have the same experience and got a way to work around this. If this cannot be solved, I doubt silk is a good tool for Web testing.

    ------------------
    jim.young@documentum.com

  8. #8
    Senior Member
    Join Date
    Jul 2002
    Location
    Paris (France)
    Posts
    182
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: A twister of a problem: link recognition

    The problem you are having sounds more like a table recognition problem.

    I have found that if a table is mal-formed (maybe the top row has 3 columns, all other columns have 4 ) HTML will accept it, Silk will go nuts on it.

    So, I found a few Internet Explorer plugins (well, HTML code that is ran with a right-click) to show what the tables actually look like. Also have one for showing the cookies. Both of these have helped both me and the developers.

    To install: Unzip, right-click on the INF file and select install. The functions will be in IE's context menu (when you right click on a web page).

    ie_content_menu.zip

    ------------------
    David Genrich
    Icarian
    555 North Mathilda Ave
    Sunnyvale, CA 94086
    davidg@icarian.com

    [This message has been edited by davidg (edited 02-26-2000).]

  9. #9
    Junior Member
    Join Date
    Dec 1999
    Location
    Pleasanton, CA
    Posts
    16
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: A twister of a problem: link recognition

    Thank you David!

    I got the plug_ins and they are great. It, however, does not solve the problem though.

    Actually I could see the tables (There are many ways we can see the actuall table, eg. use MS frontPage). The problem is: even we can see the actuall table, silk cannot. I know my link is in table is in table 58 column 3. If I refer to this silk still cannot find it because the declearation did not record the window the way it should be. Therefore, I believe we need a way to declear the window right. Another proble is that silk has a strange rule, which is that "in order to be seen as a table, an object must have at least two columne s and at least two rows" (Segue "4Test language preference" page 573). My tables happen to be those kind that only have one row. Why they have this rule? Why one row table is not qualify for a table?

    Do you have any idea how to make declearation right?

    ------------------
    jim.young@documentum.com

  10. #10
    Senior Member
    Join Date
    Jul 2002
    Location
    Paris (France)
    Posts
    182
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: A twister of a problem: link recognition

    this is getting into table recongnition rather than link recongintion, which should be in a diffrent thread so it is easier for people to follow.


    ------------------
    David Genrich
    Icarian
    555 North Mathilda Ave
    Sunnyvale, CA 94086
    davidg@icarian.com

 

 

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

vBulletin Optimisation provided by vB Optimise v2.6.0 Beta 4 (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
User Alert System provided by Advanced User Tagging v3.0.9 (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Questions / Answers Form provided by vBAnswers (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
vBNominatevBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Feedback Buttons provided by Advanced Post Thanks / Like (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Username Changing provided by Username Change (Free) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
BetaSoft Inc.
Digital Point modules: Sphinx-based search
All times are GMT -8. The time now is 02:26 AM.

Copyright BetaSoft Inc.