Character Mystery

I originally posted this in the Illustrator forum, but it's probably more appropriate here.
A web programmer friend asked for some help with a very odd character problem he's having with an XML file. There is a character within the file that is causing a parsing problem. I copied the line into Illustrator thinking that the glyphs panel would reveal what it is. It doesn't. The character appears either as a box or as a diamond with a question mark in it. When I highlight the character, nothing shows up for the glyph. Weirder yet, if I change the font for the entire line of text, the character doesn't change. Anyone have any idea on how to figure out what this thing is? This is the text (I don't know whether passing it through to the forum will change or remove it, but it's the character after the Rn).
Rn�_Mapk1 Control
The forum software converts it to an HTML entity, one I believe means 'out of range'. The following is an attempt to paste in the offending character without quoting it:
Rn�_Mapk1 Control
I've been trying to GREP the character in ID, but can't even paste it into any field. It's like the character doesn't really exist, but yet I can see it in the text.

>If you know the unicode, you can search for any character in regular mode (enter as hex: "<fffd>") or in grep mode (as escaped hex: "\x{fffd}"). <br /><br />That's just it: I don't know and can't seem to extract the unicode for it. I tried copying and pasting into different dialogs, but nothing appears when I paste. I did find out that the original text came out of an SAP system in Germany, but that hasn't really helped track down what the weird character is. There are two of them, presumably identical, in this particular file.<br /><br />I did try Kenneth Benson's suggestion to export it as tagged text, but the tagged text either shows a square, FFFD, or 0xFFFD. My real goal here is to find out what the character is so that we can tell the people creating it, "Don't do that". They're not being much help in figuring it out. There is going to be a lot more of this in the future, which is why I'm trying to find out what the character is.

Similar Messages

  • Odd Tab Behavior in InDesign CS5

    I've used CS2 for quite a few years now, but decided it was time to upgrade to CS5. Doing so in the middle of a big book layout project was PROBABLY not the best idea.
    Essentially, I have laid out a document in this fashion:
    The headings are set 0.0" to the left of the column
    The body text is set to 0.05" from the left side of the column by using the indent marker on the left side of the tabs window. (The stacked double-arrow thing)
    I have created a tab stop for paragraphs to begin at 0.15" from the left. I did this instead of a pragraph indent because I wish for the first paragraph to be flush with the left of the rest of the text.
    This worked just fine under CS2, but when I open the document in CS5, the tab character mysteriously DOES NOT INDENT the text for no reason I can understand. Even more oddly, if I adjust the body text indent to another number, say, 0.0525", the tab indent miraculously works...but ONLY for the left hand column. The right hand column, even adjusted, will never show the indent.
    Also, even if I set the tab stop to something ridulous, like 1", all it accomplishes is making the line of the paragraph with the tab indent really stranglely justified, like:
    A           word              and            spaces
    Does anyone have any idea why this is happening?
    A screenshot of the issue: http://www.wiseturtle.com/InDesignTabs.gif
    Thanks in advance,
    Clay

    There are occasional reports of oddities when converting legacy docs to CS5. Export from  the original version to .inx and open that in CS5, save as a new file with a new name (don't overwrite the old files -- there's no way for you to go back) and build a new Book in CS5.
    As far as the indents, the correct way to handle this would be to make edit the body paragraph style to lose the tab and use a first line indent (you can use the GREP expression ^\t to find leading tabs and replace with nothing in the GREP Find/Change iin CS5), then make a second style for the first paragraph based onthe body style, but without the fisrt line indent, then sect the next style for that to the body style.

  • Image offsetting

    I'm playing around with the concept of 2D fighting game sprites. I have the JFrame, image loading, painting, and everything else fine.
    However, there is a slight problem. I'm trying to figure out how to drawe the images properly. Since the sprites can be different sizes depending on the pose, there might be slight problems when they are drawn.
    For example, I've taken out two idle stance sprites. One is 36 pixels wide and the other is 38 pixels wide. Both are the same height. When I draw it, I get a slight jittery motion since the two are different widths. I fixed that with two methods that calculate an offset and draw it in a position based on the original sprite. This fixes the jitteryness with the idle stance.
    So then I extract a sprite where the character is in the middle of a kick. I put that in place of the 38 pixel wide image for some testing. When this draws, the character mysteriously teleports backwards and does her kick. This is due to the offset methods.
    I think know where the problem lies, I just don't exactly have a solution. It has to be the offset methods.
    The code for them is..
    public int calculateHeightOffset(BufferedImage img1, BufferedImage img2) {
       int height1 = img1.getHeight();
       int height2 = img2.getHeight();
       return height1 - height2;
    public int calculateWidthOffset(BufferedImage img1, BufferedImage img2) {
      int width1 = img1.getWidth();
      int width2 = img2.getWidth();
      return width1 - width2;
    }I suppose my question would be... is there a way to determine how to properly draw different-size sprites codewise? Or would I need to make it all manual. I'm assuming there is a way to do it codewise because as far as I know that's how MUGEN works. But I could be wrong though.

    So far I have it set up like this:
    Class Fighter - holds a list of all the Sprite objects, the coordinates of the fighter, and later on.. moves and all that stuff.
    Class Sprite - holds a single BufferedImage and the offsets for the image.
    Once I get enough sprites, they will be ordered in a certain fashion. There isn't any particular order they will have to be in though. This is how I'm planning to make it all work:
    The character will have a file (files?) that describes move "ranges." These ranges are actually the numbers of the sprites to use. For example, a move with a range of 0-20 would use sprites 0 to 20 for its full animation. This will allow the sprites to be in a general order within the whole file, but they must be specifically ordered in order inside a range for the move to look right.
    The sprites themselves will be named 0.gif, 1.gif and so on. I suppose the TreeMap would hold the actual Sprite object for its value and an Integer key that corresponds with its name (0 for 0.gif, 1 for 1.gif, etc.).

  • Balance dropped, score dropped. FICO scores a mystery.

    They only reason to pay for the service of credit monitoring here on myFICO is so you have some idea of where you stand when you apply for any type of credit or loan.  All the information about understanding FICO scoring is all cookie cutter garbage that you can read on any other "credit monitoring" service.  And the people that work for myFICO have no idea what they are talking about either.  They just repeat the same cookie cutter garbage I was just mentioning.  I really hope they are providing their "services" for free, because if they receive any kind of compensation, it is undeserved.  My statement balance on my American Express Everyday card went from $50 to $0 and my scores dropped by 11, 12, and 17 points (Equifax, TransUnion and Experian.)  I brought the statement balance back up to $50 the VERY next month.  Did my scores recover the lost points I just lost a month ago?  No.  And I am sure all these "credit experts" will say, "something else in your report must have changed."  That is code for, "I have no idea what happened, I am just as clueless as you why your score changed the way it did."  The precise formula for generating your FICO score is a well guarded secret.  It is the same formula used in determining battle damage to your characters in the old Sega Master System game Phantasy Star.  Sometimes when attacked, your character would lose 6 health points, other times 80 health points, and even other times 3/4ths of your total health, and so on and so on.  This is how FICO scoring works.  The formula basically goes, OOGA BOOGA, OOGA BOOGA, this is your updated score.  The only advice on this site (again which you can find on EVERY other credit monitoring service) that is worth anything is what responsible credit habits you should be practicing.  Other than that, just hope for the best.  Your score is a mystery, and it always will be.

    Beavakos wrote:
    What difference does it make if there are any negatives on my report?  I lost points for doing something negative, which is bringing my balance to zero, than returning it back to its previous balance.  THAT IS ALL I DID.  I thought time heals all wounds when it comes to credit.  Any credit missteps I might have had taken should be corrected with time.  I really don't know how to make this any clearer.  Nothing in my report changed except TIME.  If I have two dollars in my bank account, and spend one dollar, then deposit one dollar, I should have now have TWO dollars AGAIN.  ?????????????????I understand what Beavakos is saying here and we all get it. Most of us love analyzing data which is why we are here after we've already cleaned up our reports and greatly increased our scores. Again, I understand how Beavakos feels - it is said time and again that UTIL has no memory and what is lost due to UTIL will be regained - well, not always true. Looking strictly at the data two things changed on Beavakos' report: 30 days were added to age and UTIL went from something above 0% to 0%, then back again and yet Beavakos' score did not fully recoup the loss. The reason for this, I think, is very clear - we do not know as much as we think we know about FICO or any other scoring algorithm. If I lost 10 points for going from 5% UTIL down to 0% UTIL but then gained those 5 points back by raising my UTIL to 5% again then scoring algorithm's are not as complicated as they are made out to be and are thus not worth paying for. The previous scenario is simple math and if indeed it was only that I should be able to retest that hypothesis time and again and get the same exact results - not happening. That said, I do not believe that scoring is a crap shoot or something only a witch doctor can understand. It is a complicated mathematical equation that has 100's, if not 1000's, of factors all having a different weight at a different time and I believe there is no one that can determine with 100% accuracy the effect any one factor, forget about multiple factors, will have on your score. Not even the people who wrote the code would know as no one person wrote the whole code. We should probably stop getting so frustrated with how our scores react to the different changes in our reports. The reason we get upset is that we think this is some simple black and white game and it's not. It is very complicated and we will never know what we want to know to the accuracy level we want to know it. What's important is we know what hurts our score to a high degree and what helps our score to a high degree. Take care of all those things we know about and your score will go up; don't take care of those things and your score will go down. After that it is just a game we play to see how we can get to 800...

  • Several mysterious problems with Solaris 11 x86 install

    Hi folks,
    I've recently installed Solaris 11 x86 as a replacement for Solaris 9 SPARC on an Ultra 2 machine. I like the operating system, but I've promptly acquired some mysterious problems. The computer is a Fujitsu Siemens Esprimo E2500 with a Pentium 4 HT 3.06 GHz processor, 2 GB RAM, an 80 GB SATA disk and a 147 GB SCSI disk.
    I have had a few occasions on which the screen goes totally blank, the monitor claims that there is no signal, and I have to restart the machine. However, with the blank screen, the hard drive activity light is sometimes still flashing. This happens predictably when I try to log off (as opposed to shutting down) and when I use Totem media player and click on anything inside the Totem window. The motherboard has built-in graphics and brief specifications can be found here:
    http://uk.ts.fujitsu.com/rl/servicesupport/techsupport/professionalpc/ESPRIMO/Datasheets/ds_esprimo_edition_e2500.pdf
    Yesterday my PS/2 keyboard started causing the motherboard to beep crazily when I use it. It beeps rapidly 4 times, then prints the expected character. However, the problem doesn't occur at the initial log in screen. The mouse cursor has also gone beserk on one occasion, flying all over the screen (I had this problem with a previous PC, but a different OS, and thought it was a hardware fault, which I why I got this new computer). With the keyboard problem, the machine is basically unusable. However, it is configured as a dual boot machine, and the keyboard problem isn't present with Windows XP.
    There are also a few smaller, odd problems. Occasionally all the icons vanish from my Gnome desktop. The command prompt also mysteriously changed from $chris@pluto to $bash-4.1 just before the keyboard problem appeared, although that might have had something to do with me copying my home directory to a different disk and editing the /etc/auto_home file (moving the home directory worked fine). Also, unlike with Solaris 9 SPARC, I find that I have to have external SCSI devices switched on at boot time, should I want to use them later. That's a pain. With Solaris 9, I could run "boot -r" when I first attached the device, then switch on the SCSI devices and mount them whenever I wanted.
    After the useful advice I received here about "more" and "less", I thought I'd post these problems and see if anyone can help. I'd really appreciate some advice. Can anyone identify likely causes for these problems? Which problems can I expect to solve and which do I have to live with?
    Many thanks,
    Chris Tidy
    Edited by: Chris Tidy on 23-Mar-2012 03:36

    Hi folks,
    I've recently installed Solaris 11 x86 as a replacement for Solaris 9 SPARC on an Ultra 2 machine. I like the operating system, but I've promptly acquired some mysterious problems. The computer is a Fujitsu Siemens Esprimo E2500 with a Pentium 4 HT 3.06 GHz processor, 2 GB RAM, an 80 GB SATA disk and a 147 GB SCSI disk.
    I have had a few occasions on which the screen goes totally blank, the monitor claims that there is no signal, and I have to restart the machine. However, with the blank screen, the hard drive activity light is sometimes still flashing. This happens predictably when I try to log off (as opposed to shutting down) and when I use Totem media player and click on anything inside the Totem window. The motherboard has built-in graphics and brief specifications can be found here:
    http://uk.ts.fujitsu.com/rl/servicesupport/techsupport/professionalpc/ESPRIMO/Datasheets/ds_esprimo_edition_e2500.pdf
    Yesterday my PS/2 keyboard started causing the motherboard to beep crazily when I use it. It beeps rapidly 4 times, then prints the expected character. However, the problem doesn't occur at the initial log in screen. The mouse cursor has also gone beserk on one occasion, flying all over the screen (I had this problem with a previous PC, but a different OS, and thought it was a hardware fault, which I why I got this new computer). With the keyboard problem, the machine is basically unusable. However, it is configured as a dual boot machine, and the keyboard problem isn't present with Windows XP.
    There are also a few smaller, odd problems. Occasionally all the icons vanish from my Gnome desktop. The command prompt also mysteriously changed from $chris@pluto to $bash-4.1 just before the keyboard problem appeared, although that might have had something to do with me copying my home directory to a different disk and editing the /etc/auto_home file (moving the home directory worked fine). Also, unlike with Solaris 9 SPARC, I find that I have to have external SCSI devices switched on at boot time, should I want to use them later. That's a pain. With Solaris 9, I could run "boot -r" when I first attached the device, then switch on the SCSI devices and mount them whenever I wanted.
    After the useful advice I received here about "more" and "less", I thought I'd post these problems and see if anyone can help. I'd really appreciate some advice. Can anyone identify likely causes for these problems? Which problems can I expect to solve and which do I have to live with?
    Many thanks,
    Chris Tidy
    Edited by: Chris Tidy on 23-Mar-2012 03:36

  • In Lr 3.6 how can I search for filenames ending with a # character ?

    The text search seems to treat the # character in some mysterious way so as to not select the files that have it in their name. I cannot find any explanation for this in the documentation. Is there some other character(s) that I have to put beside the # to make it register as a normal character ?
    My specific problem is that I need to locate a bunch of files in my catalog that have names ending with a # character. I want to rename those files in a way that removes the trailing # but not other # characters that separate key parts of the file names. If I can't do that then I want to select that group of files and produce .xmp files for them (I don't normally use .xmp files) so that I can rename the files outside Lr and import them again complete with all previous tweaks and metadata.
    I dare say that I would not choose to name files this way if I started again but I've got far too many to change now. 
    There are also too many for me to want to rename them individually in the Metadata panel if I can avoid it.
    I'm using Lr 3.6 on Mac OSX 10.6.8
    cheers,
    - Alan

    You guys beat me to the answer, which I discovered more by accident than good management.
    Firstly, I had used the # characters because they looked good in list views such as in Windows Explorer or the Mac Finder. That was from before I got into using Lr and I was trying and using several different programs to manage my photos. Then I found that trailing # characters prevented Lr from recognizing the number at the "end" of the file name and so I wanted to delete them.
    Secondly, I got into the Lr rename facility initially to replace what I did with Downloader Pro before switching from Windows to Mac and subsequently found that there is no way to delete a trailing character, but it helped establish in my mind that to Lr the "filename" was just the first part of the name before the full stop and the extension (e.g. .cr2 or .nef or .jpg). Try it - there is a "filename" token for setting up a naming structure that ignores the extension and its full stop.
    This was further reinforced in Library Grid View where the filename and the extension are separated and the extension is not displayed.
    So having used F2 (rename) before I got around to playing with the text search I was set up for failure. Looking for {filename ends with #} or {filename ends with #.} failed because it found nothing as none of my files had a # or #. after the filename extension, and looking for {filename contains #} also failed because it found all of my files instead of just those with the trailing #. I wrongly thought that Lr had trouble seeing a # in the text search.
    Now I have been able to find and select all files with a trailing # in the base part of the filename by looking for {filename contains #.} and generate current .xmp files for them. Then I went outside Lr and used a utility file renamer to get rid of the offending # for the image files and the .xmp files in bulk. Then I re-synchronized the folder(s) in Lr to import those renamed files back into Lr along with all of the metadata, and get rid of the old files from the catalog. Had those files been a part of a collection or stack then I suspect that membership of the collection or stack would have been lost but it kept the keywords, ratings, labels, and image tweaks. Then I got rid of the .xmp files again because I prefer to work without them (less clutter and less chance of other software separating them from the image files).
    So much easier than doing it the hard way
    Thanks for your help.
    - Alan

  • Calling stored procedure - mysterious error

    Hi all,
    I have a problem with one stored procedure.
    The procedure looks like this simplified sample:
    create or replace procedure sp_sample (
    p_input1 in char,
    p_input2 in number,
    p_input10 in char,
    p_in_out in out number
    as
    l_in_out number(38) := p_in_out;
    begin
    select count(t1.id_t1)
    into l_in_out
    from table1 t1
    where
    (field1 = p_input1) and
    (field2 = p_input2) and
    (field10 like p_input10);
    p_in_out := l_in_out;
    end;
    And now the problem:
    When I call the procedure in some PL/SQL block like this:
    declare
    l_cnt number(38);
    begin
    sp_sample(
    p_input1 => 'ABC',
    p_input2 => 2,
    p_input10 => '%',
    p_in_out => l_cnt
    dbms_output.put_line(l_cnt);
    end;
    -> I receive the error message ORA-01722 in the first line of the select ("select count(t1.id_t1) ...").
    But if I call the procedure in similar PL/SQL block like this:
    declare
    l_cnt number(38);
    l_char char(10 char);
    begin
    l_char := '%';
    sp_sample(
    p_input1 => 'ABC',
    p_input2 => 2,
    p_input10 => l_char,
    p_in_out => l_cnt
    dbms_output.put_line(l_cnt);
    end;
    -> I receive a correct result.
    The only difference is in the using of a local variable instead of a string literal!
    When I simplified the conditions in the where clause in the select, the calling of the procedure passed always.
    It seems the error is caused by complicated condition in the select and some mystery concerning of passing parameters into a procedure or binding variables.
    Has anybody any experience like this?
    Jiri

    The old invalid number problem.
    Its one of the following:
    1) you are using implicit conversions
    2) you have a misunderstanding of how the oracle optimizer works.
    Probably 2, as the query works with the hardcoded literal but not the variable.
    Basically, you are getting two separate query plans, ie oracle is using a different way of getting the result depending on whether it is using a literal or a variable.
    Basically in one scenario, you are encountering a non-numeric character and an attempt is being made to convert this to a number.
    In the other, you never encounter the non-numeric character

  • Chinese Character

    I am a new user of Blackberry 8300, actually the handheld was a gift from my colleague which she
    bought in Europe, I have chosen the mobile/internet service from Hutchison 3 company for both
    mobile phone and e-mail connection, everything is working properly except the e-mail description
    cannot be displayed or shown in Chinese Character, I've been seeking for the help of the Hutchison
    people as well as download the relevant softwear, follow the guideline of the installation process, the
    situation remained the same, the Chinese Character still displayed in black square dots.
    The question is, if the hand held was not from the local dealer, it cannot be changed to Chinese language?
    Second is, where can I find the software installation or hand held modification for Chinese version.
    Third, I found the new Blackberry 3G hend held on some websites and magazines, when you are going to
    launch theses models?
    Thanks!
    Kenneth. (Kmcchan)

    americangothic wrote:
    How I am going to install this is still a mystery as I have no idea how any of this BB software works.
    http://supportforums.blackberry.com/t5/BlackBerry-Device-Software/How-To-Reload-Your-Operating-Syste...
    The search box on top-right of this page is your true friend, and the public Knowledge Base too:

  • Mysterious FM- RH style mapping override with span tag

    I have been working in a super clean FM and RH project that is has one to one mappings for all my styles in the FM to same named styles in RH, just to be sure I do not mess up. Still, though, I seem to get a mystery mapping.  For my bullets and lists, RH is adding a span tag with font set to Times Roman.  I cannot find *any* css that has Times Roman. I've checked things like my character tags, my cross reference tags, etc and cannot find a correlation.  Where would this be added if it's not in my FM Format definitions and not in my RH Style definitions. I tried reimporting the book to see if it needed a tickle of the project mappings template but that didn't help.
    I'm attaching some images in hopes someone has ideas. The example does not appear to be times roman in Design view, but in more complex topics it's pretty obvious.
    I have so many hacks for single sourcing now, about ready to give up.
    Got help?
    FM 10
    RH 11

    Thanks for the fast reply Arnis,
    I did not even notice that the span tag ended on the spaces, so it can't be what's causing the odd previews I'm seeing.  So it seems my original question is not well thought out, so I'm going to retest with some other style changes.
    Still, though I'm wanting to follow up on getting rid of the Times tag anyway.  The bullets are always defined as Verdana. Bullet is set to use "Default Font" while BulletLevel2 uses "Bold" (the char def for Bold is also Verdana).  Is there some place I have to map "tabs" to not be Times? I dont think it's the Default Font issue because bullet 2 uses Bold.
    Here are some screenshots.

  • Confirmed Bug in Siri - Texting a character not found in 7-bit ASCII Table Causes SMS to Fail

    I am 99.9% certain I have singlehandedly discovered and solved a bug in Siri that causes SMS messages to fail (without a failure notice, i.e. red exclamation mark). This problem confounded me for several weeks. (Note: This issue does not affect iMessages.) After four hours of support phone calls to AT&T and Apple, 2 replacement iPhones (set up as new iPhones, and not restoring the backup), a SIM card replacement, a reprovisioning of my entire account with AT&T, and a very upset girlfriend who was understandably skeptical that I was really texting her when I said I was, I had no solution in sight, and my SMS woes persisted. I ended up solving the problem on my own through trial and error.
    THE PROBLEM:
    When I dictate an SMS message, if I dictate the word "elipsis" it rightly interprets what I have said as "..."
    However, as of today's date, Siri interprets "elipsis" an an extended ASCII version of "...", a version where the elipsis is actually one character and not three separate periods. The one character version of elipsis is not found in 7-bit ASCII.
    If you try to text a character that is not 7-bit ASCII, your text message will not be received by the recipient, regardless of the recipient's carrier or phone, and you will not get a failure notice.
    If you have any doubt whether Siri has interpreted "elipsis" as one character or as three, try deleting it. If the entire elipsis deletes simply by depressing delete one time, you have generated the single character (non 7-bit) elipsis, a character that cannot be texted, and which will "mysteriously" cause your text to fall into an abyss.
    THE SOLUTION:
    It seems that the permanent solution is that the Siri servers or the iOS needs to be reprogrammed only to respond with 7-bt ASCII characters when the user depresses the Siri button to dictate an SMS message.  In the meantime, until Apple solves this probem, the user should refrain from dictating any punctuation that will cause Siri to respond with a non 7-bit ASCII character.
    (Apple, I want a gift certificate for solving this problem.)

    Too bad Apple will never see this post since this is a user-to-user site.
    http://www.apple.com/feedback/iphone.html

  • Reflection applied to character/paragraph style?

    Can anyone point me in the right direction to create a character or paragraph style for using the web 2.0 reflection on text? example below...
    Thanks!

    Apologies not necessary. That previewColor property has been a mystery for quite some time, and when I finally found out what it is supposed to do, it seemed to be so bug-ridden that I quickly steered away from it and I gave up attempting to make any sense of it. But your script just avoids those bugs popping up!
    As I said in the IDS post: it's rather sad the Adobe team gave up on further developing this great hidden feature into a full one; and it's downright worrying that they left that legacy code in CS5 and later, so it could get infested with all those nasty little bugs. It leads me to think: how many more 'dead' code is there, in CS5, 5.5 and now in CS6? Could that be the reason later versions are so much slower and more buggy than my trusty CS4?
    (Point in fact.
    In CS4 and earlier on the Mac, all plugin resources are gathered into neat, small packages inside a single large indexed file. That makes sense, because every single plugin resource is just a couple of bytes long -- say, a single icon, or a single string "Menu Item Name". With CS5, the programmers suddenly took a Lazy Turn and wrote out every single resource as a single small file. That means that
    1. (Annoying!) Since hard disk space is allocated by the kilobyte (typically 4Kb chunks), 10 tiny resources totalling 1 Kb now will suddenly occupy 40 Kb of disk space. Sure, hard disk space is cheap. Apparently, it's even cheaper than doing some programming -- or even just using the load code from the last version!
    2. (Frustrating!) Locating, opening, loading, and then closing a single file takes time and resources, but after a file got loaded into memory, you can do with it whatever you want. Locating, opening, loading and then closing each of the dozens of small files per plugin takes lots of time. Is that the reason CS5.5 starts up so incredibly slow? Jongware sez: yeah for sure it won't help.
    /Point in fact).

  • Error while assigning a character value to a numeric variable.

    I fire a sql statement and check the number of rows returned by the sql.
    I check this result with the application logs.
    The application logs keeps the sqls fired by the application and the no of rows returned, in the example below the sql returned 8454 rows.
    My script compares the two results.
    ***********Application Log***********
    4/14/2008 11:15:01 AM: 0059 SQL SELECT "CLUSTER_CD",
    4/14/2008 11:15:01 AM: 0060 "PRODUCT_DESC",
    4/14/2008 11:15:01 AM: 0061 "TEAM_CD"
    4/14/2008 11:15:01 AM: 0062 FROM "OPS$TMS"."MAP_CLUSTER_TEAM_PROD"
    4/14/2008 11:15:01 AM:      3 fields found: CLUSTER_CD, PRODUCT_DESC, TEAM_CD, 8,454 lines fetched
    ***********Application Log***********
    My script:
    #!/bin/ksh
    typeset -i resA
    typeset -i resB
    opstms_conn_string="abc/[email protected]"
    set `sqlplus -s $opstms_conn_string << EOF
    set pages 0
    WHENEVER SQLERROR CONTINUE
    SELECT count(*)
    FROM MAP_CLUSTER_TEAM_PROD;
    exit
    EOF`
    resA=$1 ##returns 8454
    resB=`grep '3 fields found: CLUSTER_CD, PRODUCT_DESC, TEAM_CD,' BCVP_Main_Loader.qvw.log | awk '{print $10}'`
    ##resB returns 8,454
    ## here i get syntax error
    if [ $resA -eq $resB ]; then
    echo "QA passed for sql1"
    else
    echo "QA failed for sql1"
    fi
    The problem is as resB is integer variable it does not accept character value: 8,454 so returns a syntax error:
    How do I change the value assigned to resB into a numeric variable?
    error:
    + grep 3 fields found: CLUSTER_CD, PRODUCT_DESC, TEAM_CD, BCVP_Main_Loader.qvw.log
    run.ksh[52]: 8,454: syntax error

    Change:
    resB=`grep '3 fields found: CLUSTER_CD, PRODUCT_DESC, TEAM_CD,' BCVP_Main_Loader.qvw.log | awk '{print $10}'`
    to this:
    resB=`grep '3 fields found: CLUSTER_CD, PRODUCT_DESC, TEAM_CD,' BCVP_Main_Loader.qvw.log | awk '{print $10}' | tr -d ,`
    to drop the comma. Or you could do it in awk(1):
    resB=`grep '3 fields found: CLUSTER_CD, PRODUCT_DESC, TEAM_CD,' BCVP_Main_Loader.qvw.log | awk '{ gsub( /,/, "", $10 ); print $10}'`
    Or you could to it all in awk(1):
    resB=`awk '
    /3 fields found: CLUSTER_CD, PRODUCT_DESC, TEAM_CD/ {
    gsub( /,/, "", $10 )
    print $10
    ' BCVP_Main_Loader.qvw.log`
    (This example was not tested, it's just for a model.)
    Someone more of an SQL guru than I can probably tell you how to change your numeric locale to avoid presenting those commas in the first place and avoid the problem.
    HTH

  • Problem on print a report in mode character

    hi, i have oracle 8i and reports 6i, my reports it's generate in mode character, when i send to print (printer epson fx1180), the print is bad, (i configure the prt files example: bold esc "E", Unbold esc "F"),
    MY problem is with windows 2000, i send to print on windows 98 and all good, but now i have windows 2000, anybody help me to config the printers file for windows 2000

    You can create a PDF file using the [itext |http://www.lowagie.com/iText/] library
    you can create reports using jasperreports
    you can print the reports and save them as PDF,EXCEL ,HTML and more...
    (jasperreports is using the [itext |http://www.lowagie.com/iText/] library )
    you can edit your reports using ireport
    hope that helps

  • Mail compose window jumps to back when adding address. As soon as I add a single character to the address window of a new mail message, it pops  under all other existing windows. If I have others open...and who doesn't? I have to then close them all.

    So.....deeply....disappointed by Lion...
    It honestly gets in the way of getting things done.
    When was the last time you had an upgrade, in any program or  OS that caused an immediate productivity drop?
    Mail windows, when adding a new address, pop under to become the lowest window in the stack...before you are done composing...
    no extra clicks..just adding ONE character to the actual address bar does it.
    Trying to add files to an online content management system is hard...it's hard because when you attempt to add the file, it disappears from
    the list, so a search must be conducted...oh not a search of your hard drive...no...it has to include ALL FILES and that takes a while, showing
    files totally unrelated to what you need or in whatever that view is, showing everything from movies to documents, to a whole range of other stuff
    that is irrelevant to what is being sought.
    I can get used to the disappearing icon tool bar...but the message threading in Mail is so strangely bad it's hard to describe. So you send out an email,
    and it had 5 points...but that was 3 emails before...A reply comes in addressing those 5 points but saying something like ..."I disagree with point number 4"
    Unless you have total recall the thread shows you only incoming mail and the LAST most recent folded up message you sent. You are forced to go digging
    into sent mail to see what that point you made 8 days eariler actually was...
    I am pleased that I don't have to have Launchpad (honestly the silliest 'feature' I've ever seen on a computer) in the dock and see no point in the juvenile Mission Control (the naming scheme is odd and reminiscent of Junior Spaceman...how ironic that we end America's manned launch capability at the same time as Launch Pad and Mission Control, both throwbacks to the 60's are added to an OS). In any case both are virtually useless tools. One, replacing the mildly useful Spaces and the other just genuinely useless.
    I'm going to give Apple until the second bug fix...but I can tell right now that Snow Leopard was a far more robust OS. The thousands of posts about this may rise to the level of AntennaGate...I hope so...It seems the only way to get Steve (for whom I have now honestly lost respect for the FIRST TIME) to address the issues is to embarrass the company publicly. Lion is a broken system, poorly conceived...and proof that the Apple methodology of no widespread user testing can and does fail. This needs to get fixed. Soon.

    It seems to have something to do with profile manager.
    I get stack traces in the "system messages" logs for the "Server" application, grrrr.
    I'll get that info and attempt to submit a but report tonight.

  • Print landscape report in character mode

    Hi!
    It seems that I should change xxx.prt file in DESFORMAT parameter to match different kinds of printers in order to print landscape report in character mode.
    Our customers are international wide. How can I do to make this setting meet every customer for printing reports in landscape style?
    We will run the reports on Unix on the middle tier and customers read them in browser.
    Thanks for help.
    null

    Hi,
    This sounds like a similar problem that I had (and still having). It was (is) a Bug. Go to metalink.oracle.com and reference this bug#1413754. See if this is your problem as well. My only work around was change the format of the file instead of using character mode I am now using PDF. Which is working fine for me.
    ~Vannette

Maybe you are looking for