Acrobat DC Table Editor not recognizing cells

[Guy_I]

I'm in the midst of remediating a legacy document that was untagged. I don't have the original file(s) (probably Word) used to generate the PDF.

I have manually tagged each cell for a complex table, and indicated the header items. There is one row of major headings (4 cells) and one row of subheadings (14 cells). All of these cells are tagged <TH>. All remaining cells are tagged <TD>. In order to avoid blank cells, I inserted "n/a" text colored white, tagged each one as <TD> and put it into the correct display order with the TURO tool, and ensured it was in the correct spot in the table structure via the Tabs panel.

I went into the Table Editor to start setting cell properties so I can link the data cells to the correct heading cells, and found that the Table Editor is not showing all the items's tags, and it is combining some individual items into the same cell in the Table Editor. Attached screen shots show how each item is tagged (Legacy table tags.png), and how they appear in the Table Editor (Legacy table-table editor.png).

How do I separate the combined items in the Table Editor? 

[ynwtf]

Can anyone else see the attachments? I just get an error when I try to open them. Also, please anyone jump in and correct me if I'm wrong.

Without seeing what you have going, maybe take a look at your Tags panel and see what structure is there. You should end up with a structure similar to the following:

<Table>

     <TR>

          <TH>

              Move your Header cell text stuff here 1

          <TH>

               Move your Header cell text stuff here 2

          <TH>

               Move your Header cell text stuff here 3

          <TH>

               Move your Header cell text stuff here 4

     <TR>

          <TH>

               Move your Row header cell text stuff here (assuming you want a row header, else use a <TD> instead

          <TD>

               Move your Data cell text stuff here 2

          <TD>

               Move your Data cell text stuff here 3

          <TD>

               Move your Data cell text stuff here 4

When I type "move your header/data cell text here," I mean you can drag the tags from one location to another to structure your table elements within the parent <Table> tag. Without seeing what you  have, I'm only speaking very generally here. You can see your table cell information tagged as <P> or if you've already tagged each cell text, maybe you can see either <TH> or <TD> for each cell's text. IF you're not sure what's what, you can draw a selection frame around a cell text field in your page, then in your Tags panel use the Options icon (top left of the Tags Panel) to click "Find Tag from Selection." You may need to click it twice as it can be wonky sometimes. Anyway, this should jump to and highlight the tag that is currently associated with your selected text. Check it's tagging.

My guess is, either you do not have the parent <Table> tag in place. Or, if you do, then maybe not all the tagged items are correctly nested within in. Or that perhaps some cells haven't been associated to a tag yet. Using the Tag Panel to search tags from a selection can really help find missing objects like that.

Still, I'm only speculating without seeing it. But maybe this can get you going somewhat? Be sure to save a copy of your PDF in case you end up breaking the PDF file! I've done that at least a few hundred times 😉

[Guy_I]

Thanks for the quick response. 

All cells in the table are tagged either <TH> or <TD>.  The structure is correct in the Tags panel. 

<Table>

  <TR> (major headings)

    <TH>

    <TH>

    <TH>

    <TH>

  <TR> (subheadings)

    <TH>

        .

        .   (There are 14 subheadings)

        .

    <TH>

  <TR> (first data cell row)

    <TD>

        .

        .   (There are 14 data cells)

        .

    <TD>

There are 12 data rows. 

Dunno why the forum isn't displaying the attachments. PNGs should work, right? I can email them to you, if you like. 

[try67]

No, we can't. There's currently a bug in the forums system that makes attached files not work.
Upload the images inside your message, or share them via a file-sharing website.

[Guy_I]

Thanks for the pointer, try67. 

[Guy_I]

(screen shots removed)

[Bevi Chagnon - PubCom.com]

It appears that the table was incorrectly coded when it was created and you'll have to edit the coding (not necessarily the tags).

We have a recent blog on doing this at https://www.pubcom.com/blog/2020_04-05/ACROBAT_tableeditor.shtml

And FYI, when you edited the table's content by adding the "n/a," you could very well have corrupted the PDF coding and contributed to this problem. It's a known bug in accessible PDF by Adobe: edit the content and the accessibility is corrupted somewhere in the document, not necessarily in what you edited. Tables, lists, TOCs, and other complex tag elements are more likely to fail than ordinary <P> and <Hx> tags.

Recommendation: complete the editorial changes to the content, and then tag the PDF.

[Guy_I]

I knew better than to edit the existing text, but didn't realize that adding a new text box would cause a problem. As I said in my original post, I don't have the original files -- I'm told nothing is available except the PDF I'm working on -- so I can't add the "n/a" to the blank cells in the tables and re-PDF. . 

Or did you mean go through the PDF without any tags in it, add the "n/a" text boxes, and then tag everything after that? 

[Bevi Chagnon - PubCom.com]

We use 2 methods in these situations:

1. If the PDF is untagged, then make the editorial changes to the content and run Acrobat's AutoTag utility (use the most recent version of Acrobat Pro for this). If the PDF already has tags and they're terrible, then delete them, make the edits, and run AutoTag.
2. If you don't have the original source document, export the PDF to Word, make the edits in Word, clean up any other junk/mis-formatted stuff in Word, and re-export a new PDF.

Method #2 can be invaluable for some files. Some times it's quicker to fix many accessibility problems in a newer version of Word than it is to remediate a lousy PDF. And you end up with a recovered, accessible source file in addition to a compliant PDF.

These table problems usually are found in PDFs made with older, outdated software (outdated Word and/or outdated PDF Maker). Older software didn't make PDFs per the PDF/UA standard (yea yea, I know...we can debate how well the latest versions follow the standards, but MS and Adobe have definitely improved the end result). Table problems also appear when the page or table is rotated 90 degrees, for example rotated to fit landscape in an otherwise portrait document.

[Guy_I]

Thanks for the suggestions, Bevi. The document came to me untagged. Running AutoTag returned an absolute mess where these tables were concerned, so I decided to just tag it all manually. 

From what I can see in the file, it appears that the tables were generated by some database in a 'standard' format, probably direct to PDF with those pages added to the report after the rest of it was PDFed from Word. 

I already know that I won't be permitted to recreate via the PDF-to-Word-to-PDF train for various reasons. I'm not too far into the file, so my best bet is probably starting over from an untagged copy to add the "n/a" blocks where necessary to avoid empty cells and ensure regularity. 

Thanks again.

[LKL77]

Did I ever tell you that you're my hero? I have been wrestling with this situation for YEARS and this is the first time I ever found a decent explanation! THANK YOU!!!