May I know what is this function about? This item is not mentioned/explained in the help file under “The Tools menu”.
Thanks
it should summarize all highlighted text in a new document.
I do not know how it can be used in practice, if I have a pdf with highlighted text, and I select “summarize highlights” nothing happens in DT3beta1… I did not find any explanation in the help doc.
1 Like
I see… I think this item is the same additional script provided by 2.n that will only work for .rtf document. It works in 3.0 beta if the document is a .rtf file (just checked).
oh, that’s interesting…no error message, no information in DT’s protocol, if I try to summarize hightlights of a pdf…
@cgrunenberg this should be improved (help file, and also error message if this feature is used in a non-supported file)
as written before in the forum: it would be simply great, if such a feature can be used selecting multiple pdf files, creating ONE summary file that inlcudes all highlighted sentences/paragraps with a reference link to each individual pdf file where the highlight has been made… (feature suggestion)
2 Likes
It worked (somewhat) for an indexed PDF that contained highlighted text, inasmuch as it extracted some words from each highlighted instance across all pages of the PDF. The summaries were nonsensical, but it was accurate.
It is missing in the documentation, but the function works when selecting highlighted PDFs or RTFs. It does not work with highlighted web content.
Select a highlighted PDF or two and choose the command.
- If the files are selected in a local smart group, a Summary document will be created in the root of the database.
- If the selected files are in the same group, the document will be created in that group.
@lutefish If you are experiencing unexpected behavior, please be more specific in the reporting.
1 Like
It works as described with highlighted rtf files, but not with highlighted pdfs (no summary file is created, no error message, no message in DT’s protocol)
Thanks to DT for the great effort to finally put in place this excellent utility.
For me, it works on both single and multiple selection of pdf files (kind=“PDF+Text”) in my indexed groups. The function worked smoothly almost every time except for some specific files.
I am guessing that the script is working as best as it can, but not all indexed pdfs are the same. If somehow the function seems not working, it may have to do with the OCR problems in some user-specific pdf+text files. For those who encounter issues may want to try using the function on other pdf+text files.
In the future update of 3.0, I wish we are allowed to modify the script, or by means of choosing an option, to (1) decide whether we want to paste the full name or the aliases of the file/s in the summary, and (2) able to choose to extract underlined text or highlighted text.
Updated results: the function also extracts underlined text, but displayed as a “blackened” block in the summary.rtf file. However, when I selected the block I can see the underlined text.
Just a bug, the next beta will fix this.
Just tried it a moment ago: It worked insofar as ist creates an .rtf with an entry, but this consists only of some words, actually a line of a three line highlight, and missed two other highlights completely (five lines long, two lines long). See attached the part of the pdf and the RTF.
You may want to use another/different pdfs for testing to check whether it’s a consistent pattern or is file specific? I tested the function on 50 pdf+text files and only 2 of them don’t work. Most of my files are generated by “print to pdf”, extracted from journal database, and OCRed journals (some older journal articles are not text indexed).
Also, some journal articles are intentionally scrabbled (as far as I am aware), even if you just copy and paste some selected text in a .rtf file, you won’t get what you want.
1 Like
I tried it now with a second pdf, and it worked like charme. 
By the way (feature request): It would be nice, if the generated RTF uses the PDF’s title as title/name automatically and not a simple “Zusammenfassung” (I’m German) , so the RTFs are better distinguishable.
This is not obvious, because the function works for both single and multiple selection?
Sorry, don`t geht what you mean?
If you select multiple files, the function will summarise all highlighted text of all selected files in one .rtf, with individual file name and extracted text and page index all listed in the .rtf. The script can’t aggregate all file names as title.
Okay, get it. Yeah, the feature request is only meant to one file, which is the most common case, I think.
Makes sense, thanks for the suggestion.
1 Like
Yeah, the feature request is only meant to one file, which is the most common case, I think.
Actually, it isn’t. The requests were actually skewed toward summarizing more than one document. Just FYI.
1 Like
in my case, I should have add 
1 Like