Find words with soft hyphens

[Kristinita]

1. Related issues

1. Forum question
2. Issue tracker

2. Summary

It would be nice, if SumatraPDF will find words with soft hyphens in searchable documents.

3. Data

• KiraSuperhero.pdf — PDF file with correct OCR layer.

4. Argumentation

I often need to find something in the PDF files. In SumatraPDF I may not find the word, that I need, because it moves to the next line with soft hyphen, — I don't know where soft hyphen in the documents. Hyphens required in Russian language for wrapping words; I see them in any Russian scanned paper book. This problem important to me; therefore, I use another tools (see section 7 of this issue) instead of SumatraPDF.

5. Additional information

Soft hyphen (already known as »optional hyphen») is a symbol for word-breaking in line ends.

• About optional hyphens
• Soft hyphen in PDF

6. Actual behavior

SumatraPDF doesn't recognize soft hyphens:

- symbol required in search:

7. Expected behavior

Free programs versions, that have requested feature:

• Foxit Reader:

• PDF X-Change Editor:

• Okular (open source):

Thanks.

[user1823]

This feature has been implemented by MuPDF in 2020:
https://git.ghostscript.com/?p=mupdf.git;h=2185f16814074f024800a8bcc2dcf2f68ffcb07e

So, in my opinion, all that needs to be done is to enable this in SumatraPDF.

[GitHubRulesOK]

There is agreed a difference as shown below between MuPDF 1.20 and current SumatraPDF (you can see MuPDF skips over the end of line hyphen break but then it is not the same as when used for searching other forms such as hyphenated numbers etc. so in the second view Sumatra finds the hyphenated glyphs they are mixed but generally at end of line plain text (-)Tj (not soft, but it varies as there are mixed types) but MuPDF fails to see them, since they are classed as non existent parts of a word.

[user1823]

I don't understand Russian. But, Google Translate shows the same translation for the word whether the hyphen is included or not.

It may be a problem, but I don't know.

I think that the best solution for this issue would be to add a setting controlling whether to ignore such hyphens or not so that the users can decide on their own what works best for them.

I don't think that doing this would require a significant amount of effort (I am not a dev though).