loading tesseract causes RStudio to crash

[dwyker]

loading tesseract causes RStudio to crash

[samuel-chen-mil-navy]

Yes!! tesseract causes R 4.2.0 to crash. Works okay with 4.1.x. Tested on multiple machines.
Does package need new binaries on CRAN for new 4.2.0?

[thomasairbag]

I am experiencing this problem also.
R version: 4.2.0 (2022-04-22 ucrt)
Running: Windows 11 Home 21H2

[rafaelkiko]

Same problem here
tesseract 5.0.0
R version 4.2.0 (2022-04-22 ucrt)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running: Windows 11 Pro 21H2

[jeroen]

Hmm I cannot reproduce this. Seems to load fine on my Windows...

[samuel-chen-mil-navy]

Update: On both work and home machine, library(tesseract) take about 5 - 6 minutes to load. So R didn't crash. Just it takes an unreasonably long time to load. Can this be fixed? Home machine:

R version 4.2.0 (2022-04-22 ucrt)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 22000)

[jeroen]

I tried building a new version of tesseract. Can you please test if the problem still appears with tesseract 5.1.0? To install the new test version, run this in a clean empty r session

install.packages("tesseract", repos = 'https://ropensci.r-universe.dev')

[samuel-chen-mil-navy]

The package is still taking a long time to load, but maybe a little bit faster! On home machine, I deleted the tesseract directory from my windows library folder and installed from ropensci repo. I will test on work machine and report.

[samuel-chen-mil-navy]

The newly compiled package might even be slower. This morning, I timed with stopwatch (smartphone) and got approximately five and half minutes. Now...

beginTime = Sys.time()
library(tesseract)
endTime = Sys.time()
endTime - beginTime
Time difference of 6.699568 mins
sessionInfo()
R version 4.2.0 (2022-04-22 ucrt)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 19042)

[rafaelkiko]

still getting a fatal error and aborted session with the new package,
same R version, Platform and Windows as before

[jeroen]

This has been fixed now in the upstream libtesseract code: tesseract-ocr/tesseract#3830

[jeroen]

I have deployed a fixed version to r-universe, and will submit to CRAN soon:

install.packages("tesseract", repos = 'https://ropensci.r-universe.dev')