Tesseract OCR 1.0.1

system/tesseract-ocr-1.0.1 - 02 October 2018 - Add Windows support.

downloads.xml

<?xml version="1.0"?>
<!--Phoronix Test Suite v8.2.0-->
<PhoronixTestSuite>
  <Downloads>
    <Package>
      <URL>http://phoronix-test-suite.com/benchmark-files/ocr-image-samples-1.zip</URL>
      <MD5>3766fa00fa7df769216d2a89d55fb81d</MD5>
      <SHA256>185d7823be9ea257aa0613d9acfeb8b2276ea251a72ffba0f1a11e43dcda5853</SHA256>
      <FileName>ocr-image-samples-1.zip</FileName>
      <FileSize>48594677</FileSize>
    </Package>
  </Downloads>
</PhoronixTestSuite>

install.sh

#!/bin/sh

if which tesseract>/dev/null 2>&1 ;
then
	echo 0 > ~/install-exit-status
else
	echo "ERROR: Tesseract OCR is not found on the system! This test profile needs a working tesseract installation in the PATH. Check your system package manager for tesseract / tesseract-ocr."
	echo 2 > ~/install-exit-status
	exit
fi

tesseract --version | head -n 1 | awk '{ print $NF }' > ~/pts-test-version 2>&1

unzip -o ocr-image-samples-1.zip

echo "#!/bin/sh
tesseract --oem 1 ocr-sample-1.JPG output
tesseract --oem 1 ocr-sample-2.JPG output
tesseract --oem 1 ocr-sample-3.JPG output
tesseract --oem 1 ocr-sample-4.JPG output
tesseract --oem 1 ocr-sample-5.JPG output
tesseract --oem 1 ocr-sample-6.JPG output
tesseract --oem 1 ocr-sample-7.JPG output
echo \$? > ~/test-exit-status
tesseract --version | head -n 1 | awk '{ print \$NF }' > ~/pts-test-version 2>&1" > tesseract-ocr
chmod +x tesseract-ocr

install_windows.sh

#!/bin/sh

cmd /c "C:\Program Files (x86)\Tesseract-OCR\Tesseract.exe" --version | head -n 1 | awk '{ print $NF }' > ~/pts-test-version 2>&1

unzip -o ocr-image-samples-1.zip

echo "#!/bin/sh
cmd /c \"C:\Program Files (x86)\Tesseract-OCR\Tesseract.exe\" --oem 1 ocr-sample-1.JPG output
cmd /c \"C:\Program Files (x86)\Tesseract-OCR\Tesseract.exe\" --oem 1 ocr-sample-2.JPG output
cmd /c \"C:\Program Files (x86)\Tesseract-OCR\Tesseract.exe\" --oem 1 ocr-sample-3.JPG output
cmd /c \"C:\Program Files (x86)\Tesseract-OCR\Tesseract.exe\" --oem 1 ocr-sample-4.JPG output
cmd /c \"C:\Program Files (x86)\Tesseract-OCR\Tesseract.exe\" --oem 1 ocr-sample-5.JPG output
cmd /c \"C:\Program Files (x86)\Tesseract-OCR\Tesseract.exe\" --oem 1 ocr-sample-6.JPG output
cmd /c \"C:\Program Files (x86)\Tesseract-OCR\Tesseract.exe\" --oem 1 ocr-sample-7.JPG output
echo \$? > ~/test-exit-status
cmd /c \"C:\Program Files (x86)\Tesseract-OCR\Tesseract.exe\" --version | head -n 1 | awk '{ print \$NF }' > ~/pts-test-version 2>&1" > tesseract-ocr
chmod +x tesseract-ocr

results-definition.xml

<?xml version="1.0"?>
<!--Phoronix Test Suite v8.2.0-->
<PhoronixTestSuite>
  <SystemMonitor>
    <Sensor>sys.time</Sensor>
  </SystemMonitor>
</PhoronixTestSuite>

test-definition.xml

<?xml version="1.0"?>
<!--Phoronix Test Suite v8.2.0-->
<PhoronixTestSuite>
  <TestInformation>
    <Title>Tesseract OCR</Title>
    <Description>Tesseract-OCR is the open-source optical character recognition (OCR) engine for the conversion of text within images to raw text output. This test profile relies upon a system-supplied Tesseract installation.</Description>
    <ResultScale>Seconds</ResultScale>
    <Proportion>LIB</Proportion>
    <SubTitle>Time To OCR 7 Images</SubTitle>
    <TimesToRun>3</TimesToRun>
  </TestInformation>
  <TestProfile>
    <Version>1.0.1</Version>
    <SupportedPlatforms>Linux, BSD, Solaris, MacOSX, Windows</SupportedPlatforms>
    <SoftwareType>Utility</SoftwareType>
    <TestType>System</TestType>
    <License>Free</License>
    <Status>Verified</Status>
    <EnvironmentSize>50</EnvironmentSize>
    <ProjectURL>https://github.com/tesseract-ocr/tesseract</ProjectURL>
    <Maintainer>Michael Larabel</Maintainer>
    <SystemDependencies>tesseract</SystemDependencies>
  </TestProfile>
</PhoronixTestSuite>