projections.xml and sspsTemplate.xsl do not match

[mdales]

In sspsTemplate.xsl we are looking for XML of the form:

 <horizontallines>
  <g xmlns="http://www.w3.org/2000/svg">
   <line x1="0.0" y1="43.0" x2="1277.0" y2="43.0" style="stroke:black;stroke-width:1.0;"/>
  </g>
  <g xmlns="http://www.w3.org/2000/svg">
   <line x1="782.0" y1="413.0" x2="1262.0" y2="413.0" style="stroke:black;stroke-width:1.0;"/>
  </g>
 </horizontallines>

But in the generated projections.xml we have:

 <g class="horizontallines" xmlns="http://www.w3.org/2000/svg">
  <line x1="0.0" y1="33.0" x2="1266.0" y2="33.0" style="stroke:red;stroke-width:2.0;"/>
  <line x1="758.0" y1="310.0" x2="1252.0" y2="310.0" style="stroke:red;stroke-width:2.0;"/>
 </g>

As a result the ami-forestplot --segment stage fails, as there are missing border values.

[mdales]

I have a fix for this @petermr what do you want me to do with it?

[petermr]

I am currently working with stata forest plots and the second projections.xml is the latest version. The extraction of lines is messy and depends on thresholds. I want to get it working for stata and then transfer the logic to spss. Ideally it should be possible to create a single stylesheet with different modes. The problem is that the 'ami-imageis in flux and I am updating thestatabut nospss`.

Is this critical? does a build fail? Or a test corpus? If so I can address it. But the current problem is getting the right projections.xml content!

[mdales]

Yes, changes like this, where you break thing to move forward, do stop me moving forward.

Breaking changes should only hit master once complete and tested - you actually have tests this breaks, so it should be obvious, no (AMIForestPlotTest.java spss tests surely must fail?)

I've been tasked currently with trying to come up with some way of repeatedly running PDFs through and I can't do that if master breaks so significantly.

[petermr]

We probably need a branch strategy for you and me. We should talk about this tomorrow.

OK - so this breaks the CI? **what tests are you running in the CI? I can then try to make sure that we run the same tests and that when I commit it won't break those tests. I am happy in principle to run the default tests (but not the IT ones).

Will this solve the problem?

On Wed, Jul 3, 2019 at 2:59 PM Michael Dales notifications@github.com wrote:

Yes, changes like this, where you break thing to move forward, do stop me moving forward.

-- Peter Murray-Rust Founder ContentMine.org and Reader Emeritus in Molecular Informatics Dept. Of Chemistry, University of Cambridge, CB2 1EW, UK

[petermr]

I've run mvn install on cephis and there are failures and errors. I may remove the failing tests if they are exploratory or environment dependent.

[ERROR] Failures:

[ERROR]   IntegerMultisetListTest.testBins:26 bins expected:<[[4 x 3, 5 x
2[], [6 x 2, 7, 8], [9], [12 x 2], [17, 15], [], [23], [24 x 2, ]25]]> but
was:<[[4 x 3, 5 x 2[, 6 x 2], [7, 8, 9], [12 x 2], [15], [17], [], [23, 24
x 2], []25]]>

[ERROR]   CMFileUtilTest.testRenameCaseInsensitive:57 target/A.txt

[ERROR]   ForestTest.testForestPlotBitmap:75 islands expected:<187> but
was:<100>

[ERROR]   ImageUtilTest.testScaleImagePMR:125 size expected:<1705925> but
was:<2111126>

[ERROR]   DiagramAnalyzerTest.testCrossing:138 expected:<4> but was:<1>

[ERROR]   DiagramAnalyzerTest.testFunnelSegments:88 island expected:<3560>
but was:<7505>

[ERROR]   ImagePanelTest.testMakePanels:31->writeSubimage:43 images

[ERROR]   PixelIslandTest.testBoundingBoxes:497 islands expected:<5> but
was:<0>

[ERROR]   PixelRingTest.testClipVertical:260 pixelList expected:<1637> but
was:<1648>

[ERROR]   PlotTest.testDefaultSharkPlotCLI:84 pixels 15331

[ERROR]   DarwinTest.testDarwin0big:56 darwin size expected:<171> but
was:<100>

[ERROR]   DarwinTest.testDarwinThin:76 darwin main tree size
expected:<2352> but was:<10467>

[ERROR] Errors:

[ERROR]   CrossrefTest.testCreateDownloadAgro:62 » Connect Operation timed
out

[ERROR]   ImageProcessorTest.testCommandLineDefaults:57 NullPointer

[ERROR]   ImageProcessorTest.testGetPixelIslandThroughCommandLine:108 »
Runtime Cannot w...

[ERROR]   ImageUtilTest.testSplitImages:205 » Runtime section lines out of
range

[ERROR]   ImageMomentGeneratorTest.testImageMomentGenerator:17 » NullPointer

[ERROR]   ExamplesTest.testExtractPhyloTree:322 » NullPointer

[ERROR]   ExamplesTest.testMaltoryzineEdges:60 NullPointer

[ERROR]   PixelIslandTest.testBoundingBox2:487 NullPointer

[ERROR]   PixelIslandTest.testgetTerminalMaltoryzine0:139 » NullPointer

[ERROR]   PixelIslandTest.testgetTerminalMaltoryzine1:127 » NullPointer

[ERROR]   PixelListFloodFillTest.testCreatePixelIslandsFromPixelList »
NullPointer

[ERROR]   PixelListTest.testCurvaturesBent:246 NullPointer

[ERROR]   PixelListTest.testCurvaturesStraight:236 NullPointer

[ERROR]   PixelListTest.testJoinAcyclicPaths:258 NullPointer

[ERROR]   PixelListTest.testJoinCyclicFragments:297 » NullPointer

[ERROR]   PixelListTest.testJoinLineFragments:273 » NullPointer

[INFO]

[ERROR] Tests run: 2384, Failures: 12, Errors: 16, Skipped: 207

[INFO]

most of the numerical results are extremely sensitive and should be reduced to simple ranges

This test takes 5 mins for me (probably quicker for MD).

On Wed, Jul 3, 2019 at 3:41 PM Peter Murray-Rust < peter.murray.rust@googlemail.com> wrote:

We probably need a branch strategy for you and me. We should talk about this tomorrow.

On Wed, Jul 3, 2019 at 2:59 PM Michael Dales notifications@github.com wrote:

Yes, changes like this, where you break thing to move forward, do stop me moving forward.

Breaking changes should only hit master once complete and tested - you actually have tests this breaks, so it should be obvious, no (AMIForestPlotTest.java spss tests surely must fail?)

I've been tasked currently with trying to come up with some way of repeatedly running PDFs through and I can't do that if master breaks so significantly.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ContentMine/normami/issues/14?email_source=notifications&email_token=AAFTCS6CUGVSXJ264365FS3P5SWCXA5CNFSM4H5EPAQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZERDDA#issuecomment-508105100, or mute the thread https://github.com/notifications/unsubscribe-auth/AAFTCS34AZFB2NUFJDFKS6TP5SWCXANCNFSM4H5EPAQQ .

-- Peter Murray-Rust Founder ContentMine.org and Reader Emeritus in Molecular Informatics Dept. Of Chemistry, University of Cambridge, CB2 1EW, UK

-- Peter Murray-Rust Founder ContentMine.org and Reader Emeritus in Molecular Informatics Dept. Of Chemistry, University of Cambridge, CB2 1EW, UK

[petermr]

Thanks @mdales This was not a Java test issue - it was running the script in

/normami/src/test/java/org/contentmine/ami/tools/testStack.sh

this tests the ?Stata? files. I will amend the stylesheet so that it passes and alert you. see /normami/src/main/resources/org/contentmine/ami/tools/stataTemplate1.xsl and will merge both

Michael, where is the script you are running?

[petermr]

The following test in AMIForestPlotTest will be run before push'ing

    @Test
    public void testRegression() {
        testSPSSSimple();
        testStataSegmentAndAssert();
        testStataStack();
    }