From 0f0bae18ba4c25ed6c609097519bce1bf4d2e941 Mon Sep 17 00:00:00 2001
From: "Gerber, Mike" <mike.gerber@sbb.spk-berlin.de>
Date: Tue, 4 Feb 2020 19:29:56 +0100
Subject: [PATCH] =?UTF-8?q?=E2=9C=85=20Remove=20GT=20text=20to=20not=20acc?=
 =?UTF-8?q?idently=20check=20it=20instead=20of=20OCR=20text?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 test/test_recognize.py | 20 ++++++++++++++++----
 1 file changed, 16 insertions(+), 4 deletions(-)

diff --git a/test/test_recognize.py b/test/test_recognize.py
index 11b34ad..e576ac5 100644
--- a/test/test_recognize.py
+++ b/test/test_recognize.py
@@ -3,6 +3,7 @@ import shutil
 import subprocess
 import urllib.request
 from lxml import etree
+from glob import glob
 
 import pytest
 from ocrd.resolver import Resolver
@@ -10,10 +11,15 @@ from ocrd.resolver import Resolver
 from ocrd_calamari import CalamariRecognize
 from .base import assets
 
+
 METS_KANT = assets.url_of('kant_aufklaerung_1784-page-block-line-word_glyph/data/mets.xml')
 WORKSPACE_DIR = '/tmp/test-ocrd-calamari'
 CHECKPOINT = os.path.join(os.getcwd(), 'gt4histocr-calamari/*.ckpt.json')
 
+# Because XML namespace versions are so much fun, we not only use one, we use TWO!
+NSMAP = { "pc": "http://schema.primaresearch.org/PAGE/gts/pagecontent/2019-07-15" }
+NSMAP_GT = { "pc": "http://schema.primaresearch.org/PAGE/gts/pagecontent/2013-07-15" }
+
 
 @pytest.fixture
 def workspace():
@@ -44,12 +50,20 @@ def workspace():
         ff = os.path.join(WORKSPACE_DIR, 'OCR-D-IMG', f)
         subprocess.call(['convert', ff, '-threshold', '50%', ff])
 
+    # Remove GT Words and TextEquivs, to not accidently check GT text instead of the OCR text
+    for of in workspace.mets.find_files(fileGrp="OCR-D-GT-SEG-LINE"):
+        workspace.download_file(of)
+    for to_remove in ["//pc:Word", "//pc:TextEquiv"]:
+        for ff in glob(os.path.join(WORKSPACE_DIR, "OCR-D-GT-SEG-LINE", "*")):
+            tree = etree.parse(ff)
+            for e in tree.xpath(to_remove, namespaces=NSMAP_GT):
+                e.getparent().remove(e)
+            tree.write(ff, xml_declaration=True, encoding="utf-8")
+
     return workspace
 
 
 def test_recognize(workspace):
-    # XXX Should remove GT text to really test this
-
     CalamariRecognize(
         workspace,
         input_file_grp="OCR-D-GT-SEG-LINE",
@@ -81,8 +95,6 @@ def test_word_segmentation(workspace):
     assert os.path.exists(page1)
     tree = etree.parse(page1)
 
-    NSMAP = { "pc": "http://schema.primaresearch.org/PAGE/gts/pagecontent/2019-07-15" }
-
     # The result should contain a TextLine that contains the text "December"
     line = tree.xpath(".//pc:TextLine[pc:TextEquiv/pc:Unicode[contains(text(),'December')]]", namespaces=NSMAP)[0]
     assert line